Chip Dolby Decoder

STA310
6+2-CH. MULTISTANDARD AUDIO DECODER

PRELYMINARY DATA
1 FEATURES
■ DVD Audio decoder:
❚ Meridian Lossless Packing (MLP), with
up to 6 channels,
❚ Uncompressed LPCM with 1-8 channels,
❚ Precision of up to 24 bits and sample rates
of between 44.1 kHz and 192 kHz.
■ Dolby Digital (*) decoder:
❚ Decodes 5.1 Dolby Digital Surround. TQFP80
❚ Output up to 6 channels. downmix modes: ORDERING NUMBER: STA310
1, 2, 3 or 4 channels.
■ MPEG -1 2- channel audio decoder, layers I and
II.
■ MPEG-2 6-channel audio decoder, layer II. ■ 2.5V (for core) and 3V (for I/O) power supply.
❚ 3V Capable I/O Pads .
❚ 24 bits decoding precision.
■ MP3 (MPEG layer III) decoder. ■ True-SPDIF input receiver supporting AES/
EBU, IEC958, S/PDIF.
■ Accepts MPEG-2 PES stream format for:
❚ No external chip required.
MPEG-2, MPEG-1, Dolby Digital and linear
PCM. ❚ Differential or single ended inputs can be
decoded.
■ Karaoke System.
■ Prologic decoder. APPLICATIONS
■ Downmix for Dolby Prologic compatible. ■ High-end audio equipment.
❚ A separate (2-ch) PCM output available for ■ DVD consumer players.
simultaneous playing and recording.
■ Set top box.
■ Bitstream input interface: serial, parallel or
SPDIF. ■ HDTV .
■ SPDIF and IEC-61937 input interface. ■ Multimedia PC.
■ SPDIF and IEC-61937 output interface.

■ PLL for internal PCM clock generation. (*) “Dolby “, “AC-3” and “ProLogic” are
frequencies supported: 44.1KHz family (22.05, trademarks of Dolby Laboratories.
88.2, 176.4) and 48KHz family (24, 48, 96, 192).
■ PCM: transparent, downsampling 192 to 96 Khz
and 96 to 48kHz. DESCRIPTION
■ PTS handling control on-chip.
The STA310 is a fully integrated Audio Decoder ca-
■ No external DRAM required pable of decoding all the above listed formats.
■ I2C or parallel control bus Encoded input data can be entered either by a serial
■ Embedded Development RAM for (I2S or SPDIF) or a parallel interface. A second input
customizable software capability. data stream (I2S) is available for micro input.
■ Configurable internal PLLs for system and The control interface can be either I2C or a parallel 8-
audio clocks, from an externally provided clock. bit interface. No external DRAM is necessary for a to-
■ 80-PIN TQFP package tal of 35ms surround delays.
June 2003 1/90

This is preliminary information on a new product now in development. Details are subject to change without notice.
STA310
2 STA310 AUDIO DECODER PIN DESCRIPTION
Pin Number Name Type Function
CONTROL INTERFACES
48 IRQB O (1) Interrupt Signal (level), active low
47 SELI2C I (2) Selects the Control Interface (when high: serial interface; when
low: parallel interface)
I 2C Control Interface
43 SDAI2C I/O (1) I 2C Serial Data
46 SCLKI2C I I 2C Clock
53 MAINI2CADR I (2) Determines the slave address
Parallel Control Interface
78 - 79 - 80 - 1 D0 - D1 - D2 - D3 I/O Host Data

2-3-6-7 D4 - D5 - D6 - D7
12 - 13 - 14 - 15 A0 - A1 - A2 - A3 I Host Address
16 - 18 - 19 - 20 A4 - A5 - A6 - A7
21 DCSB I Chip Select, active low
22 R/W I Read/Write Selection: read access when high, write access

when low
35 WAITB O (3) Data Acknowledge, active low
DATA INPUT INTERFACE
First Serial Data Interface (I2S)

37 DSTRB I Clock Input Data, active low
41 SIN I Serial Input Data
40 LRCLKIN I Word Clock for the Input
42 REQ O Handshake for the Data Transfer, aconfigurable by the

SIN_SETUP register
Second Serial Data Interface (I2S)

62 DSTRB2 I Clock Input Data, active low
60 SIN2 I Serial Input Data
61 LRCLKIN2 I Word Clock for the Input
63 REQ2 O Handshake for the Data Transfer, active low
DATA OUTPUT INTERFACES
69 PCMCLK I/O Oversampling Clock input for STA310 when generated externally
DAC Interface
67 SCLK O Bit Clock for the DAC
2/90
STA310
2 STA310 AUDIO DECODER PIN DESCRIPTION (continued)
68 LRCLK O Word Clock for the DAC
72 PCM_OUT0 O Data from a Prologic downmix (VCR_L/VCR_R)
73 PCM_OUT1 O Data for the first DAC (Left/Right)
76 PCM_OUT2 O Data for the second DAC (Centre/Sub)
77 PCM_OUT3 O Data for the third DAC (LeftSur/RightSur)
IEC958 Interface (S/PDIF) - One Output Port., One Input Ports.
58 I958OUT O S/PDIF Signal
25 SPDP I First differential input of S/P DIF port
24 SPDN I Second differential input of S/P DIF port
26 SPDF I External Filter
28 VDDA I Analog VDD for S/P DIF Input port
29 GNDA I Analog GND for S/P DIF Input port
STATUS INFORMATION
PCM Related Information
54 SFREQ O Then high, indicates that the sampling freq. is either 44.1Khz or
22.05Khz.
When low, indicates that the sampling frequency is either 32 Khz,
48 Khz, 24 Khz or 16Khz.
57 DEEMPH O Indicates if de-emphasis is performed.
Audio Video Synchronization
59 PTSB O Indicates that a PTS has been detected, active low.
Other Signals
31 CLK I Master Clock Input Signal.
36 RESET I(2) Reset signal input, active low.
52 TESTB I(2) Reserved pin: to be connected to VDD
49 SMODE I Reserved pin : to be connected to GND
RS232 Interface
8 RS232RX I
9 RS232TX O
PLLs INTERFACES
64 CLKOUT O System clock output with programmable division ratio
27 PLLAF I External Filter For Audio PLL.
3/90
STA310
2 STA310 AUDIO DECODER PIN DESCRIPTION (continued)
30 PLLSF I External Filter For System PLL.
Power and Ground
5 - 11 - 23 - 33 - GND GND Ground

39 - 45 - 50 - 56
- 66 - 71 - 75
4 - 17 - 34 - 38 VDD VDD 2.5V Power Supply
44 - 55 - 65 - 74
10 - 32 - 51 - 70 VDD3 VDD3 3.3V Power Supply
Notes (1) Open Drain

(2) Internal Pull-up
(3) Tri-State
PIN CONNECTION (Top view)

PCM_OUT3
PCM_OUT2
PCM_OUT1
PCM_OUT0
LRCLKIN2
PCMCLK
CLKOUT
DSTRB2
LRCLK
REQ2
VDD3
SCLK
GND
GND
GND
VDD
VDD
D2
D1
D0
80 79 78 77 76 75 74 73 72 71 70 69 68 67 66 65 64 63 62 61
D3 1 60 SIN2
D4 2 59 PTSB
D5 3 58 I958OUT
VDD 4 57 DEEMPH
GND 5 56 GND
D6 6 55 VDD
D7 7 54 SFREQ
RS232RX 8 53 MAINI2CADD
RS232TX 9 52 TESTB
VDD3 10 51 VDD3
GND 11 50 GND
A0 12 49 SMODE
A1 13 48 IRQB
A2 14 47 SELI2C
A3 15 46 SCLKI2C
A4 16 45 GND
VDD 17 44 VDD
A5 18 43 SDAI2C
A6 19 42 REQ
A7 20 41 SIN
21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
SPDF
PLLAF
VDDA
GNDA
PLLSF
CLK
VDD3
GND
VDD
WAITB
HRSTB
DSTRB
VDD
GND
LRCKLIN
DCSB
HRWB
GND
SPDN
SPDP
D00AU1225
4/90
STA310
ABSOLUTE MAXIMUM RATINGS

Symbol Parameters Value Unit
Vdd 2.5V Power Supply Voltage -0.5 to 3.3 V
2.5V Input or Output Voltage -0.5 to (Vdd+0.5) V
Vdd3 3.3V Power Supply Voltage -0.5 to 4 V
3.3V Input or Output Voltage -0.5 to (Vdd+0.5) V
ELECTRICAL CHARACTERISTICS (VDD = 3.3V +/-0.3V; Tamb = 0 to 70°C; Rg = 50 Ω unless otherwise spec-
ified
DC OPERATING CONDITIONS
Symbol Parameters Value Unit
Vcc Power Supply Voltage 2.5 V
Tj Operating Junction Temperature -20 to 125 °C
GENERAL INTERFACE
Symbol Parameters Conditions Min Typ Max Unit Note
Iil Low level input current without Vi = 0V 1 µA 1
pull-up device
Iih High level input current without Vi = Vdd 1 µA 1
pull-down device
Ioz Tri-state output leakage without Vi = 0V or Vdd 1 µA 1
pull-up/down device
Ilatchup I/O Latch-up current V<0V, V>Vdd 200 mA 2
Vesd Electrostatic protection Leakage <1µA 2000 V 3
Note: 1. The leakage currents are generally very small, <1nA. The value given here, 1µA, is a maximum that can occur after an Electrostatic
Stress on the pin.
2. V> Vdd3 for 3.3V buffers.
3. Human Body Model
LVTTL & LVCMOS DC Input Specification 2.7V <Vdd3 <3.6V
Vil Low level input voltage 0.8 V 1
Vih High level input voltage 2.0 V 1
Vilhyst Low level threshold input falling 0.8 1.35 V 1
Vihhyst High level threshold input rising 1.3 2.0 V 1
Vhyst Schmitt trigger hysteresis 0.3 0.8 V 1
Note: 1. Takes into account 200mV voltage drop in both supply lines.
2. X in the source/sink current under worst case conditions and is reflected in the name of the I/O cell according to the drive capability
5/90
STA310
ELECTRICAL CHARACTERISTICS (continued)

Ipu Pull-up current Vi = 0V -66 0.8 µA 1
Rpu Equivalent Pull-up resistance Vi = 0V 50 KΩ

Note: 1. Min condition: VDD = 2.7V, 125°C Min precess Max condition: VDD = 3.6V, -20°C Max
POWER DISSIPATION
PD Power Dissipation Sampling frequecy ≤ 24KHz t.b.d. mW 1
@VDD = 2.4V
Sampling frequecy ≤ 32KHz t.b.d. mW 1
Sampling frequecy ≤ 48KHz t.b.d. mW 1
INTRODUCTION
The STA310 is a fully integrated multi-format audio decoder. It accepts as input, audio data streams coded with
all the formats listed above.
Host I/F I2C

8 PCM
2 I2S
AC3 STA310
DVD Audio
MPEG
MMDSP+ S/P DIF
S/P DIF
2.1 Inputs and Outputs

2.1.1 Data Inputs
- Through a parallel interface (shared with the control interface)
- Through a serial interface (for all the I2S formats)
- Through a S/P DIF (SPDIF or IEC-61937 standards).
- Trough a second, independent,I2S (for application like i..e. Karaoke mixing).
2.1.2 Data outputs

- The PCM audio ooutput interface, which provide:
PCM data on 4 outputs:
• Left/Right,
• Centre/Subwoofer
• Left Surround/Right Surround.
6/90
STA310
• Data From a Prologic downmix (encoder)

“Lrclk” “Sclk” “PcmClk”
- S/P DIF Output
2.1.3 Control I/F

I2C slave or parallel interface:
The device configuration and the command issuing is done via this interface. To facilitate the contact with the
MCU, 2 interrupt lines (IRQB and INTLINE) are available.
3 ARCHITECTURE OVERVIEW
3.1 Data flow

The STA310 is based on a programmable MMDSP+ core optimized for audio decoding algorithms.
Dedicated hardware has been added to perform specific operations such as bitstream depacking or IEC data
formatting.
The arrows in Figure 3 indicate the data flow within the chip.
The compressed bitstream is input via the data input interface.
Data are transferred on a byte basis to the FIFO. This FIFO allows burst input data at up to 33Mbit/s.
The input processor, which is composed of a packet parser and an audio parser, unpacks the bitstream (Packet
parser) and verifies the syntax of the incoming stream (audio parser).
The compressed audio frames with their associated information (PTS) are stored into the circular frame buffer.
While a second frame is stored in the circular frame buffer, the first frame is extracted by the audio core decoder
which decodes it to produce audio samples.
The PCM unit converts the samples to the PCM format. The PCM unit controls also the channel delay buffer in
order to delay each channel independently.
In parallel, the IEC unit transmits non compressed data or compressed data according to the selected mode. In
the compressed mode, the data are extracted directly from the circular buffer and formatted according to the
IEC-61937 standard. In non compressed mode, the left and right PCM channels formatted by the PCM unit are
output by the IEC unit, according to the SPDIF standard
Figure 1. Architecture and data flows
1 INPUT DATA 2 FIFO INPUT

DATAIN INTERFACE 256 x 8 PROCESSOR
3
IEC958
IEC958
(1937)
CIRCULAR FRAME BUFFER FORMATER
HOST OUT
INTERFACE
4
CONTROL,
STATUS 5 8
CORE AUDIO PCM UNIT PCMOUT
CLOCKS
DECODER
7 6
CHANNEL DELAY
BUFFER (35ms)
7/90
STA310
3.2 Functional diagram
Figure 2. Audio decoder top level functional diagram
STA310 IEC 1937 (AC-3 / MPEG 2)
FORMATTER
STA310
SPDIF mode switch

IEC958
NULL DATA 58 I958OUT
DCSB 21
6
IRQ 48 Pink Noise Gen
Beep Tone Gen
PCM
2 Switctch
SDAI2C 43
MP3
SCLKI2C 46 CONTROL
DOWN-SAMPLING
MAINI2CADR 53 LPCM
video
96/48kHz
Prologic Decoder
L DELAY
Bass Redirection / Vol ctrl

A[0..7] L/Lt
PCM 1..4 L
D[0..7] 73 PCMOUT1
PCMmixing
R DELAY
2 2 2
R
R/Rt C DELAY
SWITCH
SIN 41 CDDA
PES PARSER
PES PARSER
C lfe 76 PCMOUT2
LRCLKIN 40 FRAME DELAY
I2S_IN1 MPEG 1 C
BUFFER Layer 1-2
DSTRB 37 Lfe Ls
lfe DELAY
REQ 42 6
AC-3
Ls
Rs
77 PCMOUT3
Downmix
DELAY
6 1..6 Ls
6
PACKET MPEG 2 Rs
Rs
FORMATTER PTS
2 to 6 ch
LVCR DELAY 72 PCMOUT0

MLP Downmix Lt/Rt
2/0
2to2 RVCR DELAY
6to2
SIN2 60 63 SCLK
LRCLKIN2 61 Voice Effects: Sample Gain 68 LRCLK

I2S_IN2 Echo, Rate System
Level Sensitive 64 CLKOUT
DSTRB2 62 Chorus Converter 2 and
Reverb Cancel Audio
REQ2 63 Clocks 31 CLK
69 PCMCLK
3.3 Control interface description

The IC can be controlled either by a host using an I²C interface, or by a general purpose host interface.
These interfaces provide the same functions and are described in the following sections. The selection is per-
formed by the means of the pin SELI2C: when high, this pin indicates that the I²C interface is used. When low,
the parallel interface is used.
3.3.1 Parallel control interface

When the pin SELI2C is low, the control of the chip is performed through the parallel interface. When accessing
the device through the parallel interface, the following signals are used:
- The address bus A[7..0]. It is used to select one of the 256 register locations.
- The data bus DATA[7..0]. If a read cycle is requested, the data lines D[7:0] will be driven by the IC.
For a write cycle, the STA310 will latch the data placed on the data lines when the WAIT signal is
driven high.
- The signal R/W. It defines the type of register access: either read (when high), or write (when low).
Some registers can be either written or read, some are read only, some are write only.
- The signal DCSB. A cycle is defined by the assertion of the signal DCSB.
Note: 1. The address bus A[7..0], and read/write signal R/W must be setup before the DCSB line is activated.
8/90
STA310
- The signal WAIT. This signal is always driven low in response to the DCSB assertion.
The timing diagrams for the parallel control interface are given in Electrical specifications on page 5.
3.4 I2C control interface

When the pin SELI2C is high, the chip is controlled through the I²C interface. The I²C unit works at up to 400kHz
in slave mode with 7-bit addressing.
- The Pin MAINI2CADR selects the device address. When MAINI2CADR is high the slave address is
0x5C, when low the device address is equal to the value on the address bus (A0...A6).
- The pin SDAI2C is the serial data line.
- The pin SCLKI2C is the serial clock.
The I²C Bus standard does not specify sub-addressing. There are thus potentially multiple ways to implement
it. Any implementation that respects the standard is of course legal but a particular implementation is used by
many companies. The following paragraphs describe this implementation.
3.4.1 Protocol description

For write accesses only, the first data which follows the slave address is always the sub-address.
This is the one and only way to declare the sub-address. It should be noticed that the sub-address is implement-
ed as a standard data on the I²C Bus protocol point of view. It is a sub-address because the slave knows that it
must load its address pointer with the first data sent by the master.
See in the Appendix X.x for I2C message format examples.
3.5 Decoding process

The decoding process in the STA310 is done in several stages:
- Parsing,
- Main decoding,
- Post decoding,
- Bass redirection,
- Volume and Balance control.
Each of the stages can be activated or bypassed according to the configuration registers.
Parsing
The bitstream parsing (performed by the input processor) is in charge of discarding all the non audio information
in order to transmit to the next stage (the circular frame buffer) only the audio elementary stream (AC3, MPEG1/
2, LPCM, PCM, DVD Audio).
The parsing stage operates in two phases: the packet parser unpacks the stream, the audio parser checks the
syntax of the bitstream.
Main Decoding
The input of this stage is an elementary stream, the outputs are decoded samples. The number of output chan-
nels is defined by the downmix register (1 channel up to 6 channels). For details, please refer to the description
of the register.
The decoding formats currently supported are AC3, MPEG1 layers I and II, MPEG2 layer II, LPCM. It is neces-
sary to select the appropriate stream format by configuring the registers STREAMSEL and DECODESEL before
running the decoder.
9/90
STA310
Post Decoding
The post decoding includes specific PCM processing: DC filter, de-emphasis filter, downsampling filter. These
filters can be independently enabled or disabled through the register DWSMODE.
It provides also a Pro Logic decoder, which is described in detail in a next section.
Bass Redirection
This stage redirects the low frequency signals to the subwoofer.
The subwoofer is extracted from the other channels (L, R, C, Ls, Rs, LFe). There are six possible configurations
to extract the subwoofer channel, which can be selected thanks to the OCFG register.
Volume and Balance Control

The volume is a master volume (no independent control for each channel). It is controlled by the PCMSCALE
register, which enables to attenuate the signals by steps of 2dB.
Two balance controls are available: one for Left/Right channels, one for Left Surround/Right Surround channels.
They are configurable by means of registers BAL_LR (Left-Right Balance) and BAL_SUR (Left Surround-Right
Surround Balance), which provide attenuation of signals by steps of 0.5dB.
4 OPERATION
4.1 Reset
The STA310 can be reset either by a hardware reset or by a software reset:
- The hardware reset is sent when the pin RESET is activated low during at least 60ns. This is equiv-
alent to a power-on reset.
This resets all the configuration registers, i.e. PLL registers (PLLSYS, PLLPCM), Interrupt registers
(INTE, INT, ERROR), interface registers (SIN_SETUP, CAN_SETUP) and command registers
(SOFTRESET, RUN, PLAY, MUTE, SKIP_FRAME, REPEAT_FRAME).
- The software reset is sent when the register SOFTRESET is written to 1 (the register is automatically
reset once the software reset is performed). It resets only the interrupt related registers (INTE, INT,
ERROR) and the command registers (SOFTRESET, RUN, PLAY, MUTE, SKIP_FRAME,
REPEAT_FRAME). All other decoding configurations are not changed by softreset.
Some information concerning the post-processing are anywayt of date after a soft-reset
Note: 1. The chip must be soft reset before changing any configuration register.
10/90
STA310
4.2 Clocks
There are two embedded PLLs in the STA310: the system PLL and the PCM PLL.
The following is the block diagram of the system and audio clocks used in the STA310
Figure 3. PLL Block Diagram

CLKOUT CLK RXN RXP PCMCLK
/N
sys_clockout
pcmclk_en
PLL Sys PLL Audio SPDIF
/2
/2
78
plls_config
sys_clk
pcm_clk
SCLOCK
LRCLK
PCM_OUT
PCMOUT0,1,2,3
DSP Core
Periph 1 Periph 2 Periph 3
R I W
Figure 4. Block Diagram of Functional PLL
analog part Filter (external)
ClkIn Ip
DIV N+1
PFD Charge
(27MHz) Pump
R
C3 C
pll_disable
DIV M+1 VCO

Uvco
dN
Switching
Frac Circuit DIV (X+1)
update_frac
Oclk
11/90
STA310
4.2.1 System clock

The system clock sent to the DSP core and the peripherals can be derived from 4 sources and the selection is
performed through an Host Register; external clock, external clock divided by 2, internal system PLL and inter-
nal system PLL divided by 2.
The system PLL is used to create the system clock from the input clock. This PLL is software programmable
through the Host Registers mechanism. The system PLL is used to set the any frequency up to the maximum
allowed device speed. After hard reset the system clock is running at 47.25MHz. An RC network must be con-
nected to the filter Pin PLLSF.
The system clock is output on the pin CLKOUT after a programmable divider ranging from 1 to 16.
4.2.2 DAC clocks
4.2.2.1 PCM clock

The PCM clock can be either input to the device or generated by the internal PLL or recovered by the embedded
SPDIF receiver. The selection is done via the Host Registers.
After a hardware reset, the internal PLL is disabled and the PCMCLK pad is an input. PCMCLK may be equal
to the PCM output bit rate, or it may be an integer multiple of this, allowing the use of oversampling D-A con-
verters.
The internal fractional PLL is able to generate PCMCLK at any “FsX Oversampling Factor” frequencies, where
Fs is any multiple or sub-multiple of the two 44.1kHz and 48kHz sampling frequencies. An RC network must be
connected to the filter pin PLLAF; refer to External circuitry on page 9 for recommended values.
If the PCMCLK is recovered from the embedded SPDIF receiver, the only supported overampling frquency is
128 Fs.
4.2.2.2 Bit clock SCLK
The PCM serial clock SCLK is the bit clock. It provides clocks for each time slot (16 cycles for each channel in
16-bit mode, 32 cycles for each channel in 18-, 20-, 24-bit modes). The frequency of SCLK is therefore fixed to
2 x Nb time slots x Fs, where Fs is the sample frequency.
The clock is derived from the clock PCMCLK. The register PCMDIVIDER must be configured according to the
selected output precision and the frequency of PCMCLK, so that the device can construct SCLK:
Fsclk = Fpcmclk / (2 x (PCMDIVIDER+1)) gives
Table 1.
PCM Divider Value Mode Description
5 PCMCLK = 384 Fs, DAC is 16-bit mode
3 PCMLK = 256 Fs, DAC is 16-bit mode
The value of PCMDIVIDER = 0 is reserved. If this number is loaded, the divider is bypassed and the frequency
of SCLK equals the frequency of PCMCLK. The PCMDIVIDER register must be setup before the output of SCLK
starts.
This can be done by first disabling PCM outputs, by de-asserting the MUTE and PLAY commands and then
writing into the PCMDIVIDER register. Once the register is setup, the MUTE and/or PLAY commands can be
asserted. PCMDIVIDER can not be changed “on the fly”.
12/90
STA310
4.2.2.3 Word clock LRCLK

The frequency of LRCLK is given by:
- Flrclk = Fsclk/32; for 16 bit PCM output,
- Flrclk = Fsclk/64; for 18, 20 or 24 bits PCM output.
No special configuration is required. The polarity can be changed in the register PCMCONF, by setting up the
field INV as needed.
4.3 Decoding states

There are two different decoder states: Idle state and decode state (see <Blue HT>Figure 3). To change states,
register
Figure 5. Decoding States

Time
Idle mode Init mode Decode mode
Soft reset Run command Decoder ready to play sample
Idle Mode
This is the state entered after a hardware or software reset. In this state, the embedded DSP does not decode,
i.e. no data are processed. The chip is waiting for the RUN command, and during this state all configuration
registers must be initialized. In this state, even if the chip is not processing data, the DACs clocks can be output,
which enables to setup the external DACs. Once the PCMCLK, SCLK and LRCLK clocks are configured, it is-
possible to output them by setting the MUTE register.I
Table 2. Idle mode. play and mute commands effects

Play Mute Clock (SCLK, LRCLK) State PCM Output
X 0 Not running 0
X 1 Running 0
Note: 1. The PLAY command has no effect in this state as the decoder is not running. It can however be sent and it will be taken into account
as soon as the decoder enters the decode state.
Decode Mode
This state is entered after the RUN command has been sent (i.e. RUN register = 1). In this mode, the data are
processed. The decoder can play sound, or mute the outputs, by using the PLAY and MUTE registers:
- To decode streams, the PLAY register must be set. When decoding, the sound will be sent to outputs
if the MUTE register is reset. The outputs are muted if the MUTE register is set.
- To stop decoding, the PLAY register should be reset. Resuming decoding is performed by writing
PLAY to 1 again
13/90
14/90
STA310
1
1
0
0
Play
SEL_I2C MAIN_I2C_ADR REQ
PCM Out Block

1
0
1
0
HOST INTERFACE NULL

DATA SPDIF IEC958
Mute
SDA_I2C MODE FORMATTER SPDIF_O

SCL_I2C SWITCH
WAIT IEC 1937 (AC3/MPEG 2/DTS)

4.4 Data input interface description.
Figure 6. Block Diagram of Data Flow

HOST HOST
DCSB CONTROL REGISTERS
256 BYTE
RWB PCM
A0 to A7 SWITCH
R0
D0 to D7
FRAME
8 BUFFER PCMOUT1
&
DMA DBIT/ PCMOUT2
3 FIFO R1
3 W NBIT PCMOUT3
Running
Running
Running
LRCLKIN
I2S DSP
Not running
Clock State
DSTRB 8 CORE PCMOUT0

PACKET AUDIO
SIN
PARSER PARSER
Table 3. Decode Mode. Play and Mute commands effects
SPDIF1_A AUDIO SCLK

STA120 &
SPDIF1_B First Data Input Stream LRCLK
SYSTEM
CLOCKS PCMCLK
Second Data Input Stream
3 8
CLKOUT
I2S in
LRCLKIN2 TEST CLK
INTERFACE
DSTRB2 PLLs RESET
0
0
0
SIN2
PCM Output
REQ2 TEST SMODE D00AU1228

Decoded Samples
isters can be changed “on-the-fly”: PCM_SCALE, BAL_LR, BAL_SUR, OCFG, DOWNMIX registers.
No
No
Yes
Yes
Decoding
Note: 1. It is not possible to change configuration registers in this state. It is necessary to soft reset the chip before. Only the following reg-
STA310
Two independent inputs are available on the STA310.

The main one allows to enter input data stream through through:
- A serial interface (referred to as Data Serial Interface),
- And a parallel interface (referred to as Data Parallel Interface).
The choice is performed by the register SIN_SETUP.
4.4.1 Data serial interface

When the serial mode is selected, the bitstreams can be entered into the STA310 through either:
- a four-signal data interface or ,
- trough a SPDIF input (no external circuit is required).
The four-signal data interface (see Figure 5) provides:
- An input data line SIN,
- An input clock DSTR,
- A word clock input LRCLKIN
- And a hand-shake output signal REQ.
-
Note: 1. Only 16-bit PCM streams are supported. For 20-bit or 24-bit PCM, the 4 or 8 least significant bits are ignored
.
The specifications of those signals can be configured by the means of the register CAN_SETUP.
Two modes exist in serial mode, one that uses the LRCLKIN pin and one that does not use the LRCLKIN pin.
4.4.1.1 Modes without the LRCLKIN pin
In this mode the signal LRCLKIN is not used by the STA310. The input data SIN is sampled on the rising edge
of DSTR. When the STA310 input buffer is full the REQ signal is asserted. The polarity of REQ signal is pro-
grammable through the register SIN_SETUP. The data must be sent most significant bits first.
When the decoder cannot accept further data the REQ is de-asserted and the DSTR clock must be stopped as
soon as possible to avoid data loss. After the REQ is de-asserted, the decoder is still able to accept data for a
limited number of clock cycles.
The maximum number of data that can be transmitted with respect to the change of REQ is given by the follow-
ing formula: Nbits = 23 - 6 * FDSTR/33MHz, where: FDSTR is the DSTR clock frequency, (max is 33 MHz).
4.4.1.2 Modes using the LRCLKIN pin
When receiving data from an A/D converter or from an S/PDIF receiver, the signal LRCLKIN is used.
The LRCLKIN signal is used to make the distinction between the left and right channels. Any edge of the LR-
CLKIN signal indicates a word boundary.
The data transfer between the input interface and the FIFO is done on a byte basis. After the edge (rising or
falling) of the LRCLKIN, a new byte is transferred to the first stage of the STA310 every 8 DSTR clock cycles.
If the number of time slots is not a multiple of 8, the remaining data is lost. The polarity of LRCLKIN and DSTR
is programmable.
The LRCLKIN can be delayed by one time slot, in order to support PCM delayed mode. All these configurations
are programmable through the CAN_SETUP register.
The register CAN_SETUP has 4 significant bits, and each bit has a specific meaning, see CAN_SETUP on page
41.
Only the first byte is transferred to the STA310 because the number of time slots is 12 (8 + 4). SIN and LRCLKIN
are sampled on the falling edge of DSTR In this case SIN_SETUP = 3 and CAN_SETUP = LeftFirstChannel +
FallingStrobe + AllSlot = 2 + 4 + 8 = 14
15/90
STA310
Table 4.
When Set When Clear Name
Bit 0 The input data is one slot delayed with The input data is not delayed DelayMode
respect to LRCLKIN
Bit 1 First channel when LRCLKIN is set First channel when LRCLKIN is reset LeftFirstChannel
Bit 2 Data are sampled on falling edge of Data are sampled on rising edge of FallingStrobe
DSTR DSTR
Bit 3 All the bytes are extracted Only the first 16 data bits are extracted AllSlot
Figure 7.
00 00 00
0 0 0
LRCLKIN
00 00 00
DSTR 00 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 000 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00
00 00 00
00000000000000000000000000000000000000000000000000000000000000000
BIT 7
BIT 6
BIT 5
BIT 4
BIT 3
BIT 2
BIT 1
BIT 0
BIT 7
BIT 6
BIT 5
BIT 4
BIT 7
BIT 6
BIT 5
BIT 4
BIT 3
BIT 2
BIT 1
BIT 0
BIT 7
BIT 6
BIT 5
BIT 4
SIN
0 0 0
Transferred data Discarded data
Example 2: Only the first 2 bytes are transferred to the STA310 because the number of slots is 20 (16 + 4). SIN
and LRCLKIN are sampled on the falling edge of DSTR. The data is in delayed mode.
The register configuration is SIN_SETUP=3 and CAN_SETUP = DelayMode + LeftFirstChannel + FallingStrobe
+ AllSlot = 1 + 2 + 4 + 8 = 15.
This mode is a specific mode where only the first 16 data bits are transferred. The remaining bits are discarded.
The register configuration is SIN_SETUP = 3 and CAN_SETUP = DelayMode + FallingStrobe = 1 + 4 = 5.
4.4.1.3 SPDIF Input

A true SPDIF Input SPDIF (PCM audio samples) or IEC-61937 (compressed data) is selectable as a main serial
input.
4.4.1.4 Autodetected formats

The STA310 cut 2.0 is able the following audio format changes on the s/pdif input
Table 5. Audio Format detection
BEFORE AFTER
AC3 PCM
AC3 MPEG
MPEG AC3
MPEG PCM
PCM AC3
PCM MPEG
16/90
STA310
4.4.1.5 Second Input

A second independent input allows to input bitstreams in serial mode.
This second input can be used, to input audio stream from a microphone, while we decode a data stream trough
the main input.
4.4.2 Data parallel interface

Two ways are available to input data in parallel mode:
- Either through the parallel data bus, shared with the external controller,
- Or through the DATAIN register
4.4.2.1 Using the parallel data bus

In this mode the data must be presented on the 8-bit parallel host data bus D[7..0]. Note that this bus is shared
with the external controller. On the rising clock of DSTR the data byte is sampled by the STA310. The signal
REQ is used to signal when the input FIFO is full. When REQ is de-asserted the transfer must be stopped to
avoid data loss.
After the REQ is de-asserted, the decoder is still able to accept data for a limited number of clock cycles.
The maximum number of data that can be transmitted with respect to the change of REQ is given by the follow-
ing formula: Nbits = 23 - 6 * FDSTR/33MHz, where: FDSTR is the DSTR clock frequency, (max is 33 MHz).
The signals DSTR and DCSB are used to make the distinction between Stream Data (strobed by DSTR) and
Control Data (strobed by DCSB). To avoid conflicts, the DSTR signal and the DCSB signal must respect given
timing constraints.
4.4.2.2 Using the DATAIN register

The data can be input by using the control parallel interface as if accessing any other register.
The signal DCSB is therefore used. When using this register to input data stream, there is no need to byte-align
the data.
Figure 8.
00 00
00 00
LRCLKIN
00 00
0 0 000 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0000 0 0
DSTR
0 0 000 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 000 0 0
4
7
6
5
4
3
2
1
0
7
6
5
4
3
2
1
0
7
6
5
4
7
6
5
4
3
2
1
0
7
6
5
4
3
2
1
0
7
6
5
4
7
SIN
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
0 Transferred data Discarded data

0
17/90
STA310
Figure 9.
00 00
00 00
LRCLKIN
00 00
DSTR 00 00
00 00
4
7
6
5
4
3
2
1
0
7
6
5
4
3
2
1
0
7
6
5
4
3
2
1
0
7
6
5
4
3
2
1
0
7
6
5
4
7
00 00
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT
SIN
0 Transferred data Discarded data

0
4.5 Streams parsers
The parsing stage is operated by two parts: the packet parser and the audio parser.
The packet parser unpacks stream, sorts packets and transmit data to the audio parser. The audio parser ver-
ifies the stream syntax, extracts non-audio data and sends audio data to the frame buffer.
Packet parser
Before unpacking packets and transmitting data, the packet parser needs to detect the packet start by recog-
nizing the packet synchronization word. It is possible to force the parser to search for two packet synchronization
words before starting to unpack and transmit.
This is done by setting the register PACKET_LOCK to 1. Otherwise, the packet parser will start handling the
stream once it has detected information matching the packet synchronization word.
The packet parser is also able to perform selective decoding: it can decode audio packets that are matching a
specified Id. This Id is specified in AUDIO_ID and AUDIO_ID_EXt registers, and the function is enabled by set-
ting the AUDIO_ID_EN register.
Audio parser
The audio parser needs to detect the audio synchronization word corresponding to the type of stream that must
be decoded. It is possible to force the audio parser to detect more than one synchronization word before pars-
ing.
This is done by setting the SYNC_LOCK register to a value between 1 and 3 - number of supplementary sync
words to detect before considering to be synchronized.
The status of synchronization of both parsers is provided in the register SYNC_STATUS. Each time the syn-
chronization status of one of the two parsers changes, the interrupt SYN is generated (if enabled) and the status
can be read in SYNC_STATUS.
4.6 Decoding modes

4.6.1 AC-3
The STA310 is Dolby Digital certified for class A products. The decoder must be programmed so to specify the
stream format as AC-3 encoded: register DECODESEL = 0.
In the sections below are provided the modes specific to the AC-3 decoding.
4.6.1.1 Compression modes

Four compression modes are provided in the STA310:
18/90
STA310
- Custom A (also named custom 0 in Dolby specifications),

- Custom D (also named custom 1 in Dolby specifications),
- Line mode,
- RF mode.
These modes refer to different implementation of the dialog normalization and dynamic range control features.
The mode is selected by programming the register COMP_MOD to the appropriate value.
Line Mode
In Line Mode (COMP_MOD = 2), the dialog normalization is always enabled. It is done by the decoder itself and
the dialog is reproduced at a constant level.
The dynamic range control variable encoded in the bitstream is used and can be scaled by the two scaling reg-
isters HDR (for high-level cut compression) and LDR (for low-level boost compression). In case of 2/0 downmix,
the high-level cut compression is not scalable.
RF Mode
In RF Mode (COMP_MOD=3), the dialog normalization is always performed by the decoder. The dialog is re-
produced at a constant level.
The dynamic range control and heavy compression variables encoded in the bitstream are used, but the com-
pression scaling is not allowed. This means that the HDR and LDR registers can not be used in this mode. A
+11dB gain shift is applied on the output channels.
Custom A Mode
In Custom A mode (COMP_MOD=0), the dialog normalization is not performed by the decoder and must be
done by another circuit externally.
The dynamic range control variable encoded in the bitstream is used and can be scaled by the two scaling reg-
isters HDR (for high-level cut compression) and LDR (for low-level boost compression).
Custom D Mode
In Custom D mode (COMP_MOD=1), the dialog normalization is performed by the decoder. The dynamic range
control variable encoded in the bitstream is used and can be scaled by the two scaling registers HDR (for high-
level cut compression) and LDR (for low-level boost compression).
4.6.1.2 Karaoke mode

The AC-3 decoder is karaoke aware and capable.
A karaoke bitstream can be composed of 5 channels: L for Left, R for Right, M for guide Melody, V1 for vocal
track 1 and V2 for Vocal track 2.
- When in karaoke aware mode, the channels L,R and M are reproduced, and the channels V1 and V2
are reproduced at a level fixed by the bitstream.
- When in karaoke capable mode, it is possible to choose to reproduce one, two or none of the two
incoming vocal tracks, V1 and V2.
The karaoke decoder is activated by the use of KARAMODE register, which specifies the downmix for the dif-
ferent modes. This register replaces DOWNMIX register. It is however possible to consider the incoming
karaoke channels as any other multichannel stream and output it with a downmix specified in DOWNMIX reg-
ister. For details, refer to the Digital Audio Compression AC-3 ATSC standard, annex C.
19/90
STA310
4.6.1.3 Dual Mode

The Dual Mode corresponds to a mode where two completely independent mono program channels (e.g. bilin-
gual) are encoded in the bitstream, referenced to as channel 1 and channel 2.
The possible ways to output channels on left/right outputs are:
- Output channel 1 on both L/R outputs,
- Mix channels 1 and 2 to monophonic and output on both L/R,
- Output channel 1 on Left output, and channel 2 on Right output.
This channels downmix is specified in the register DUALMODE.
4.6.2 MPEG
The STA310 is able to decode MPEG-1 layerI and layerII encoded data, as well as MPEG-2 layer I, layer II data
without extension (i.e. 2-channel streams).
The MPEG input format should be specified in the DECODESEL register:
- DECODESEL=1 for MPEG1. The MC bit in MC_OFF register should be set.
- DECODESEL=2 for MPEG2. The MC bit in MC_OFF register should be set.
4.6.3 MP3
The STA310 is able to decoder MPEG2 layer III (MP3) data.
The MP3 input format aboved be specified in the DECODESEL register:
- DECODESEL=9 for MP3.
4.6.3.1 Dual Mode

The Dual Mode corresponds to a mode where two completely independent mono program channels (e.g. bilin-
gual) are encoded in the 2-channel incoming bitstream, referenced to as channel 1 and channel 2.
The audio decoder allows to:
- Mix channels 1 and 2 to monophonic and output on both L/R,
- Output channel 1 on Left output, and channel 2 on Right output.
The output configuration is chosen by special downmix for dual mode through register MPEG_DUAL.
20/90
STA310
4.6.3.2 Decoding flow
Figure 10. AC-3 Decoding Flow
PCM_OUT
6-Channel AC-3 Data
L L L L Delay
Data Input Interface
Volume, Balance
Bass Redirection
Fifo 256 Bytes
Packet Parser
AC-3 Decoder
Frame Parser
R R R R
Frame Buffer
Delay
Downmix
PCM_OUT
C C C C Delay
LFe LFe Sub Sub
Delay
Ls Ls Ls Ls
PCM_OUT
Delay
Rs Rs Rs Rs
Delay
Figure 11. MPEG Decoding Flow
2-Channel MPEG1/2 Data
PCM_OUT
L L
L
MPEG1/2 Decoder
Delay
Bass Redirection
Volume, Balance
Packet Parser
Fifo 256 Bytes
R
Frame Parser
L R R
Frame Buffer
Delay
Downmix
PCM_OUT
Sub Sub Zeros
Delay
R
PCM_OUT
Zeros
4.6.4 PCM/LPCM
The decoder supports PCM (2-channels) and LPCM Video (8-channels) and Audio (6-channels) streams. This
is selected by DECODESEL=3.
4.6.4.1 Downsampling filter
When decoding PCM/LPCM streams encoded at 96kHz, it is possible to use a filter that downsamples the
stream from 96kHz to 48kHz. The chip can not output streams at 96kHz. The register DWSMODE is used to
configure the use of this filter.
Figure 12. PCM/LPCM Decoding flow
2-Channel PCM/LPCM Data
PCM_OUT0
L
Downsampling Filter
L L
Delay
Bass Redirection
Volume, Balance
96kHz -> 48kHz
Packet Parser
Fifo 256 Bytes
Frame Parser
R R R
Frame Buffer
Delay
PCM_OUT1
Zeros
Sub Sub
Delay
PCM_OUT2
Zeros
21/90
STA310
4.6.5 MLP
MLP is a lossless coding system for use on digital audio data originally represented as linear PCM. MLP is man-
datory in DVD Audio. It allows transmission and storage of up to 6 channels. each up to 24 bits precision and
with sample rates between 44.1 KHz and 192KHz.
- DECODESEL = 8
4.6.6 CDDA
- DECODESEL = 5
4.6.7 Beep Tone

- DECODESEL = 7
4.6.8 Pink noise generator

The pink noise generator can be used to position the speakers in the listening room so to benefit of the best
listening conditions.
The decoder must be programmed so to generate pink noise by writing 4 in the DECODESEL register. The
DOWNMIX register is used to select independently the channels on which the pink noise will be output.
When generating pink noise, the output configuration should be: OCFG=0 and PCM_SCALE=0.
Figure 13. Pink Noise Generator Flow

PCM_OUT0
L L
Pink Noise Generator
No Bass Redirection:
R R
PCM_OUT1
C C
Downmix
Pink
ocfg = 0
LFe LFe
Noise
Ls Ls
PCM_OUT2
Rs Rs
4.7 Post Processing

The following post processing alghorithms are available
4.7.1 Prologic
Pro Logic Compatible Downmix
The STA310 can decode an AC-3 multichannel bitstream and encode it to provide a 2-channel Pro Logic com-
patible output (Lt, Rt). These 2 channels are the result of a specific downmix referred to as Pro Logic compatible.
This downmix is selected by the register DOWNMIX. The 2 channels can be used as the input of a Pro Logic
decoder and player (e.g. home theatre).
Pro Logic Decoding

The STA310 can decode a 2-channel Pro Logic bitstream. The 2 channels could come from a CD player, an
AC-3 2-channel bitstream or an MPEG1 bitstream. The 2-channel bitstream can be converted into a 4-channel
output (L, R, C, S). The surround (S) is simultaneously sent on Ls and Rs channels. A Pro Logic downmix en-
22/90
STA310
ables to configure which channels to output on PCM data. This is done through the register PL_DWN.
An auto-balance feature is available and activated through PL_AB register. The delay on surround channel is
configurable thanks to the LSDLY register (while resetting the RSDLY register).
The bass redirection is performed after the Pro Logic decode. The same bass redirection configuration than
those available in non-Pro Logic modes can be used except that the surround channels will not be added to the
bass redirection. In the case of AC-3 or MPEG the STA310 is therefore capable of first decoding the AC-3 or
MPEG stream then performing the Pro Logic decode.
4.7.2 Others
- Karaoke system
- Bass Management + Volume Control
- Deemphasis
- DC Remove
4.8 How to choose a decoder

To set up the device you have to select two registers.
The first one is DECODESEL for Audio data type,
The second one is STREAMSEL for Transport data type,
The STREAMSEL can be set-up as follows:
0= PES
1= PES DVD Video
2= Packet MPEG1
3= Elementary stream or IEC.60958
4= reserved
5= IEC.61937
6= PES DVD Audio
So the possible configurations on listed in the following table:
Table 6. Possible configurations:
STREAMSEL DECODESEL MODE

0 0 MPEG2 PES carrying Dolby Digital (ATSC)
0 1 MPEG2 PES carrying MPEG1 frames
0 2 MPEG2 PES carrying MPEG2 frames
1 0 MPEG2 PES carrying Dolby Digital frames for DVD Video
1 2 MPEG2 PES carrying MPEG2 frames for DVD Video
1 3 MPEG2 PES carrying Linear PCM frames for DVD Video
1 1 MPEG1 packet carrying MPEG1 frames
3 0 Dolby Digital frame elementary streams
3 1 MPEG1 frame elementary streams
3 2 MPEG2 frame elementary streams
3 3 Stereo PCM (16bits samples)
3 4 Pink Noise Generator
23/90
STA310
STREAMSEL DECODESEL MODE

3 5 CDDA frames
3 7 Beep Tone Generator
3 9 MP3 frame elementary streams
5 0 IEC61937 Input with Dolby Digital frames
5 1 IEC61937 Input with MPEG1 frames
5 2 IEC61937 Input with MPEG2 frames
6 3 MPEG2 PES carrying Linear PCM for DVD Audio
6 8 MPEG2 PES carrying MPL for DVD Audio
4.9 How to Program a Post Processing

4.9.1 2 registers for the mode:
PDEC (0x62) to define the type of PostProcessing
PDEC MODE
0x01 Prologic
0x02 MPEG 1/2 Dynamic Range
0x08 Double Stereo
0x10 DC Remove
0x20 Deemphasis filter
4.9.2 1 or 2 registers to control the “PostProcessing”

Prologic decoder (PDEC = 0x01):
PL ABL WS (0x64) Effect
1 AutoBalance
2 WideSurround
PL DOWNMIX (0x65) Prologic DownMix
0,1,2 Prologic not applied
3 3/0 (L, R, C)
4 2/1 (L, R, Ls) Phantom
5 3/1 (L, R, Ls)
6 2/2 (L, R, Ls, Rs) Phantom
7 3/2 (L, C, R, Ls, Rs)
Remark: When playing “Dolby Digital Prologic encoded”, if PL_DOWNMIX is correctly set, Prologic decoder’ is
automatically applied even if the register “PDEC” different to 1.
24/90
STA310
4.10 What Can Be Processed at the Same Time

Same Time 1
Decoder
MPEG1 Commands
MP3 S/PdifOutput
AC3 Mute
MPEG2 Post Bass
Volume Control PCM (Left
,Right
Skipframe Pcrocessing Management PCM (VCRs)
LPCM Video Encoded
PCM Pause Mute
MLP Off
LPCM Audio Pause block
PinkNoise
Beep Tone
Same Time 2
Post Post Post Post

Pcrocessing Pcrocessing Pcrocessing Pcrocessing
Prologic
Karaoke
Karaoke Channel Delay Karaoke
Channel Delay Channel Delay
5 PCM OUTPUT CONFIGURATIONS
5.1 Output configurations

The figure below shows the different configurations supported at PCM output stage. They are selected by the
OCFG register contents.
- In configuration 1, 3 and 4, the main channels are attenuated by 18.5dB, and the LFE by 8.5dB before
summing.
After digital/analog conversion, the subwoofer preamplifier has to compensate for the different gains
of the main channels and subwoofer.
- In configuration 2, the main channels are attenuated by 16dB and the LFE by 6dB before processing.
- In configuration 0, outputs are only scaled and rounded (see next section).
The same configurations will be used in case of a decoded Pro Logic program with the exception that the sur-
round channels will not be added to the bass redirection (the surround channels of a Pro Logic program are
band limited and bass is considered as leakage).
25/90
STA310
Figure 14. PCM Output Configurations

00 00
Not used with Prologic
L L L
00 00 L
00 0 0 0 0 0 00
C C
C C
00 0 0 0000 0000000000
R R
R R
LS LS
0000000
LS LS
RS RS
RS RS -18.5dB
LFE LFE LFE -8,5dB SUB
Configuration 0 Configuration 1
00 00
00 00
00 00 00 0 0000000 00 00
L -16dB L L L
C C
C -16dB C
00 0 0 0 0 0 0 000
00 0 0 0 00000 000 00
R R
00 0 00 00 00 00
R -16dB R
LS LS
LS -16dB LS RS RS
RS -16dB
0000000 RS
-18.5dB Not used in
configuration 4
LFE -6dB -4dB LFE -8,5dB SUB

Configuration 2 Configurations 3 and 4
5.2 PCM scaling

PCM scaling is needed for every decoding mode (AC3, Pro Logic, MPEG, PCM). It is applied at the end of the
filtering steps before PCM output, allowing maximum effective word width for most of the signal processing be-
fore.
Master volume (PCM_SCALE register) and balances (BAL_LR and BAL_SUR registers) are implemented for
PCM scaling.
5.3 Output quantization

For optimal results for 16/18/20-bit DACs, a quantization with rounding is applied together with the PCM scaling.
The sample value is multiplied by a rounding factor and rounded to 24 bits. The result is then left shifted (4/6/8)
for PCM output.
The output precision is selectable from the 16bits/word to 24 bits/word by configuring the field PREC in the reg-
ister PCMCONF.
5.4 Interface and output formats

The decoded audio data are output in serial PCM format.
The interface consists of the following signals
PCM_OUT0, 1, 2 PCM data, output,
SCLK Bit clock (or serial clock), output,
LRCLK Word clock (or Left/Right channel select clock), output,
PCMCLK PCM clock, input or output (see <CrossRef><BlueHT>Clocks <BlueHT>on page 11 for details).
26/90
STA310
5.4.1 Output precision and format selection

Output precision is selectable from 16 bits/word to 24 bits/word by setting the output precision select, in the PC-
MCONF (16-, 18-, 20- and 24-bit mode) register.
In 16-bit mode, data may be output either with the most significant bit first or least significant bit first. This is
configured by the contents of the field ORD in the PCMCONF register.
When PCMCONF.PREC is more than 16 bits, 32 bits are output for each channel. In this configuration, the field
FOR of register PCMCONF is used to select Sony or I²S- compatible format. The field DIF of PCMCONF is used
to position the 18, 20 or 24 bits either at the beginning or at the end of each 32-bit frame.
Figure 15. Output formats

16 SCLK cycles
LRCLK
16 SCLK cycles
M L M L
PCM_OUT[2:0] PCMCONF.ORD = 0, PCMCONF.PREC is 16 bits mode
S S S S
L M L M
PCM_OUT[2:0] S PCMCONF.ORD = 1, PCMCONF.PREC is 16 bits mode
S S S
32 SCLK cycles
LRCLK 32 SCLK cycles
M L M L PCMCONF.FOR = 1
PCM_OUT[2:0] S 18, 20 or 24 bits 0 18, 20 or 24 bits 0
S S S PCMCONF.DIF = 1
PCM_OUT[2:0] 0 18, 20 or 24 bits 0 18, 20 or 24 bits
S S S S PCMCONF.DIF = 0
PCM_OUT[2:0] 0 S 18, 20 or 24 bits
S
0 0 18, 20 or 24 bits 0
PCMCONF.DIF = 1
S S
M L MSB M L PCMCONF.FOR = 1
PCM_OUT[2:0] MSB 18, 20 or 24 bits 18, 20 or 24 bits
S S S S PCMCONF.DIF = 0
DATA IN SAMPLE DATA SENT ON THE

PCM PCM PCM
M CONF.PEC MEMORY DATA PCM SERIAL OUTPUT
CONF.ORD CONF.FOR CONF.DIF
[23:0] (LEFT BIT FIRST)
0:16-bit mode 1 NA NA {d23-d8}-{8*0} {d8-d23}: 16 bits
0:16-bit mode 0 NA NA {d23-d8}-{8*0} {d23-d8}: 16 bits
1:18-bit mode NA 0 0 {d23-d6}-{6*0} {13*0}{0}{d23-d6}: 32 bits
1:18-bit mode NA 0 1 {d23-d6}-{6*0} {0}{d23-d6}{13*0}: 32 bits
1:18-bit mode NA 1 0 {d23-d6}-{6*0} {14*d23}{d26*d6}: 32 bits
1:18-bit mode NA 1 1 {d23-d6}-{6*0} {d23-d6}{14*0}: 32 bits
2:20-bit mode NA 0 0 {d23-d4}-{4*0} {11*0}{0}{d23-d4}: 32 bits
2:20-bit mode NA 0 1 {d23-d4}-{4*0} {0}{d23-d4}{11*0}: 32 bits
2:20-bit mode NA 1 0 {d23-d4}-{4*0} {12*d23}{d23-d4}: 32 bits
2:20-bit mode NA 1 1 {d23-d4}-{4*0} {d23-d4}{12*0}: 32 bits
3:24-bit mode NA 0 0 {d23-d0} {6*0}{0}{d23-d0}: 32 bits
3:24-bit mode NA 0 1 {d23-d0} {0}{d23-d0}{7*0}: 32 bits
3:24-bit mode NA 1 0 {d23-d0} {8*d23}{d23-d0}: 32 bits
3:24-bit mode NA 1 1 {d23-d0} {d23-d0}{8*0}: 32 bits
27/90
STA310
How to read the above table:

The first 4 columns list the possible configurations for output formats on the PCM outputs. The 5th column gives
the description of the internal 24-bit decoded, scaled and rounded audio samples as they are stored in memory.
These 24 bits are referred to as d23, d22,..., d0, where MSB=d23, LSB=d0. The last column describes the se-
quence of bits that are output on PCM_OUT according to the selected format.
Example 1: in 16-bit mode, with PCMCONF.ORD=1: In memory, 24 bits are stored, where only the 16 MSB bits
(d23, d22,... to d8) are significant and the 8 remaining bits are 0. This is noted: {d23-d8} {8*0}. The data are sent
LSB first, i.e. d8 is sent first and d23 is sent last. This is noted {d8-d23}. 16 bits only are transmitted per channel.
Example 2: in 20-bit mode (PCMCONF.ORD field is meaningless in this mode), with PCMCONF.FOR=1 and
PCMCONF.DIF=0: In memory, 24 bits are stored, where only the 20 MSB (d23 to d4) are significant and the
remaining 4 LSB are 0.This is noted: {d23-d4} {4*0}. 32 bits are transmitted per channel on the PCM outputs:
the 12 first transmitted bits are d23, the last bits are d23 to d4, where d23 is transmitted first. This is noted:
{12*d23} {d23-d4}.
5.4.2 Clocks polarity selection

The polarity of the PCM serial output clock, SCLK and the polarity of the PCM word clock LRCLK are selected
by the field SCL and INV respectively, in the PCMCONF register.
5.4.3 I2S format compatible outputs

To output I²S compatible data, the PCMCONF register must be configured as follows
PCMCONF.DIF =1 not right padded,
PCMCONF.FOR =0 I²S format,
PCMCONF.INV =0 do not invert LRCLK,
PCMCONF.SCL =0 do not invert SCLK.
5.4.4 Sony format compatible outputs

PCMCONF.FOR =1 Sony format,
PCMCONF.INV =1 Invert LRCLK.
Figure 16. SCLK Polarity

00 00
00 00
SCLK SCLK
0 0
LRCLK LRCLK
PCM_OUT0, 1, 2 PCM_OUT0, 1, 2
SCL = 0 SCL = 1
0 0 0 0 0
Figure 17. LRCLK Polarity
00 00 00 00 00 00
00 00 00 00 00 00
Left Right
LRCLK
0
Right Left
INV = 1 INV = 0
28/90
STA310
6 S/PDIF OUTPUT
The S/PDIF output pad is a TTL output pad with slew rate control. The output DC capability is 4 mA. The voltage
drop is 3V. This output must be connected to a TTL driver before the transformer.
The S/PDIF output supports SPDIF and IEC-61937 standards. Several registers must be initialized to configure
the SPDIF output:
- The category code must be entered in the IEC958_CAT register. It is related to the type of application.
The category code is specified in the Digital Output Interface standard.
- The status bits that will be transmitted on the SPDIF output, must be programmed in the
IEC958_STATUS register.
- IEC clock setting must be specified in the IEC958_CONF register.
- The data type dependent information can be specified in the IEC958_DTDI register.
- The S/PDIF type is selected through the IEC958_CMD register: the IEC unit can output decoded data
(PCM mode), encoded data, or null data.
Note: 1. The SPDIF output handles only 48kHz or 44.1kHz sample rates.
6.1 SPDIF output

When configured in SPDIF mode, the S/PDIF output is used to transmit either the L/R channels (PCMOUT1) or
VCR_L/VCR_R (PCMOUT0).
The selection is done by choosing the PCM mode and AUX = 1 in the register SPDIF_CMD and resetting the
COM status of SPDIF_STATUS register.
6.2 IEC-61937 output

When configured in IEC-61937 mode, the S/PDIF output is used to transmit encoded data taken directly from
the frame buffer.
The selection is done by choosing the encoded mode (ENC mode) in the register IEC958_CMD and setting the
bit COM in IEC958_STATUS register.
The decompressed data are output simultaneously on the PCM_OUT outputs.
Latency in software versions 6 and later
For software versions 6 and later, when choosing to output encoded S/PDIF data, a latency is automatically in-
serted between S/PDIF output and PCM outputs. The PCM outputs are delayed compared to the SPDIF output.
The latency value is defined by standards and applied when the auto-latency mode is selected.
AC3 decoding
Latency = 1/Fs * (1/3 * Framesize + 256)
= 1/Fs * (32 * Datarate/Fs + 256)
MPEG decoding
Latency = 1/Fs * (36 * Datarate/Fs + 96)
where Fs is the sampling frequency in kHz, Framesize is expressed in 16-bit words, Datarate is the bit rate in
kbits per second.
The latency insertion can not be disabled however it can be programmed to values different from those required
in the standard by selecting the user-programmable-latency mode (by setting the bit 7 of IEC858_CONF regis-
29/90
STA310
ter). In this case, the latency is specified in the IEC958_LATENCY register.
Note that there are minimum and maximum values to respect
Table 7.
AC-3 MPEG
Min. Latency Max. Latency Min. Latency Max. Latency
256 1536 96 1152
samples / Fs samples / Fs samples / Fs samples / Fs
If those limits are not respected, an error interrupt occurs corresponding to error type: LATENCY_TOO_BIG,
which automatically makes the chip switch to auto_latency mode.
For software versions prior to 6, the latency is not implemented.
6.3 PCM null data

When configured in muted mode (in the IEC958_CMD register), the outputs are PCM null data. This can be
used to synchronize the external IEC receiver.
7 INTERRUPTS
7.1 Interrupt register

The decoder can signal to the external controller that an interrupt has occurred during the execution.
The register INTE enables to select which interrupts will be generated and output on the IRQ output pin.
When an interrupt occurs, the signal IRQ is activated low and the controller can check which interrupt was de-
tected by reading the register INT.
According to the type of interrupt detected, other information can be obtained by reading associated registers
(such as stream header, type of error detected, PTS value).
7.2 IRQ Signal

This signal, IRQ, is a three-state line. This signal indicates (by going low) when an interrupt occurs. It returns to
high level once the corresponding bit in the interrupt register has been cleared.
7.3 Error concealment

Errors are signaled as interrupts by the audio core. The error list is provided in. Most of the errors are automat-
ically handled by the core, some require that software be changed.
AC-3 decoding errors:

Those errors are signaled in the ERROR register but handled directly by the core. Nothing can be done by the
software. They signal that something wrong happened during the decoding. The core soft mutes the frame and
continues to decode.
MPEG decoding errors:

Those errors are also signaled in the ERROR register but handled directly by the core. Nothing can be done by
the software. They signal that something wrong happened during the decoding. The core soft mutes the frame
and continues to decode. Only one error in this category indicates a programming error: if triggering the
30/90
STA310
MPEG_EXT_CRC_ERROR, the bit MC_OFF must be set. This indicates that the decoder tries to decode more
than 2 channels whereas the incoming stream contains only 2 channels.
Packet and audio synchronization errors:

Those errors are handled internally, and usually indicate that the incoming bitstream is incorrect or incorrectly
input to the chip. In those cases, the decoder resets the corresponding parsing stage (packet or audio parser)
then searches for the next correct frame.
Miscellaneous errors:
- LATENCY_TOO_BIG error indicates a problem of latency programming which is superior to the max-
imum authorized value.
Change the latency value or switch to auto-latency mode to solve the problem. Other miscellaneous
errors are internally handled.
31/90
STA310
8 AUDIO/VIDEO SYNCHRONIZATION
8.1 Presentation time stamp detection

8.1.0.1 PTS Signal
This signal, PTS, is used to signal the detection of a Presentation Time Stamp in a stream, for audio/video syn-
chronization. When a PTS is detected, the signal PTS goes low during one LRCLK period. It is generated while
the PCM are output, so to enable the use of an external counter to synchronize the STA310 with a video decod-
er.
The signal is activated, even if PTS interrupt is not enabled.
8.1.1 PTS interrupt
When enabled through the INTE register, the interrupt PTS is generated when a PTS is detected. The interrupt
is signalled on the IRQ output, which goes low. The IRQ signal is de-activated once the PTS bit has been
cleared in INT register by reading the PTS Most Significant Bit.
8.1.2 PTS interrupt and signal relative timings
The IRQ configured as PTS interrupt is output before the PTS signal. The PTS signal is activated at last one
period of LRCLK after the IRQ signal.
8.2 Frames skip capability

When the audio decoder is late compared to the video decoder, the decoder is able to skip frames. Writing 1 in
the SKIP_FRAME register makes the decoder ignore the next incoming frame. Once skipping the frame, it con-
tinues to decode the stream, and the SKIP_FRAME register is automatically reset.
8.3 Frames repeat capability

When the audio decoder is ahead of the video decoder, the decoder can repeat frames. Writing 1 in the
REPEAT_FRAME register makes the decoder repeat the current frame. Once repeating the frame, the chip
plays the next incoming frame, and the REPEAT_FRAME register is reset.
9 REGISTER MANUAL
9.1 Introduction
The STA310 device contains 256 registers.
Two types of registers exist:
- From address 0x00 to 0x3F, the registers are real registers that can be initialized after reset.
- From address 0x40 to 0x100, they are memory locations. This means that the registers located at the
address 0x40 to 0x100 can have different meanings and usage according to the mode in which the
device operates.
Be careful that they can not be hardware reset: they contain undefined values at reset and require to
be initialized after each hardware reset.
In this document, only the user registers are described.
The undocumented registers are reserved. These registers must never be accessed (neither in Read nor in
Write mode).
The Read only registers must never be written
32/90
STA310
9.2 Register map by function

Code Description
(a) Register modification is ALWAYS taken into account by the audio decoder.
Any change to these registers is taken into account immediately.
(b) Register modification is taken into account AFTER EVERY DECODED DATA BLOCK or JUST AFTER
RESET (soft or hard).
The decoded block is related to the granularity of the computation in the audio decoder software.
A block is 256 samples in Dolby Digital, 96 samples in MPEG, 80 samples in LPCM/PCM.
(f) Register modification is taken into account AFTER EVERY DATA FRAME.
A frame is: 1152 samples in MPEG I/II, 1536 samples in Dolby Digital, 384 samples in MPEG-1 layer 1, 80
samples in LPCM/PCM.
(r) Register modification is taken into account ONLY WHEN THE DSP IS RUN AFTER RESET (soft or hard).
(1) The delay registers are updated when bit 0 of the UPDATE register is set to 1.
(2) The volume is updated when CHAN_IDX is set to the appropriate value.
(3) The Karaoke mode is updated when KAR_UPDATE is set to ’1’.
The following tables list the register map by address and function, then each audio decoder register is described
individually
Table 8.
Register function HEX DEC Name
VERSION 0x00 0 VERSION
0x01 1 IDENT
0x71 113 SOFT‘VER
SETUP + INPUTS 0x0C 12 SIN_SETUP (a)
0x0D 13 CAN_SETUP (a)
PCM CONFIGURATION 0x54 84 PCMDIVIDER (b)
0x55 85 PCMCONF (b)
0x56 86 PCMCROSS (b)
DAC AND PLL CONFIGURATION 0x05 5 SFREQ (f)
0x12 18 PLLCTRL (f)
0x18 24 PLLMASK (a)
0x0E 14 DATA IN
0x12 18 PLLCTRL (f)
0x11 17 PLL_DATA (a)
0x1D 29 PLL_CMD(f)
0x12 18 PLL_ADD (f)
0xB5 181 ENA_ALL FRACPLL
0xB6 182 AU_PLL_FRACL_192
0xB7 183 AU_PLL_FRACH_192
0xB8 184 AU_PLL_XDIV_192
0xB9 185 AU_PLL_MDIV_192
0xBA 186 AU_PLL_NDIV_192
0xBB 187 AU_PLL_FRACL_176
0xBC 188 AU_PLL_FRACH_176
0xBD 189 AU_PLL_XDIV_176
0xBE 190 AU_PLL_MDIV_176
33/90
STA310

CHANNEL DELAY SETUP 0x57 87 LDLY (1)
0x58 88 RDLY (1)
0x59 89 CDLY (1)
0x5A 90 SUBDLY (1)
0x5B 91 LSDLY (1)
0x5C 92 RSDLY (1)
0x5D 93 UPDATE (f)
0xAF 91 LVDLY (1)
0xB0 92 RVDLY (1)
SPDIF OUTPUT SETUP 0x5E 94 SPDIF_CMD (r)
0x5F 95 SPDIF_CAT (f)
0x60 96 SPDIF_CONF (b)
0x61 97 SPDIF_STATUS (b)
0x75 117 SPDIF_REP_TIME (b)
0x7E 126 SPDIF_LATENCY (f)
0x7F 127 SPDIF_DTDI (f)
COMMAND 0x10 16 SOFTRESET (a)
0x13 19 PLAY (a)
0x14 20 MUTE (a)
0x72 114 RUN (a)
0x73 115 SKIP_MUTE_CMD (f)
0x74 116 SKIP_MUTE_VALUE(f)
INTERRUPT 0x07, 08 7, 8 INTE (a)
0x09, 0A 9, 10 INT (a)
INTERRUPT STATUS 0x40 64 SYNC_STATUS
0x41 65 ANCCOUNT
0x42 66 HEAD4 (f)
0x43 67 HEAD3 (f)
0x44, 45 68, 69 HEADLEN (f)
0x46 - 4A 70 PTS (f)
0x0F 15 ERROR
DECODING ALGORITHM 0x4C 76 STREAMSEL (r)
0X4D 77 DECODSEL (r)
SYSTEM SYNCHRONIZATION 0x4F 79 PACKET_LOCK (r)
0x50 80 ID_EN (a)
0x51 81 ID (a)
0x52 82 ID_EXT (a)
0x53 83 SYNC_LOCK (r)
POST DECODING AND PRO LOGIC 0x62 98 PDEC1 (b)
0xB1 177 PDEC2
0x64 100 PL_AB (b)
0x65 101 PL_DWNX (b)
0x66 102 OCFG
0x70 112 DWSMODE (b)
34/90
STA310

BASS REDIRECTION 0x4E 78 VOLUME0 (2)
0x63 100 VOLUME1 (2)
0x66 102 OCFG (b)
0x67 103 CHAN_IDX (b)
DOLBY DIGITAL CONFIGURATION 0x68 104 AC3_DECODE_LFE (b)
0X69 105 AC3_COMP_MOD (b)
0x6A 106 AC3_HDR (b)
0x6B 107 AC3_LDR (b)
0x6C 108 AC3_RPC (b)
0x6D 109 AC3_KARAMODE (b)
0x6E 110 AC3_DUALMODE (b)
0x6F 111 AC3_DOWNMIX (b)
0x76 118 AC3_STATUS0 (f)
0x7A 122 AC3_STATUS4 (f)
0x7B 123 AC3_STATUS5 (f)
0x7C 124 AC3_STATUS6 (f)
0x7D 125 AC3_STATUS7 (f)
MPEG CONFIGURATION 0x68 104 MP_SKIP_LFE (b)
0x69 105 MP_PROG_NUMBER (b)
0x6E 106 MP_DUALMODE (b)
0x6A 110 MP_DRC (b)
0x6C 108 MP_CRC_OFF (b)
0x6D 109 MP_MC_OFF (b)
0x6F 111 MP_DOWNMIX (b)
0x76 118 MP_STATUS0 (f)
0x7A 122 MP_STATUS4 (f)
0x7B 123 MP_STATUS5 (f)
PINK NOISE GENERATION 0x6F 111 PN_DOWNMIX
REGISTERS
PCM BEEP-TONE CONFIGURATION 0x68 104 PCM_BTONE (b)
35/90
STA310

KARAOKE 0x81 129 KAR_MCh0VOL (3)
0x82 130 KAR_MCh1VOL (3)
0x83 131 KAR_KEYCONT (3)
0x84 132 KAR_KEYVALUE (3)
0x85 133 KAR_VCANCEL (3)
0x86 134 KAR_VVALUE (3)
0x87 135 KAR_MMUTE (3)
0x88 136 KAR_VCh0VOL (3)
0x89 137 KAR_VCh1VOL (3)
0x8A 138 KAR_DUET (3)
0x8B 139 KAR_DUETTHRESH (3)
0x8C 140 KAR_VOICE (3)
0x8D 141 KAR_VDELAY (3)
0x8E 142 KAR_VBAL (3)
0x8F 143 KAR_VMUTE (3)
0x90 144 KAR_PLAY (3)
0x91 145 KAR_MODE (3)
0x92 146 KAR_DIN_CTL (3)
0x93 147 KAR_UPDATE
SECOND SERIAL INPUT 0x94 148 SFREQ2
0x95 149 CANINPUT_MODE
LINEAR PCM (DVD AUDIO) 0x6F 111 LPCMA_DOWNMIX
REGISTERS 0x70 112 LPCMA_FORCE_DWS
0x76 118 LPCMA_STATUS0
0x7A 122 LPCMA_STATUS4
0x7B 123 LPCMA_STATUS5
0x97 151 LPCMA_DM_COEFT_0
0x9A 154 LPCMA_DM_COEFT_3
0x9B 155 LPCMA_DM_COEFT_4
0x9C 156 LPCMA_DM_COEFT_5
0x9D 157 LPCMA_DM_COEFT_6
0x9E 158 LPCMA_DM_COEFT_7
0x9F 159 LPCMA_DM_COEFT_8
0xA0 160 LPCMA_DM_COEFT_9
LINEAR PCM (DVD VID & PCM) 0x6F 111 LPCMV_DOWNMIX
REGISTERS
36/90
STA310

0x70 112 LPCMV_FORCE_DWS
0x76 118 LPCMV_STATUS0
0x97 151 LPCMV_DM_COEFT_0
0x9A 154 LPCMV_DM_COEFT_3
0x9B 155 LPCMV_DM_COEFT_4
0x9C 156 LPCMV_DM_COEFT_5
0x9D 157 LPCMV_DM_COEFT_6
0x9E 158 LPCMV_DM_COEFT_7
0x9F 159 LPCMV_DM_COEFT_8
0xA0 160 LPCMV_DM_COEFT_9
0xA8 168 LPCMV_CH_ASSIGN
0xA9 169 LPCMV_MULTI_CHS
DE-EMPHASIS REGISTER 0xB5 181 DEEMPH
AUXILLIARY OUTPUTS REGISTERS 0xAE 174 VCR_MIX
0xAF 175 VCR_LDLY
0xB0 176 VCR_RDLY
MISCELLANEOUS 0x2B 43 BREAKPOINT
0x3A 58 CLOCKCMD
0xFF 255 INIT_RAM
SPDIF AUTODETECTION 0xE0 224 AUTODETECT_ENA
0xE1 225 AUTODETECT_SENS
0xE2 226 AUTODETECT_ALIGN
37/90
STA310
9.3 VERSION REGISTERS

IDENT
Identify
7 6 5 4 3 2 1 0
1 0 1 0 1 1 0 0
Address: 0x01
Type: RO
Software Reset: 0x31
Hardware Reset: 0x31
Description:
IDENT is a read-only register and is used to identify the IC on an application board. IDENT always has the value
“0x31”.
SOFTVER
Software version
7 6 5 4 3 2 1 0
Address: 0x71
Type: R/W
Software Reset: NC
Hardware Reset: UND
Description:
This SOFTVER register is the version of the code which is running on the device. This regiter is updated by the
embedded software just after a soft reset of the device:
- For STA310 cut 1.0 the register contain the value 0x0A
- For STA310 cut 2.0 the register contain the value 0x14
Loading a patch into the STA310 will automatically change the register content.
Please contact ST to have the correct value according to the patch being used.
This register must be readonly after the STA310 has finished booting, in order to get a correct value (when
INIT_RAM register hold the value 1)
VERSION
Version
7 6 5 4 3 2 1 0
0 0 0 1 0 0 0 0
Address: 0x00
Type: RO
38/90
STA310
Software Reset: NA shown below:

Hardware Reset: NA IMODE[1:0] Mode
0 Parallel input (DSTR+Data [7:0] + REQ)

Description:
1 Serial input (DSTR + SIN + REQ)
This version register is read only and is used to iden-
tify the audio hardware version. The version register 2 Reserved - Not used
holds a number which refers to the cut number. The 3 A/D input (DSTR+LRCLKIN+REQ+SIN;
version numbers are defined as below: SPDIF input; PCM input
- STA310 cut 1.0, version number is : 0x10
When the IC is configured in mode 1 or 3, the
- STA310 cut 2.0, version number is : 0x10 CAN_SETUP register is used to configure the IC with
repect to the data format.
9.4 SETUP & INPUT REGISTERS

CAN_SETUP
The STA310 can get receive an input bitstream either
from the I2s input or ffrom the Spdif input, the selec- A/D converter setup
tion and the configuration is done through 2 registers
7 6 5 4 3 2 1 0
SIN_SETUP @ 12 and CAN_SETUP @ 13.
S16 SAM FIR PAD
SIN SETUP Address : 0x0D
Input data setup Type: R/W
7 6 5 4 3 2 1 0
Software Reset: NC
Hardware Reset: 0
SPDIF POL IMODE [1...0]
Address: 0x0C Description:

Type: R/W CAN_SETUP is used to configure the data serial in-
Software Reset: NC terface. The register is only taken into account when
the register sin_setup [1...0] = 3.
Hardware Reset: 0
Also see SIN_SETUP register./
- S16 When set, the slot count is 16. When re-
Description:
set, the slot count is 32 but only the first 16
This register is used to configure the input data inter- are extracted.
face. The register must be setup before sending data
to the IC. The mapping of the register isescribed be- - SAM When set data is sampled on the falling
low. Remember that the data must be sent to the de- edge of the DSTR. When reset, the data is
vice MSB first. sampled on the rising edge of DSTR
- SPDIF data frpm SPDIF when set to 1, data - FIR When set the first channel (Left) is input
from main I2S input. when Lrclkin=1. When reset, the first channel
- POL Polarity of the REQ signal. When set, is input when Lrclkin=0.
the REQ pin is active low: data must be input - PAD When set, data Lrclkin is delayed by one
when REQ is low. When reset, the REQ pin is cycle (padding mode).
active high and data must be input when REQ When the IC is configured with the S/PDIF input, reg-
is high. ister CAN_SETUP must be set to 2
- IMODE[1...0] Input mode. Indicates which DATAIN
data input interface is used. Data input
The configuration of the 3 possible interfaces is
7 6 5 4 3 2 1 0
39/90
STA310
Address : 0x0E PCM divider value Mode description

Type: WO 2 DAC_PCMLK = 384 Fs,
DAC is 32-bit mode
Software Reset: NA
1 DAC_PCMLK = 256 Fs,
Hardware Reset: NA DAC is 32-bit mode
Description:
PCMCONF
Data can be fed into the STa310 by using this register
PCM configuration
instead of the dedicated interface. there is no need to
byte align the bitstream when using this register. 7 6 5 4 3 2 1 0
ODR DIF INV FOR SCL PREC[1:0]
Address: 0x55
Type: R/W
9.5 PCM CONFIGURATION RESISTERS Software Reset: NC
PCMDIVIDER
Hardware Reset: UND
Divider for PCM clock Description:
7 6 5 4 3 2 1 0 Bitfield Description
ORD PCM Order: This bit is significant only
when in 16-bit mode. When set, LSB
Address : 0x54 is sent first. When reset, MSB is sent
first.
Type: R/W
DIF PCM_DIFF: This bit is not significant
Software Reset: UND in 16-bit. When set, indicates that the
bits are not right-padded in the slot.
Hardware Reset: UND
When reset, Ii is right padded.
INV INV_LRCLK: When set the polarty of
Description: LRCLK is inverted: Left channel is
output when LRCLK is high.
The PCM divider must be set according to the formu- When reset, the polarity of LRCLK is
la below, where DAC_SCLK is the bit clock for the such that the left channel is outout
DAC. When Div is set to 0, DAC_SCLK is equal to when LRCLK is low.
DAC_PCMCLK:
FOR FORMAT: This bit selects the data
output format: When set, the Sony
Div = (DAC_PCMCLK/ (2 x DAC_SCLK)) -1 format is chosen. When reset 0 the
format is IS format.
SCL INV_SCLK: When set, the polarity of
When the internal PLL is used, DAC_PCMCLK=384 SCLK is inverted, the PCM outputs
x fs or 256 x fs. If DAC_PCMCLK = 384 x fs, the for- and LRCLK will be stable for the
mula becomes: DACs on the falling edge of SCLK.
When reset, PCM outputs and LRCLK
are stable on the rising edge of SCLK.
Div = (192 x Fs/DAC_SCLK) -1
PREC[1:0] PCM Precision
0: 16 bit mode (16 slots)
If DAC_SCLK is 32 x Fs (common case with the 16 1: 18 bit mode (32 slots)
bit DAC), Div must be set to 5.
PCM divider value Mode description
5 DAC_PCMCLK = 384Fs,
PCMCROSS
DAC is 16-bit mode
7 6 5 4 3 2 1 0
3 DAC_PCMLK = 256 Fs,
DAC is 16-bit mode VCR CLR[1:0] CSW[1:0] LRS[1:0]
40/90
STA310
Address: 0x56
Type: R/W Fs
46 44.1 32 - 96 88.2 64 - 24 22.05
(KHz)
Software Reset: NC
Value 0 1 2 3 4 5 6 7 8 9
Hardware Reset: UND
Description: Fs
16 - 12 11.025 8 - 192 176.4 128 -
(KHz)
The PCMCROSS register only acts if bit PFC of reg-
ister SPDIF_DTDI is set. Value 10 11 12 13 14 15 16 17 18 19
Bitfield Description
LRS[1:0] Cross left and right surround. PLLCTRL
CSW[1:0] Cross centre and subwoofer. PLL Control
CLR[1:0] Cross left and right channels.
00: Left channel is mapped on the left 7 6 5 4 3 2 1 0
output, Right channel is mapped on the SYSCLSCK[1..0] OCLK[2..0]
Right output.
01: Left channel is duplicated on both Address: 0x12
outputs.
Type: R/W
10: Right channel is duplicated on both
outputs. Software Reset: NA
11: Right channel and Left channel are
Hardware Reset: 0x19
toggled.
VCR[1:0] These 2 bits manage the VCR outputs.
Description:
Bitfield Value Description
OCLK Configure PCMCLK PCMCLK pad
[2:0] source and direction direction
9.6 PDAC and PLL configuration registers
-01 Audio master Clock Input
SFREQ from PCMLCK pad.
Sampling frequency 011 Audio master Clock Input
from internal audio
7 6 5 4 3 2 1 0 PLL
111 Audio master Clock Input
from internal S/PDIF
Address: 0x05 receiver
Type: R/WS -00 Forbidden
Software Reset: NC 010 Audio master Clock Output
from internal audio
Hardware Reset: 0 PLL
110 Audio master Clock Output
Description: from internal S/PDIF
receiver
This status register holds the code of the current SYSCLC 0 System Clock from CLK pad
sampling frequency. If the audio stream is encoded K[1:0] Output
(Dolby Digital, MPEG) or packetized (DVD_LPCM), 1 System Clock from CLK pad divided
the sampling frequency is automatically read in the by 2
audio stream and written into this register by the au- 2 System Clock from internal system
dio DSP. The register is automatically updated by the PLL
DSP when it performs a down-sampling (for exam- 3 System Clock from internal system
ple, 96kHz to 48kHz). PLL divided by 2
The DSP resets SFREQ to 0.
PLL_DATA
For PCM stream or CDDA, this register is written to
by the application. The value in SFREQ corresponds
to the following frequencies:.
41/90
STA310
PLLData PLL Address
7 6 5 4 3 2 1 0 7 6 5 4 3 2 1 0
ADDRESS
Address: 0x12
Address: 0x11
Type: R/W Type: R/W
Software Reset: NA
Software Reset: NA
Hardware Reset: 0
Hardware Reset: 0
Description:
Description:
Data that must be written (has been read) at (from)
the address specified by PLL_ADD. Value Address of PLLs configuration registers
Address 2: Disable System PLL
3: System PLL frac Low
PLL_CMD
4: System PLL frac High
PLL Command 6: System PLL N divider
7: System PLL X divider
7 6 5 4 3 2 1 0 8: System PLL M divider
9: System PLL update
AUPLLCTL SYSPLLCTL RWCTL[1:0] 10: Disable Audio PLL
11: Audio PLL Frac Low
Address: 0x1D 12: Audio PLL Frac High
14: Audio PLL N divider
Type: R/W
15: Audio PLL X divider
Software Reset: NA 16: Audio PLL M divider
17: udio PLL update
Hardware Reset: 0
Description: ENA_AU_FRACPLL
Bitfield Description Audio PLL Enable
RWCTL [1:0] Configure PCMCLK source and 7 6 5 4 3 2 1 0
direction.
00: No action is performed on the ENA_PLL
configuration registers of the level 1
01: Read action of the configuration Address: 0xB5
registers. During this phase, the
content of a selectable (by PLL_ADD)
Type: R/W
register of the level 1 is copied into the Software Reset: 1
PLL_DATA register.
10: Write action of the configuration
Hardware Reset: 0
registers. During .this phase, the
content of a selectable (by PLL_ADD)
register of the level 1 is copied into the
Description:
PLL_DATA register. This register is used to enable the audio PLL of the
11: Forbidden STA310. This register must be always set to “1” after
SYSPLLCTL System PLL coefficients transfert either a soft or hardware reset.
0: No Transfer
1: Transfer the data between the level 1
and the level 2 for the system PLL AU_PLL_FRACL_192
Frac Low Coefficient
AUPLLCTL Audio PLL coefficient transfert
0: No Transfer
7 6 5 4 3 2 1 0
1: Transfer the data between the level 1
and the level 2 for the audio PLL FRACL
PLL_ADD Address: 0xB6
42/90
STA310
Type: R/W This register must contain a XDIV value that enables
Software Reset: 0x34 the audio PLL to generate a frequency of
ofact*192KHz for the PCMCK.
Hardware Reset: UND
Default value at soft reset assume:
– Oversampling factor (ofact) = 384. PCMLCK =
Description: 384 x SF (where SF is the sampling frequency)
This register must contain a FRACL value that en- – External crystal provide a clock running at
ables the audio PLL to generate a frequency of
27MHz
– Oversampling factor (ofact) = 384. PCMLCK = AU_PLL_MDIV_192
384 x SF (where SF is the sampling frequency) M Divider Coefficient
– External crystal provide a clock running at 7 6 5 4 3 2 1 0
27MHz
MDIV
Address: 0xB9
AU_PLL_FRACH_192
Type: R/W
Frac High Coefficient
7 6 5 4 3 2 1 0
Hardware Reset: UND
FRACH
Address: 0xB7 Description:

Type: R/W This register must contain a MDIV value that enables
Software Reset: 0xEC the audio PLL to generate a frequency of
Hardware Reset: UND
Description: 384 x SF (where SF is the sampling frequency)
This register must contain a FRACH value that en- – External crystal provide a clock running at
27MHz
– Oversampling factor (ofact) = 384. PCMLCK = AU_PLL_NDIV_192
384 x SF (where SF is the sampling frequency) N Divider Coefficient
27MHz
NDIV
Address: 0xBA
AU_PLL_XDIV_192
Type: R/W
X Divider Coefficient
7 6 5 4 3 2 1 0 Hardware Reset: UND
XDIV
Description:
Address: 0xB8
This register must contain a NDIV value that enables
Type: R/W the audio PLL to generate a frequency of
Software Reset: 0x01 ofact*192KHz for the PCMCK.
Hardware Reset: UND Default value at soft reset assume:
Description: – Oversampling factor (ofact) = 384. PCMLCK =
43/90
STA310
384 x SF (where SF is the sampling frequency) X Divider Coefficient

– External crystal provide a clock running at
7 6 5 4 3 2 1 0
27MHz
XDIV
AU_PLL_FRACL_176 Address: 0xBD

Frac Low Coefficient Type: R/W
7 6 5 4 3 2 1 0
Hardware Reset: UND
FRACL
Address: 0xBB Description:

Type: R/W This register must contain a XDIV value that enables
Hardware Reset: UND
Description:
384 x SF (where SF is the sampling frequency)
This register must contain a FRACL value that en-
ofact*176KHz for the PCMCK. 27MHz

– Oversampling factor (ofact) = 384. PCMLCK = AU_PLL_MDIV_176
384 x SF (where SF is the sampling frequency) M Divider Coefficient
7 6 5 4 3 2 1 0
27MHz
MDIV
AU_PLL_FRACH_176 Address: 0xBE

Frac High Coefficient Type: R/W
7 6 5 4 3 2 1 0
Hardware Reset: UND
FRACH
Address: 0xBC Description:

Type: R/W This register must contain a MDIV value that enables
Hardware Reset: UND
Description:
This register must contain a FRACH value that en-
ofact*176KHz for the PCMCK. 27MHz

– Oversampling factor (ofact) = 384. PCMLCK = AU_PLL_NDIV_176
384 x SF (where SF is the sampling frequency) N Divider Coefficient
27MHz
NDIV
AU_PLL_XDIV_176
44/90
STA310
Address: 0xBF Hardware Reset: 0

Type: R/W Description
Bitfield
Hardware Reset: UND HALF_FS If the incoming bitstream is encoded with
half sampling frequency, the device
generates a PCM clock (for audio DAC)
Description:
1: At 256 x half_fs or 384 x half_fs (half_fs
This register must contain a NDIV value that enables is equal to 24KHz, 22.05KHz, 16KHz).
the audio PLL to generate a frequency of
ofact*176KHz for the PCMCK. 0: At 256 x fs or 384 x fs (fs is equal to
Default value at soft reset assume: 48KHz, 44.1KHz, 32KHz).
This function is mainly use for DAC
– Oversampling factor (ofact) = 384. PCMLCK = frequency adaptation.
27MHz
9.7 Channel delay set-up registers
The six delay setup registers are used to set the rel-
INIT_RAM ative delays to the (up to) six loud speaker channels
in order to give the sound effects of, for example, a
STa310 boot Done large room or to compensate for the listener not being
in the centre of the loud speaker system. The sum of
7 6 5 4 3 2 1 0
the delays on the channels must be less than or
RAM_INIT equal to 35ms.
The unit for the register delay contents is a group of
Address: 0xFF 16 samples.
Type: RO Each register value is chosen using the expression:
Software Reset: 1 desired channel delay’*’sampling frequency’/16 sam-
Hardware Reset: 0 ples and taking care to ensure that the sum of the ’de-
sired channel delays’ is not more than 35ms.
Description: For example, when the sampling frequency is 48kHz,
the sum of the values programmed in the six delay
This register is used to signal when the STA310 has registers must be less than or equal to:
finished to boot. After a soft reset or a hardware re-
set, the host processor must wait until INIT_RAM
hold the value “1”. 35 ms * 48 KHz /16 samples = 105.
The host can then start to configure the STA310 ac- When only one surround channel is present (in Pro
cording to its application. Logic or other mode), the right surround delay must
be cleared, and the left delay channel is used for both
surround channels.
PLLMASK LDLY
PCMCLK mask for half sampling frequency Left channel delay
7 6 5 4 3 2 1 0 7 6 5 4 3 2 1 0
HALF_FS
Address: 0x18 Address: 0x57

Type: W Type: R/W
Software Reset: NC Software Reset: NC
Hardware Reset: UND
45/90
STA310
Delay on left channel, expressed in number of group LSDLY

of 16 samples. LDLY = delay (ms) * Fs (kHz) / 16 Left surround channel delay
7 6 5 4 3 2 1 0
RDLY
Right channel delay
Address: 0x5B
7 6 5 4 3 2 1 0
Type: R/W
Software Reset: NC
Address: 0x58 Hardware Reset: UND
Type: R/W Delay on left surround channel, expressed in number
of group of 16 samples. LSDLY = delay (ms) * Fs
Software Reset: NC
(kHz) / 16
Hardware Reset:UND
Delay on right channel, expressed in number of
RSDLY
group of 16 samples. RDLY = delay (ms)*Fs (kHz)/16
Right surround channel delay
CDLY 7 6 5 4 3 2 1 0
Centre channel delay
7 6 5 4 3 2 1 0 Address: 0x5C
Type: R/W
Software Reset: NC
Address: 0x59
Hardware Reset: UND
Type: R/W
Software Reset: NC
Description:
Hardware Reset: UND
Delay on right surround channel, expressed in num-
ber of group of 16 samples. RSDLY = delay (ms) * Fs
Description: (kHz) / 16.
Delay on center channel, expressed in number of When only one surround channel is used, this regis-
group of 16 samples. CDLY = delay (ms) * Fs (kHz) / ter must be reset at initialization.
16
LVDLY
SUBDLY Left VCR channel delay
Subwoofer channel delay
7 6 5 4 3 2 1 0
7 6 5 4 3 2 1 0
Address: 0xAF
Address: 0x5A
Type: R/W
Type: R/W
Software Reset: NC
Software Reset: NC
Hardware Reset: UND
Hardware Reset: UND
Description:
Description:
Delay on left VCR channel, expressed in number of
Delay on subwoofer channel, expressed in number of group of 16 samples.
group of 16 samples. SUBDLY = delay (ms) * Fs LSDLY = delay (ms) * Fs (kHz) / 16
(kHz) / 16
46/90
STA310
RVDLY Software Reset: 0

Right VCR channel delay Hardware Reset 0
7 6 5 4 3 2 1 0
Description:
Address: 0xB0
DLY This bit must be set to 1 to force the DSP to
Type: R/W
update its delays values (read from the audio
Software Reset: NC delay registers).
Hardware Reset: UND
0: Delay values held in the audio delay
registers are NOT updated in the DSP
(i.e. the DSP will keep the delay values set
Description:
previously)
Delay on right VCR channel, expressed in number of
group of 16 samples. RSDLY = delay (ms) * Fs (kHz) 1: The delay values held in the audio delay
/ 16. registers are updated in the DSP (i.e. the
DSP will use the new values). This bit is
When only one VCR channel is used, this register automatically reset to zero after it the update
must be reset at initialization. has been carried out.
TM Set to “0”
UPDATE
PCM delay update
7 6 5 4 3 2 1 0
9.8 SPDIF output set-up
TM DLY
SPDIF_CMD
Address: 0x5D SPDIF control
Type: R/W
7 6 5 4 3 2 1 0
AUX reserved SPDIF_MODE[1:0]
Address: 0x5E This register controls the SPDIF mode:

Type: R/W
Software Reset: NC Description:
Hardware Reset: UND
SPDIF_MODE[1:0] 00: OFF, the IEC60958 is not working and the output line is idle,
AUX = ’0’ 01: MUTE, the outputs are PCM null data,
10: PCM, the outputs are PCM data and only the Left/Right channels are transmitted,
11: EMC, in this "encoded" mode the compressed bitstream is transmitted in IEC61937
standard.
SPDIF_MODE[1:0] 10: PCM, the outputs are PCM data and only the "VCR" channels are transmitted.
AUX = ’1’ All other values are reserved.
Category code
SPDIF_CAT
7 6 5 4 3 2 1 0
CATCODE
47/90
STA310
Address: 0x5F
Type: R/W
Software Reset: NC
Hardware Reset: UND
Description:
The table below defines the category codes, values not listed are reserved.
Category code Description
0000000 General
1000000 Experimental
XXX0000 Reserved
XXX1000 Solid State Memory
0000100 Broadcast reception of dig. audio Japan
1100100 Broadcast reception of dig. audio United states
0001100 Broadcast reception of dig. audio Europe
1000100 Broadcast reception of dig. audio Electronic Software delivery
XXXX100 Broadcast reception of dig. audio All other states are reserved
0000010 Digital / Digital converters and signal processing PCM encoder/decoder
0100010 Digital / Digital converters and signal processing Digital / Digital Digital sound sampler
0010010 converters and signal processing r Digital signal mixe
0011010 Digital / Digital converters and signal processing Sample rate converter
XXXX010 Digital / Digital converters and signal processing All other states are reserved
XX00110 A/D converter W/o copyright
XX10110 A/D converter W/ copyright (using copy and L bits)
XXX1110 Broadcast reception of dig. audio
0000001 Laser optical CD - Compatible with IEC 908
0001001 Laser optical CD - Not compatible with IEC 908 (Magneto optical)
X X X X 0 01 Laser optical All other states are reserved
0 0 0 010 1 Musical instruments, microphones, etc. Synthesizer
0 0 0 110 1 Musical instruments, microphones, etc. Microphone
X X X X101 Musical instruments, microphones, etc. All other states are reserved
0000011 Magnetic tape or disks DAT
0001011 Magnetic tape or disks Digital audio sound VCR
X X X X 011 Magnetic tape or disks All other states are reserved
X X X X111 Reserved
7 Only cat. codes XXXX100, XXX1110, XXXX001 -> L bit
0 Original, commercially pre-recorded data
1 No indication of 1st generation or higher
7 All other categories
0 No indication of 1st generation or higher
1 Original, commercially pre-recorded data
SPDIF_CONF
SPDIF PCMCLK divider
7 6 5 4 3 2 1 0
LAT SM=0 RND DIV[4:0]
Address: 0x60
Type: R/W
Software Reset:NC
48/90
STA310
Hardware: Reset:UND
Description:
DIV[4:0] This field is the DAC_PCMCLK divider. It must be set according to the formula:
in 16 bit mode: IECDIV=(1+PCMDIV)/2-1; in 32 bit mode: IECDIV=PCMDIV
RND This bit is used to have a "16-bit rounding" on the SPDIF (when in PCM mode):
0: no rounding,
1: rounding.
This bit has no effect on the precision of analogue data
SM SYNC MUTE Mode, must be set to zero.
LAT Configures the latency mode between the SPDIF output (in mode compressed) and the Audio output.
0: Auto-Latency: The latency is the transmission time for 2/3 of the payload, plus the time to decode
an audio block.
For MPEG Auto-Latency, the latency is the following time depending of the sampling frequency in the
incoming bitstream: MPEG 48KHz: 20.90ms, MPEG 44.1KHz: 22.95ms, MPEG 32KHz: 32.53ms.
1: User-programmable latency - the SPDIF_LATENCY register is used.
The table below shows the relationship between the value of the IEC divider and the value of the PCM divider.
PCM Divider Value Mode Description IEC Divider Value
5 DAC_PCMCLK = 384Fs, DAC is 16-bit mode 2
3 DAC_PCMLK = 256 Fs, DAC is 16-bit mode 1
SPDIF_STATUS
SPDIF status bit
7 6 5 4 3 2 1 0
SFR PRE COP COM
Address: 0x61
Type: R/W
Software Reset: NC
Hardware: Reset UND
Description:
49/90
STA310
This register is used to set the value of the status bit in the IEC958 data stream.
COM Compress data bit.
1: compressed mode
0: non compressed mode.
COP 1: copy allowed
0: copy not allowed
PRE 1: output has pre-emphasis
0: output does not have pre-emphasis
SFR 0000: if sampling frequency = 44.1KHz
0010: if sampling frequency = 48KHz
SPDIF_REP_TIME
SPDIF repetition time of a pause frame
7 6 5 4 3 2 1 0
Address: 0x75
Type: R/W
Software Reset: NC
Hardware: Reset: UND
Description:
In compressed mode, a burst of pause frames is sent when there are no more data to transmit (due to an error
or a gap in the incoming bitstream, for example).
This register sets the size of a pause frame in IEC frames: Dolby Digital =4, MPEG=32 and DTS=3.
SPDIF_LATENCY
Latency value
7 6 5 4 3 2 1 0
Value
Address: 0x7E
Type: R/W
Software Reset: NC
Description:
If bit LAT of register SPDIF_CONF is set, a delay can be configured between the output of IEC61937 in com-
pressed mode and the output of the audio decoder. To configure a latency (in unit of seconds) this register has
to be set according the following formula:
Value = L x FS/8 where, L=Latency in s and FS=Sampling frequency in Hz
50/90
STA310
The minimum latency delay is 0; the maximum laten- Hardware: Reset: UND
cy delay is the time to decode a frame:
Bitfield Description Description:
Value Dolby Digital: L = 1536 samples / Sampling The feature is only available for STA310 cut 2.0. This
frequency register is used to enable the autodetection on the S/
MPEG: L = 1152 samples / Sampling PDIF. When high, the autodetection is present. When
frequency low, autodetection is disable.
The STA310 cut 2.0 is able to detect the following au-
dio format changes on the S?PDIF input.
SPDIF_DTDI FROM TO
SPDIF data-type information AC3 PCM

AC3 MPEG
7 6 5 4 3 2 1 0
MPEG AC3
PFC DTD INF
MPEG PCM
Address: 0x7F PCM AC3
Type: R/W PCM MPEG
Software Reset: NC
The host must respond to the RST and LCK interrupt
Hardware: Reset: UND in order for the STA310 to take into account the
change of the audio format.
Description::
Bitfield Description AUTODETECT_SENS
S/PDIF Autodetection Sensitivity
DTDI[4:0] in Dolby Digital mode: 4 3 2 1 0
7 6 5 4 3 2 1 0
0 0 BSMOD
[2..0] SENS
in MPEG mode: 4 3 2 1 0 Address: 0xE1
0 0 0 DR K
Type: R/W
Software Reset: 0
DTD 1: Data-type dependent information used
for the SPDIF in compressed mode, can be Hardware: Reset: UND
set by the user.
Refer to IEC958 standard for more
information. Description:
The feature is only available for STA310 cut 2.0. This
0: Transmitted DTDI are extracted from the register is used to configure the autodetection sensi-
stream. tivity. The lower is SENS, the faster is the autodetec-
PFC 1: PCMCROSS function enabled tion. Typical value is 0.
AUTODETECT_ ALIGN
AUTODETECT_ENA
S/PDIF Autodetection Alignement
S/PDIF Autodetection Enable
Address: 0xE2
7 6 5 4 3 2 1 0 Type: R/W
ENA Software Reset: 0
Address: 0xE0
Type: R/W
Description:
Software Reset: 0
51/90
STA310
The feature is only available for STA310 cut 2.0. This

register is used to configure the Left/Right sample ■ When in idle mode, the PLAY value is not taken
alignement of the S/PDIF. Typical value is 10 into account by the decoder.
■ When in init mode, the PLAY value is not taken

into account by the decoder.
■ When in decode mode, PLAY enables the

9.9 Audio command registers decoding, see table below:
SOFTRESET DAC_SCLK,
PLAY MUTE DAC_PC
Soft reset DAC_LRCLK Decoding
value value MOUT
state
7 6 5 4 3 2 1 0
0 0 Not running 0 No
0 1 Running 0 No
Address: 0x10 1 0 Running Decoded Yes

samples
Type: W0
1 1 Running 0 Yes
Software Reset: NA
Hardware: Reset: NA
MUTE
Description: Mute
When bit 0 of this register is set, a soft reset occurs. 7 6 5 4 3 2 1 0
The command registers and the interrupt registers MUTE
listed below are cleared.
The decoder goes into idle mode and the volumes Address: 0x14
are cleared. Type: R/W
Command registers: Software Reset: 0
MUTE, RUN, PLAY, SKIP_MUTE_CMD and Hardware: Reset: 0
SKIP_MUTE_VALUE
Interrupt registers:
INTE, INT and ERROR Description:
The MUTE command is handled differently according
to the state of the decoder:
■ When in idle mode after hardware reset, setting
PLAY MUTE to “1” automatically runs the DAC_SCLK
Play and DAC_LRCLK clocks and outputs them to
the DACs.
7 6 5 4 3 2 1 0
■ When playing, setting MUTE to “1” mutes the
PLAY PCM outputs.
■ The MUTE register has no effect on the SPDIF
Address: 0x13 output.
Type: R/W
Software Reset: 0 RUN
Hardware: Reset: 0 RUN decoding
7 6 5 4 3 2 1 0
Description:
RUN
The PLAY command is treated according to the state
of the decoder: Address: 0x72
52/90
STA310
Type: R/W Skip frames or mutes blocks of frame

Software Reset: 0
7 6 5 4 3 2 1 0
Hardware Reset: 0
Description: Address + 0x74

This register enables to exit from idle mode. After a Type: R/W
soft or hard reset the decoder is in idle mode. It stays Software Reset: 0
in this mode until the RUN is set.
Hardware Reset:0
In run mode the decoder takes into account the state
of all the configuration registers and begins to de-
code. Description:
The RUN register can only be reset by the SOFTRE- The value in this register gives either the number of
SET command. frames to skip or the number of blocks during which
the decoder will be stopped. This is used in conjunc-
tion with register .
SKIP_MUTE_CMD
Skip or mute commands
7 6 5 4 3 2 1 0
9.10 Interrupt register
reserved MUTE reserved PAU BLK SKP SMUT INTE
Address: 0x73 Interrupt enable
Type: R/W 7 6 5 4 3 2 1 0
Software Reset: 0 @0x08 INTE[15:8]
Hardware Reset: 0 @0x07 INTE[7:0]
Address: 0x08 - 0x07

Description:
Type: R/W
This register cannot be used in MP3 decoding
Software Reset: 0
mode.The register is taken into account at a begin-
ning of decoding a frame. Hardware Reset: 0
Value Description
Description:
SMUT, If one or both of these bits is ’1’ then the This register is used to enable each interrupt inde-
MUTE decoder continues the normal decoding
pendently. Setting a bit in the register enables the
process, but the output samples are soft-
muted to zero. When both these bits are ’0’
corresponding interrupt.
muting is disabled and the decoder plays
the incoming frame.
INT
SKP Skip frame. The decoder skips the number Interrupt
of frames programmed in register
7 6 5 4 3 2 1 0
BLK Pause block. The decoder introduces a
delay equal to the number of blocks @0x0A INTE[15:8]
programmed in register @0x09 INTE[7:0]
PAU The decoder is stopped whilst this bit is ’1’. Address: 0x0A - 0x09
Reserved Set to ’0’. Type: RO
Software Reset: 0
Hardware Reset:0
SKIP_MUTE_VALUE
53/90
STA310
Description: These registers indicate which interrupt Bit

occurred. Provided an interrupt is enabled through Name Condition Signalled
Numer
the register INTE, if the corresponding bit of INT is
11 FIO FIFO Input has Overflowed (2)
set, the corresponding interrupt has occurred. The
signal IRQ is activated whenever one of the bits of 12 RST(3) The STA310 has detected a change
INT become set. Depending on the nature of the con- in the incoming audio format. The
dition, clearing a bit in INT register is performed by ei- soft Reset produce must be applied
ther reading the MSB of INT register, or by reading and the device must be re-initialized
the MSB of the associated condition registers (see according to the new audio format
below). This register is reset by software reset. detected. Registers DECODESEL
(0x4d) and STREAMSEL (0x4c)
The table below shows the interrupt nature indicated
contain the new audio format (1)
by each bit..
13 LCK(3) A break has occurred in the S/PDIF
Bit stream causing the internal S/PDIF
Name Condition Signalled
Numer PLL to get unlocked. The soft reset
procedure must be applied and the
0 SYN Change in Synchronization Status (1) device must be re-initialized
1 HDR according to the current audio format
Valid Header Registered (1) decoding contained in the registers
2 ERR DECODESEL (0x4d) and
Error Detected (1)
SRTREAMSEL (0x4c). (1)
3 SFR Sampling frequency changed (2) 14 USD Reserved
4 DEM De-emphasis Changed (2)
15 TBD Reserved
5 BOF First Bit of New Frame at Output Notes: 1. Cleared when a reset occurs or when the MSB of the in-
Stage (2) terrupt register is read
2. Cleared when a reset occurs or when the MSB of the
6 PTS First Bit of New Frame with PTS at corresponding register is read. Affected registers are
Output Stage (2) listed in the following table
3. Only available in STA310 cut 2.0
7 ANC Not implemented
Address Name
8 PCM PCM Output Underflow (2)
0x0F ERROR
9 FBF Frame Buffer Full: The frame buffer
memory contains 2 frames: one 0x40 SYNCSTATUS
decoded, and one parsed for next
decoding 0x41 ANCCOUNT
10 FBE Frame Buffer Empty: The frame 0x42 HEAD 4
buffer memory contains 1 frame
which begins to be decoded. The 0x46 PTS [32]
next frame begins to be parsed
9.11 Interrupt status registers

SYNC_STATUS
Synchronization status
7 6 5 4 3 2 1 0
PAC FRA
Address: 0x40
Type: RO
Software Reset: UND
Hardware Reset: UND
54/90
STA310
MPEG_2
Description:
0 0 0 0 0 0 DR K
This register indicates the status of the audio parser
for synchronization. It is used in conjunction with OTHER
PACKET_LOCK and SYNCK_LOCK registers. On
read the synchronization status interrupt bit is 0 0 0 0 0 0 0 0
cleared (INT.SYN is cleared).
Address: 0x42
Type: RO
FRA Frame Status
0 0: Research audio synchronization Software Reset: UND
0 1: Wait for confirmation - a synchro word has Hardware Reset: UND
been detected but the parser has not yet
detected SYNC-LOCK+1 synchro words.
1 0: Synchronized - SYNC_LOCK + 1 synchro Description:
words have been detected
1 1: Not used This register contains header data HEAD[31:24]. The
contents depend on the type of the
PAC Packet Status
0 0: Research packet synchronization word
frame.HEAD4[7:3] = 00000, in all cases.
0 1: Wait for confirmation - - a synchro word When the host reads this register, the corresponding
has been detected but the parser has not yet interrupt bit (HDR) is cleared.
detected
PACKET_LOCK+1 synchro words.
1 0: Synchronized - PACKET_LOCK + 1 Dolby Digital
synchro words have been detected
1 1: Not used Bitfield Description
ANCCOUNT HEAD4[2:0] BSMOD if an Dolby Digital frame

Ancillary data
MPEG-2
7 6 5 4 3 2 1 0
HEAD4[2] 0
Address: 0x41 HEAD4[1] DR=1 Dynamic range exists
Type: RO HEAD4[0] K=0 in normal mode, K=1 in Karaoke
mode.
Software Reset: UND
Hardware Reset: UND OTHER
In all other types of frame HEAD4[2:0] = “000”
Description:
This value gives the number of ancillary data in the HEAD3
stream. The ancillary data interrupt bit ANC of the HEADER 3 register
register is cleared by a read.
7 6 5 4 3 2 1 0
HEAD4 0 0 0 DTYPE
HEADER 4 register Address: 0x43

AC_3 Type: RO
7 6 5 4 3 2 1 0 Software Reset: UND
0 0 0 0 0 BSMOD
Hardware Reset: UND
Description:
55/90
STA310
This register contains header data Address: 0x46 to 0x4A

HEAD[23:16].HEAD3[7:5]=“000”, in all cases Type: R/W
HEAD3[4:0] = DTYPE
Software Reset: UND
DTYPE is the data type and is defined as follows:
Hardware Reset: UND
Bit Description
DTYPE 0000: Null data or Linear PCM Description:
0001: Dolby Digital When the PTS interrupt is activated, a new PTS val-
0100: MPEG-1 Layer I ue is stored in this register. Once the PTS[32] value
0101: MPEG-1 Layer II or MPEG-2
is read bit PTS of the PTS register is cleared.
word extension
0110: MPEG-2 Layer II with extension ERROR
1001: MPEG-2 Layer II low sample ERROR code
rate
7 6 5 4 3 2 1 0
(11) 1011: DTS-1 (Frame Size 512)
(12) 1100: DTS-2 (Frame Size 1024)
(13) 1101: DTS-3 (Frame Size 2048) Address + 0x0F
This register can not detect the data-type of data in a stream. Type: RO
Software Reset: 0
HEADLEN Hardware Reset:0
Frame length
Description:
7 6 5 4 3 2 1 0
This is a status register, when read by the ST20, this
HEADLEN[15:8] and the corresponding interrupt register are cleared.
HEADLEN[7:0] This 7-bit register is ANDed with 0x7F to get the cor-
rect value. The value in the ERROR register indi-
Address: 0x44 - 0x45 cates the type of error that has occurred. These
Type: RO errors are defined in the table below.
Software Reset: UND Value
Error Name
Hardware Reset: UND (decimal)
Dolby Digital decoding
No error 0
Description:
EXPAND_DELTA_PAST_END_ARRAY 1
The HEADLEN register contains the bit length of the
compressed data frame HEAD[15:0]. HEADER reg- XDCALL_TRY_TO_REUSE_REMAT_FLG 2
isters are all updated as soon as the decoder begins XDCALL_TRY_TO_REUSE_COUPLING_ST 3
to decode a frame. RA
XDCALL_CANT_COUPLE_IN_DUAL_MODE 4
PTS XDCALL_TRY_TO_REUSE_CPL_LEAK 5
XDCALL_TRY_TO_REUSE_SNR 6
PTS
XDCALL_TRY_TO_REUSE_BIT_ALLOC 7
7 6 5 4 3 2 1 0 XDCALL_TRY_TO_REUSE_COUPLING_EX 8
0x46 PTS[32] PONENT_STRA
0x47 PTS[31:24] XDCALL_TRY_TO_REUSE_EXPONENT_S 9
0x48 PTS[23:16] TRA
0x49 PTS[15:8] XDCALL_TRY_TO_REUSE_LFE_EXPONEN 10
T_STRA
0x4A PTS[7:0]
XDCALL_CHBWCOD_IS_TOO_HIGH 11
56/90
STA310
Value Value
Error Name Error Name
(decimal) (decimal)
BSI_ERR_REV 12 MPEG_BITRATE_ERROR 58
BSI_ERR_CHANS 13 MP3 Decoding
CRC_NOT_VALID 14 CRC_ERROR 01
Packet Synchronization DATA_AVAILABLE_ERROR 02
SYNCHRO_PACKET_NOT_FOUND 16 ANC_PARTIAL_READ_ERROR 03
BAD_MPEGI_RESERVED_WORD 17 ANC_NOT_READ_ERROR 04
BAD_MPEG2_RESERVED_WORD 18 BAD_ID_AND_IDEX_VALUES 33
BAD_LPCM_SYNCHRO 19 LAYER_IS_NOR_LAYER3 34
UNKNOWN_STREAM_ID 20 BAD_audio_sampling_freq 35
MARKER_ERROR 21 FREE_FORMAT_NOT_SUPPORTED 36
UNKNOWN_SUB_STREAM_ID 22 BIT_RATE_NOT_SUPPORTED 37
IEC958_INPUT_MISMATCH_CONF 23 BIG_VALUE_ERROR 48
IEC958_MPEG2_LAYERI_NOT_SUPPORTE 24 MODE_CHANGE_ERROR 49
D FS_CHANGE_ERROR 50
IEC958_PAUSE_FRAME_NOT_SUPPORTE 25 Miscellaneous
D
IEC_958_READ_ERROR 64
IEC958_BAD_DATA_TYPE_DEPENDANT 26
MPEG_FB_BYPASS_AREA_ERROR 65
MISMATCH _HOST_SEL_CONFIGURATION 27
SKIPPING_BITS_IN_FB_ERROR 66
Audio Synchronization
LATENCY_TOO_BIG 67
SYNCHRO_AUDIO_NOT_FOUND 32
SKIP_MUTE_ERROR 68
BAD_CRC_AC3 33
UNKNOW_SFREQ_FOR_LATENCY 69
BAD_LPCM_QUANTIZATION_WORDLENG 34
TH LATENCY_TOO_SMALL 70
BAD_AUDIO_SAMPLING_FREQUENCY 35 BAD_INPUT_CHAN 71
BAD_MPEG_LAYER 36 INVALID_ALPHA_COEFF 72
MPEG_BITRATE_FREE_FORMAT 37
9.12 Decoding algorithm registers
NOT_SUPPORTED_AC-3_FRMSIZECOD 38
The table below shows how the STREAMSEL and
BAD_CRC_MPEG_EXTENDED 39 DECODSEL registers should be programmed for dif-
BAD_MPEG_EXTENDED_RESERVED_BIT 40 ferent types of bitstream.
MPEG_EXTENDED_SYNC_NOT_FOUND 41
MPEG_EXTENDED_LENGTH_TOO_SMALL 42 Table 9.
BAD_SAMPLES_PER_CHANNEL 44 STREAMSEL and DECODESEL programming
BAD_FRAME_BIT_SIZE 45
definitions
MPEG Decoding STREAMSEL DECODSEL
Mode
(0x4C) (0x4D)
MPEG_EXTENSION_ERROR 48
0 0 MPEG2 PES carrying
MPEG_MC_MUTE 49
Dolby Digital (ATSC)
NOT USED 50
NOT USED 51 MPEG1 frames
MPEG_LAYER_ERROR 52 0 2 MPEG2 PES carrying
MPEG_CHCONFIG_ERROR 53 MPEG2 frames
MPEG_MC_PREDICTION_ERROR 54 1 0 MPEG2 PES carrying

Dolby Digital frames for
MPEG_CRC_ERROR 55 DVD video
MPEG_EXT_CRC_ERROR 56
MPEG_TOO_SMALL_FOR_MC_HEADER 57
57/90
STA310
1 2 MPEG2 PES carrying This register identifies the audio data-type.

MPEG2 frames for DVD
video Bitfield Description
DEC[3:0] 0000: Dolby Digital Decoding
linear PCM for DVD
video 0001: MPEG1
0010: MPEG2
2 1 MPEG1 packet carrying 0011: PCM/LPCM
MPEG1 audio
0100: PINK NOISE generator
3 0 Dolby Digital frames 0101: CD_DA
elementary streams 0111: PCM beep tone generator
1001: MP3
3 1 MPEG1 frame
elementary streams
3 2 MPEG2 frame
elementary stream STREAMSEL
3 3 Stereo PCM (16bits STREAM selection
samples) and PCM2
channels 7 6 5 4 3 2 1 0
3 4 Pink noise generator
STRSEL
3 5 CDDA (Stereo PCM 16
bits samples) Address: 0x4C
3 7 Activate PCM beep tone
Type: R/W
3 9 MP3 elementary
streams Software Reset: NC
5 0 IEC61937 Input with Hardware Reset: UND
Dolby Digital frames Description:
5 1 IEC61937 Input with
MPEG1 frames Bitfield Description
5 2 IEC61937 Input with STRSEL 000: PES
MPEG2 frames
001: PES DVD video
5 6 IEC61937 Input with 010: Packet MPEG1
DTS frames 011: Elementary Stream/IEC60958
6 3 Linear PCM for DVD 100: Reserved
audio 101: SPDIF IEC61937
110: PES DVD audio
DECODSEL
Decoding algorithm
7 6 5 4 3 2 1 0
DEC
9.13 System synchronization registers
Address: 0x4D PACKET_LOCK
Type: R/W Packet lock
Software Reset: NC
7 6 5 4 3 2 1 0
Hardware Reset: UND
Description: Address: 0x4F

Type: R/W
Software Reset: NC
Hardware Reset: UND
58/90
STA310
Description: For MPEG1 packets or PES, the 5 LSB bits are sig-
This register specifies the number of supplementary nificant. For DVD PES (LPCM, Dolby Digital or
packet synchro words that the packet parser must MPEG), the 3 LSB bits are significant (see audio
detect before it is considered as synchronized, and pack definition in DVD specifications).
can send data to the audio parser (max=1, min=0). In These bits correspond to the stream number defined
this way, stream data can not be sent to the audio in the STREAM_ID field of the audio packet header,
parser instead of packet sync words. except for DVD, Dolby Digital or LPCM packets,
where they correspond to the stream number defined
PACKET_LOCK = 0: the packet parser is synchro- in the SUB_STREAM_ID field.
nized when it has detected one packet synchro word.
PACKET_LOCK = 1: the packet parser is synchro-
nized when it has detected two packet synchro ID_EXT
words. Audio extension
7 6 5 4 3 2 1 0
ID_EN
Enable audio ID
Address + 0x52
7 6 5 4 3 2 1 0
Type: R/W
Software Reset: NC
Type: R/W
Software Reset: NC Description:
Hardware Reset:UND The 3 LSB bits of this register are significant. In case
of DVD MPEG2 audio with extension bitstream (see
DVD specifications), this register is used to select the
Description: stream defined in the STREAM_ID of the packets
If set to 1, the audio decoder decodes only the stream containing MPEG2 extension bit stream data.
corresponding to the stream-id or sub-stream-id of
the packet layer. This selection is done through
AUDIO_ID or AUDIO_ID_EXT registers. If set to 0, SYNC_LOCK
the decoder decodes all the audio packets. SYNC lock
7 6 5 4 3 2 1 0
ID
Audio ID
Address: 0x53
7 6 5 4 3 2 1 0
Type: R/W
Software Reset: NC
Address : 0x51 Hardware Reset: UND

Description:
Type: R/W
Software Reset: NC This register specifies the number of supplementary
audio synchro words that the audio parser must de-
Hardware Reset: UND tect before it is considered as synchronized, and can
send data to the decoder.
Description: In this way, stream data can not be sent to the decod-
When decoding packets, it is possible to specify an er instead of audio sync words. Max value = 3; min
identifier for a selected program. AUDIO_ID must be value = 0. SYNC_LOCK = 0, the audio parser is syn-
chronized when it has detected one audio synchro
written with the packet ID. This feature is enabled
word. SYNC_LOCK = n > 0, when the audio parser
when the register AUDIO_ID_EN is set, and only
has detected one synchro word, it waits until it de-
packets with matching ID are decoded.
tects n supplementary audio sync words.
59/90
STA310
When it has detected (SYNC_LOCK+1) sync words, Description:

it sends the data to the decoder.
PL_AB When high, select the auto-balance
function (used to track out gin
mismatches between Lt and Rt).When
9.14 Post decoding and pro logic registers low, disable the autobalance function
PDEC1 PL_WS When high, enable Wide Surround mode.
Post decoder register The wide surround option is provided for
users who plan to do further post-
processing of the Pro Logic outputs, and
7 6 5 4 3 2 1 0
want to fold the lowpass filtering of the
surround channel into their downstream
VMAX DEM DCF DB PVIRT MPEG_DR PL
processing Lowpass filtering and B-type
shelf filtering are both disabled. When
Address: 0x62 low, disable ide surround mode.
Type: R/W
Software Reset: NC
PL_DWNX
Hardware Reset: UND
Pro Logic Decoder Downmix
7 6 5 4 3 2 1 0
Description:
LFE_BYP PL_DWNX[2:0]
This register controls the post decoder operations.
Address : 0x65
Type: R/W
PL When high Pro Logic decoding is forced,
when low the ProLogic decoder is
Software Reset: NC
automatically enabled when the stream Hardware Reset: UND
contains the info that it is ProLogic
encoded
Description:
MPEG_DR When high enable MPEG dynamic range.
DB When high the "double stereo" procedure PL_DWNX 0,1,2: Pro Logic disabled
is enabled. Double stereo is a copy of [2:0] 3: 3/0 (L, R, C) three stereo
Left/Right channels on to Left/Right 4: 2/1 (L, R, Ls) phantom
surround channels in order to have a 5: 3/1 (L, C, R, Ls)
pseudo 5 -channel decoder effect. 6: 2/2 (L, R, Ls, Rs) phantom
7: 3/2 (L, C, R, S, S)
DCF When set the DC filter is activated.. When
reset, it is disabled. LFE_BYP 0: LFE channel is clear
1: LFE channel is bypassed
DEM When set the de-emphasis filter is
activated.. When reset, it is disabled. Prologic decoder support only 4 sampling frequen-
cies: 48KHz, 44.1KHz, 32KHz, 22.05KHz.
PL_AB If we active prologic with sampling frequency differ-
Pro Logic auto balance ent from those frequencies, the decoder will be auto-
matically disable the prologic call.
7 6 5 4 3 2 1 0
PL_WS PL_AB OCFG
Address : 0x64 Output Configuration
Type: R/W 7 6 5 4 3 2 1 0
Software Reset: NC LFE_BYP BOOST OCFG[2:0]
Hardware Reset: UND
Address: 0x66
60/90
STA310
Type: RW The description is provided in the output configura-

Software Reset: NC tion section.
Hardware Reset: UND This register is used to indicate the output configura-
tion chosen.
- LP means Low pass filter
Description:
- HP means High pass filter
6 output configurations are provided that redirects
bass on subwoofer channel, and applies some filters
on channels.
Bitfield conf Meaning Description
OCFG[2:0] Bass management configuration according to the bass direction scheme from
Dolby. For configurations 2,3,4 the subwoofer can be output is bit LFE asset to
high. For all other configurations, the LFE bit has no effect.
0 ALL All channels are rounded according to the selected output precision, (24b -> 16b,
24 -> 18b.) and scaled (volume control) only.
1 LSW Low frequencies are extracted from the six input channels and redirected to the
subwoofer. SUB = LP(L+R+Ls+Rs+C+LFE).Low frequencies are removed from all
channels L = HP(L), R = HP(R), C = HP(C), Ls = HP(Ls), Rs = HP(Rs).
2 LLR Low frequencies are extracted from C, LFE, Ls and Rs channels and redirected to
left and right channels: C = HP (C), Rs = HP (Rs), Ls = HP(Ls), L = L + LP
(C+LFE+Ls+Rs), R = R + LP(C+LFE+Ls+Rs). If subwoofer is output, SUB = LP
(LFE+C+Ls+Rs).
3(1) SLP Low frequencies are redirected to the left, right and surround channels or cab be
output on the subwoofer.
If sub-woofer is output, SUB = LFE, L = L + LP(C), R = R + LP(C), Ls = Ls, Rs = Rs
If sub-woofer is not output, L = L + LP(C) + LFE, R = R + LP(C) + LFE, Ls = Ls +
LFE, Rs = Rs + LFE.
4 SIMP Simplified configuration. Low frequencies are exrtacted from C, Ls, Rs and LFE.
If subwoofer is output, SUB = LP(C+Ls+Rs) + LFE, L = L, R = R.
If sub-woofer is not output, SUB = LFE, L = L + (C+Ls+Rs), R = R + (C+Ls+Rs).
5 BYP BYPASS, All channels are directly routed to PCM outputs.
6 configuration 1 without filters.
BOOST(1) Channel level, enables boost:

if OCFG_num = 2 :
0 : No +12dB boost on left and right channels
1 : +12dB boost on left and right channels
When configuration = 3 :
If sub-woofer is output :
0 : No +4dB boost on all channels
1 : +4dB boost on all channels
If sub-woofer is not output :
0 : No +8dB boost on all channels
1 : +8dB boost on all channels
LFE_BYOP 0; LFE channel is clear
1: LFE channel is bypassed
Note: 1. In configuration 3 with subwoofer enabled, the output of
the subwoofer is 10dB greater than expected. Therefore
when using this mode, the subwoofer output level needs DWSMODE
to be attenuated 10dB in order to match the subwoofer
output levels of other bass management configurations.
In general be carefull while using the boost option since
it has the potential of causing the woofer output to over-
load
61/90
STA310
Downsampling filter ■ If CHAN_IDX = 2, then VOLUME0 can be

written with the attenuation that will be applied
7 6 5 4 3 2 1 0 to Left Surround channel.
■ If CHAN_IDX = 5, then reading VOLUME0
provides the attenuation that is applied to Left
Address: 0x70 channel.
Type: R/W
provides the attenuation that is applied to
Software Reset: NC Center channel.
Hardware Reset: UND ■ If CHAN_IDX = 7, then reading VOLUME0
provides the attenuation that is applied to Left
Surround channel.
Description:
■ Other values of CHAN_IDX are reserved.
This register controls the downsampling filter for the
LPCM video, LPCM audio modes. When decoding a
96kHz DVD-LPCM stream, it might be necessary to VOLUME1
downsample the stream to 48kHz. Volume of second channel
Value Mode 7 6 5 4 3 2 1 0
0 Automatic (according to bitstream) AVol1
1 Force Downsampling
Address: 0x63
2 Suppress Downsampling
Type: RWS
Software Reset: 0
Hardware Reset: UND
9.15 Bass redirection registers

Description:
VOLUME0
This register reads or writes the attenuation that is
Volume of first channel applied to the channel selected by CHAN_IDX. The
volume of the right channel can be set up with a
7 6 5 4 3 2 1 0 0.5dB step.
AVol0 ■ If CHAN_IDX = 0, then VOLUME1 can be
Address: 0x4E to Right channel.
Type: RWS ■ If CHAN_IDX = 1, then VOLUME1 can be
Software Reset: 0 to Subwoofer channel.
Hardware Reset: UND ■ If CHAN_IDX = 2, then VOLUME1 can be
Description: to Right Surround channel.
This register reads or writes the attenuation that is
applied to the channel selected by CHAN_IDX. The provides the attenuation that is applied to Right
volume of the left channel can be set up with a 0.5dB channel.
step. ■ If CHAN_IDX = 6, then reading VOLUME1
■ If CHAN_IDX = 0, then VOLUME0 can be

provides the attenuation that is applied to
written with the attenuation that will be applied Subwoofer channel.
to Left channel. ■ If CHAN_IDX = 7, then reading VOLUME1
■ If CHAN_IDX = 1, then VOLUME0 can be

provides the attenuation that is applied to Right
written with the attenuation that will be applied Surround channel.
to Center channel. ■ Other values of CHAN_IDX are reserved.
62/90
STA310
and CHAN_IDX value is automatically

CHAN_IDX changed to 4.
Channel Index
7 6 5 4 3 2 1 0
9.16 Dolby Digital configuration registers
Reserved CHAN_IDX
AC3_DECODE_LFE
Address: 0x67 Decode LFE
Type: R/W
7 6 5 4 3 2 1 0
Software Reset: 4
Hardware Reset: UND
Address: 0x68
Description:
Type: R/W
This register identifies the pair of channels and the
type of access: Software Reset: NC
Hardware Reset: UND
Bitfield value Channel pair Access comment
CHAN_I 0 Left and Right write Description:

DX When this register is set to 1, the device decodes
1 Center and write LFE channel (if present).
Subwoofer
2 Left surround write AC3_COMP_MOD

and right
Compression mode
surround
7 6 5 4 3 2 1 0
3 reserved reserved
4 no pair none Indicates

selected that volume
can be read
Address + 0x69
or written Type: R/W
Software Reset: NC
5 Left and Right read
Hardware Reset: UND
6 Center and read
Subwoofer
Description:
7 Left surround read The value of this register defines the compression
and right mode. In custom A mode, the dialog normalization
surround function is not done by the audio decoder, it has to be
done by an external analog part. In all other modes
■ To read a volume, the register CHAN_IDX the normalization is done by audio decoder.
must be set to the appropriate value. The DSP
indicates that the attenuation is readable Value Meaning
through registers VOLUME0 and
VOLUME1 by changing automatically the 0 Custom A (Analog)
CHAN_IDX to value 4.
1 Custom D (Digital)
■ To write a volume, the attenuation of the pair of
channel should be written in VOLUME0 and 2 Line Out
VOLUME1 registers. Then the CHAN_IDX
3 RF Mode
register is written to the appropriate value. The
attenuation is updated on the next audio block
63/90
STA310
AC3_HDR level signals amplification is maximum.

High dynamic range A value of zero disables the low-level amplification.
This word is ignored if the compression mode is set
7 6 5 4 3 2 1 0 to RF mode.
AC3_RPC
Address: 0x6A Repeat count
Type: R/W
7 6 5 4 3 2 1 0
Software Reset: NC
Hardware Reset: UND
Address + 0x6C
Description:
Type: R/W
This register corresponds to the Dynamic range
scale factor for high level signals, also called cut fac- Software Reset: NC
tor in the Dolby specifications. Hardware Reset: UND
HDR = 255 * Cut Factor (in decimal), where the cut
factor is a fractional number between 0 and 1. It is Description:
used to scale the dynamic range control word for
high-level signals that would otherwise tend to be re- When a CRC error is detected, previous blocks can
duced. be repeated or muted.
When HDR = 0xff (cut factor = 1.0), the high level sig- This register specifies the number of audio blocks to
nals reduction is the one given in the stream. repeat before muting. If this is zero, then blocks are
muted until the next frame is decoded
A value of zero disables the high-level compression.
This word is ignored if the compression mode is set
to RF mode. AC3_KARAMODE
Karaoke downmix
AC3_LDR
7 6 5 4 3 2 1 0
Low dynamic range
7 6 5 4 3 2 1 0
AudioBaseAddress + 0x6D
Type: R/W
Address : 0x6B Software Reset: NC
Type: R/W Hardware Reset: UND
Software Reset: NC
Hardware Reset: UND Description:
Downmix mode when a karaoke bit stream is re-
Description: ceived.
This register corresponds to the Dynamic range A Karaoke bitstream can be composed of 5 chan-
scale factor for low level signals, also called boost nels, which are: L (left), R (right), M (Music), V1(Vocal
factor in the Dolby specifications. 1), V2 (Vocal 2).
LDR = 255 * BoostFactor (in decimal), where the There are two major modes when receiving a
boost factor is a fractional number between 0 and Karaoke bitstream: aware and capable.
1.0. The boost factor scales the dynamic range con- When in 'aware' mode ( KARAMODE = 0), a pre-
trol-word for low-level signals that would otherwise defined downmix is applied on all incoming channels.
tend to be amplified. When in 'capable' mode ( KARAMODE = 4, 5, 6, 7),
When LDR = 0xff (boost factor = 1.0), and the low the user can choose to reproduce or not the two in-
64/90
STA310
coming vocal channels, V1 and V2. when in 2/0 output mode or when receiving a “Dual
An additional mode is added ( AC3_KARAMODE = mode” incoming bitstream (example: A disk with 2 dif-
3) to allow multi-channel reproduction. In this case, ferent languages on channel 1 and channel 2). In the
the downmix specified by the AC3_DOWNMIX and following table, channel 1 and 2 represent the output
AC3_DUALMODE registers is applied. channels after downmix performed with
AC3_DOWNMIX.
The following table summaries the different modes:
This register enables Mono downmix when
Value Mode Comment AC3_DOWNMIX = 2 and AC3_DUALMODE = 3.
0 Aware Left = L + clev*M + slev*V1, Right = R + Value Description
clev*M + slev*V2
0 Output as Stereo
1 Not used
1 Output Channel 1 on both output L/R
2 Not used
2 Output Channel 2 on both output L/R
3 Multicha Consider bitstream as multi-channel:
nnel Perform downmix according to 3 Mix Channel 1 and 2 to monophonic and
DOWNMIX and DUALMODE registers output on both L/R
4 Capable Do not reproduce V1, V2: Left = L +

clev*M, Right = R + clev*M
AC3_DOWNMIX
5 Reproduction V1 only: Left = L + clev*M
Downmix
+ 0.707*V1, Right = R + clev*M +
0.707V1
7 6 5 4 3 2 1 0
6 Reproduction V2 only: Left = L + clev*M
+ 0.707*V2, Right = R + clev*M +
0.707V2
Address: 0x6F
7 Reproduction V1, V2: Left = L + clev*M
Type: R/W
+ V1, Right = R + clev*M + V2
Software Reset: NC
Left = Output Channel, Hardware Reset: UND
Right = Output Channel, L, R, M, V1, V2 = Input
Channels (coded in Dolby Digital karaoke bitstream),
clev = Center Mix Level (value provided in the bit- Description:.
stream),
Value Description
slev = Surround Mix Level (value provided in the bit-
stream). For further information ref. to annex C of 0 2/0 Dolby Surround (LT, RT)
ATSC standard “Digital Audio Compression (AC-3)”. 1 1/0 (C)
2 2/0 (L, R)
AC3_DUALMODE
3 3/0 (L, C, R)
Dual downmix
4 (L, R, S)
7 6 5 4 3 2 1 0 5 3/1 (L, C, R, S)
6 2/2 (L, R, LS, RS - Dolby Phantom Mode
7 3/2 (L, C, R, LS, RS)
Address: 0x6E
Type: R/W Note: in notation, 3/2 represents 3 front speakers and
Software Reset: NC 2 surround speakers.
Hardware Reset: UND
AC3_STATUS0
Description
This register allows additional downmix to be set
65/90
STA310
Dolby Digital status register Dolby Digital status register 2
7 6 5 4 3 2 1 0 7 6 5 4 3 2 1 0
Not used fs_cod Bitrate code Bsmod Bsid
AudioBaseAddress + 0x76 Address + 0x78

Type: RO Type: RO
Hardware Reset:UND Hardware Reset:UND
Description: Description:
This register contains bit stream information extract- This register contains bit stream information extract-
ed from the stream. ed from the stream.
Bitfield Description Bitfield Description
Bitrate code Code identifying the bitrate. Bitrate[4..0] = Bsid Bit stream identification, indicates the version
frmsizecod[5..1] of the standard
fs_cod Code identifying the sampling frequency Bsmod Bbit stream mode, indicates the type of service
AC3_STATUS1
Dolby Digital status register 1 AC3_STATUS3
Dolby Digital status register 3
7 6 5 4 3 2 1 0
Reserved LFE Acmod 7 6 5 4 3 2 1 0
Reserved Cmixlevel SurMixlevel

Address: 0x77
Type: RO Address: 0x79
Software Reset: NC Type: RO
Hardware Reset:UND Software Reset: NC
Hardware Reset: UND
Description:
This register contains bit stream information extract- Description:
ed from the stream. This register contains bit stream information extract-
ed from the stream.
Acmod Audio coding mode. Indicates which channels Bitfield Description

are in use.
Cmixlevel Downmix level of center channel
LFE Indicates if LFe channel is present in the
SurMixlevel Downmix level of surround channel
stream
AC3_STATUS2
AC3_STATUS4
66/90
STA310
Dolby Digital status register 4 Type: RO

Software Reset: NC
7 6 5 4 3 2 1 0
Hardware Reset: UND
Reserved Dsurmod Copyright Origbs Lancode
Address: 0x7A Description:

Type: RO This register contains the code indicating the dialog
Software Reset: NC normalization level extracted from the stream.
Hardware Reset: UND
AC3_STATUS7
Description: Dolby Digital status register 7
This register contains bit stream information extract- 7 6 5 4 3 2 1 0
ed from the stream.
Room type Mix level Audprodie
Address: 0x7D
Lancode When at 1, indicates that a language code
is provided in the stream Type: RO
Origbs When at 1, indicates that the stream is an Software Reset: NC
original Hardware Reset: UND
Copyright When at 1, indicates that the stream is
protected by copyright
Description:
Dsurmod In 2/0 mode, indicates if the stream is
This register contains bit stream information extract-
Dolby surround encoded
ed from the stream.
AC3_STATUS5
Audprodie Audprodie: if set, indicates that room type
Dolby Digital status register 5 and mix level are provided
Mix level If audprodie is set, mix level indicates the
7 6 5 4 3 2 1 0
sound level
Lancode Room type If audprodie is set, mix level indicates the
sound level
Address : 0x7B
Type: RO
Software Reset: NC
Hardware Reset: UND 9.17 MPEG configuration registers
MP_SKIP_LFE
Description: Channel skip
This register contains the code of the language of the
7 6 5 4 3 2 1 0
audio service, extracted from the stream.
Reserved
AC3_STATUS6 Address : 0x68

Dolby Digital status register 6 Type: R/W
7 6 5 4 3 2 1 0 Software Reset: 0x00
Reserved Dialog Normalization (see Dolby Hardware Reset: UND
specifications)
Address:0x7C Description:
67/90
STA310
When this register is set to 1, the LFE channel is Dynamic range control
skipped. When this register is set to 0 the LFE chan-
nel is decoded (if present). 7 6 5 4 3 2 1 0
DRC
MP_PROG_NUMBER
Program number Address: 0x6A
Type: R/W
7 6 5 4 3 2 1 0
Reserved Prog
Hardware Reset: UND
Address: 0x69
Type: R/W Description;
Software Reset: 0x00 When bit DRC=1, dynamic range control is enabled.
The dynamic range is set according to the data trans-
Hardware Reset: UND
mitted in the DVD MPEG stream.
Description:
MP_CRC_OFF
When the stream is in Second Stereo mode, this reg-
CRC check off
ister specifies which program is played.
7 6 5 4 3 2 1 0
Prog Select program #0 or #1 where 0: L0,R0 in front
channels, 1: L2,R2 in front channels
Address: 0x6C
Type: R/W
MP_DUALMODE Software Reset: NC
MPEG setup dual mode Hardware Reset: UND
7 6 5 4 3 2 1 0
Description:
When register is set to 1, the CRC in MPEG frame is
Addres: 0x6E not checked. When register is set to 0, the CRC in
MPEG frame is checked if exists. If a CRC error oc-
Type: R/W curs, the decoder soft mutes the frame (but does not
Software Reset: 0x00 stop).
Hardware Reset: UND
MP_MC_OFF
The MPEG DUAL_MODE is active in downmix mode Multi-channel
1 and 9.
7 6 5 4 3 2 1 0
Value Description
Reserved DEN Reserved MC
0 Output as Stereo
Address: 0x6D
1 Output Channel 1 on both outputs L/R
Type: R/W
2 Output Channel 2 on both outputs L/R Software Reset: NC
3 Mix Channel 1 and 2 to monophonic, and Hardware Reset: UND
output on both L/R
MP_DRC
68/90
STA310
Description: (KS x RS) by (KS x S) in the above equations.
Bitfield Description Value Output Mode Comment

0x00 1/0 (C) = Mono CO = Kj x L + C + Kr x R + KS
MC When MC=1, the multi-channel part of the
(LS + RS)
bitstream is not decoded, only the MPEG-1
compatible bitstream is decoded. Bit MC must 0x01 2/0 (L, R) = LO = (L + KC x C + KS x LS)/
be set to 1 for an MPEG-1 bitstream. Stereo (1 + KC + KS),
RO = (R + KC x C + KS x RS)/
DEN De-normalization: set DEN=0 for MPEG1
signals, and set DEN=1 for MPEG2 multi- (1 + KC + KS)
channel signals 0x02 3/0 (L, C, R) LO = L + KS x LS, RO = R +
When DEN=1, MPEG2 multi-channel signals KS x RS, CO = C
L, C, R, LS and RS can be de-normalized.
The signals must first be inverse-weighted 0x03 2/1 (L, R, S) LO = L + KC x C, RO = R + KC
then multiplied by the de-normalization factor. x C, LsO = RsO = KS x (LS +
This undoes the attenuation carried out at the RS )
encoder side to avoid overload when
calculating the compatible signals (see MPEG 0x04 3/1 (L, C, R, S) LO = L, RO = R, CO = C, LsO
13818-3 specifications). = RsO = KS x (LS + RS)
0x05 2/2 (L, R, LS, RS) LO = L + KC x C, RO = R + KC

x C, LsO = LS, RsO = RS
MP_DOWNMIX
0x06 3/2 (L, C, R, LS, LO = L, RO = R, CO = C, LsO
MPEG downmix RS) = LS, RsO = RS
7 6 5 4 3 2 1 0 0x09 2/0 (Dolby LT = (L + 0.707C - 0.707 x 0.5
surround LT, RT) (LS + RS)) /2.414,
RT = (R + 0.707C + 0.707 x
0.5 (LS + RS)) /2.414
Address: 0x6F
0x0A 2/0 Karaoke Lk = L + 0.707 A1 + 0.707 G,
Type: R/W capable: V1 ON, Rk = R + 0.707 A2 + 0.707 G
Software Reset: 0x08 V2 ON
Hardware Reset: UND 0x0B 2/0 Karaoke Lk = L + 0.707 A1 + 0.707 G,
Capable: V1 ON, Rk = R + 0.707 G
V2 OFF
Description:
0x0C 2/0 Karaoke Lk = L + 0.707 G, Rk = R +
In the table below, LO, RO, CO, LsO, RsO represent Capable: V1 OFF, 0.707 A2 + 0.707 G
the output channels after downmix, and L, R, C, LS, V2 ON
RS are the audio channels.
0x0D 2/0 Karaoke Lk = L + 0.707 G, Rk = R +
The coefficients Kj, KC, Kr, KS, depend on the number Capable: V1 OFF, 0.707 G
of input channels. In the above table, the equations V2 OFF
are given for a 5 channels input bitstream. If the input
0x0E 2/0 Karaoke Lk = L + 0.707 A1 + 0.707 G,
bitstream does not contain five channels (L, C, R, LS, Capable: V1 ON, Rk = R + 0.707 A1 + 0.707 G
RS), the coefficient “Kj” corresponding to the channel V2 OFF (Dolby
not present is equal to 0. Digital like)
If the MPEG bitstream contains only one surround 0x0F 2/0 Karaoke Lk = L + 0.707 A2 + 0.707 G,
channel (S), replace (KS x (LS + RS)), (KS x LS and Capable: V1 OFF, Rk = R + 0.707 A2 + 0.707 G
V2 ON (Dolby
Digital like)
0x1A 3/0 Karaoke Lk = L + 0.707 A1, Ck = G, Rk
Capable: V1 ON, = R + 0.707 A2
V2 ON
69/90
STA310
0x1B 3/0 Karaoke Lk = L + 0.707 A1, Ck = G, Rk Description:

Capable: V1 ON, = R
V2 OFF Bitfield Description
0x1C 3/0 Karaoke Lk = L, Ck = G, Rk = R + MEX[1:0] Mode Extension

Capable: V1 OFF, 0.707 A2
V2 ON MOD[1:0] Mode
0x1D 3/0 Karaoke Lk = L, Ck = G, Rk = R PRI Private Bit

Capable: V1 OFF,
V2 OFF PAD Padding Bit
0x1E 3/0 Karaoke Lk = L, Ck = G + A1, Rk = R SFR[1:0] Sampling Frequency
Capable: V1 ON,
V2 OFF
(Dolby Digital like)
MP_STATUS2
0x1F 3/0 Karaoke Lk = L, Ck = G + A2, Rk = R
Capable: V1 OFF, MPEG status register 2
V2 ON
(Dolby Digital like) 7 6 5 4 3 2 1 0
not used C OCB EMP[1:0]

MP_STATUS0
Address: 0x78
MPEG status register 0
Type: RO
7 6 5 4 3 2 1 0 Software Reset: UND
ID LAY[1:0] P BRI[3:0] Hardware Reset: UND
Address : 0x76
Description:
Type: RO
Software Reset: UND Bitfield Description
Hardware Reset: UND EMP[1:0] Emphasis rate index

Description: OCB Original/Copy Bit
Bitfield Description C Copyright
BRI[3:0] Bit rate index
P Protection Bit
LAY[1:0] Layer MP_STATUS3
ID Identifier MPEG status register 3
7 6 5 4 3 2 1 0
MP_STATUS1
CEN[1:0] SUR[1:0] LFE AMX DEM[1:0]
Address : 0x79
7 6 5 4 3 2 1 0
Type: RO
SFR[1:0] PAD PRI MOD[1:0] MEX[1:0]
Software Reset: UND
Type: RO
Software Reset: UND
Hardware Reset: UND
70/90
STA310
Description: 9.18 Pink noise generation registers

PN_DOWNMIX
Pink noise downmix
DEM[1:0] Dematrix procedure
AMX Audio mix 7 6 5 4 3 2 1 0
LFE LFE RS LS LFE C R L
SUR[1:0] Surround
Address: 0x6F
CEN[1:0] Centre
Type: R/W
Software Reset: NC
MP_STATUS4 Hardware Reset: UND
7 6 5 4 3 2 1 0 Description:
EXT NML[2:0] MFS MLY CIB CIS Bitfield Description
Address: 0x7A L 1: Left channel contains pink noise
0: Left channel is forced to zero
Type: RO
R 1: Right channel contains pink noise
Software Reset: UND
0: Right channel is forced to zero
Hardware Reset: UND
C 1: Center channel contains pink noise
0: Center channel is forced to zero
Description: LFE 1: LFE channel contains pink noise
0: LFE channel is forced to zero
LS 1: Left surround channel contain pink noise
CIS Copyright ID Start 0: Left surround channel is forced to zero
CIB Copyright ID Bit RS 1: Right surround channel contains pink zero
MLY Multi-lingual Layer 0: Right surround channel is forced to zero
MFS Multi-lingual FS After this processing, the OCFG stage is applied on

NML[2:0] Number of Multi-lingual Channels these channels. OCFG must be configured to 0 and
attenuation on all channels must be set to 10dB at-
EXT Extension bitstream present tenuation.
The other values must not be used because low fre-
quency extraction must not be done when generating
MP_STATUS5 pink noise.
MPEG status register 5 Pink noise selection is made through the STREAM-
SEL and DECODSEL registers.
7 6 5 4 3 2 1 0
Address: 0x7B 9.19 PCM beep-tone registers

Type: RO PCM_BTONE
Software Reset: UND PCM beep tone frequency
Hardware Reset: UND
7 6 5 4 3 2 1 0
Description:
The number of extended ancillary data bytes is con-
Address: 0x68
tained in this register
71/90
STA310
Type: R/W Address: 0x82

Software Reset: 0 Type: R/W
Hardware Reset: UND Reset value: 0xFF
Description: Description:
The value in this register sets the PCM beep tone fre- This register has the same function as
quency according to the formula: KAR_MCh0VOL for the right music channel.
Beep_tone_frequency = (Fs/2)/(Register_value+1)
KAR_KEYCONT
Key control (Pitch Shift) ON/OFF
9.20 Karaoke registers 7 6 5 4 3 2 1 0

This section describes the registers which select the
Reserved PSHIFT
karaoke effects, for example: volume, chorus, echo,
reverb and mute. Any change to these registers must
be signalled to the DSP by writing 1 to the register. Address: 0x83
Type: R/W
KAR_MCh0VOL Reset value: 0
Music channel 0 (L) volume

Description:.
7 6 5 4 3 2 1 0
value
PSHIFT 0: pitch shift disabled, 1: pitch shift enabled
Address: 0x81
Type: R/W
KAR_KEYVALUE
Reset value: 0xFF
Key value
Description; 7 6 5 4 3 2 1 0
This register contains the scaling factor applied to the KEYVALUE
left channel of the music input. It specifies a fractional
multiplication factor whose value varies from 0 to 1.0: Address: 0x84
Music Left channel = original music left channel Type: R/W
scale_factor.
Reset value: 0x00
Value 0x00: Scale factor 0 = left channel mute Description:
0x7F: Scale factor 0.5 = half restitution of left The pitch shift can be changed from -3.5 to 3.5 tones,
channel
in steps of 1/4 tone. This register sets the number of
0xFF: Scale factor 1.0 = full restitution of left
channel
tones according to the following table:
Key -1/ -1/ -3/ -1 - - - - -2.49 -3.5
control 4 2 4 1.1 1.3 1.5 1.9
KAR_MCh1VOL (tone) 6 4 8 3
Music channel 1 (R) volume KEYVAL 0 1 2 3 4 5 6 7 8 9
UE
(decimal)
7 6 5 4 3 2 1 0
value
72/90
STA310
Key 1/4 1/2 3/4 1 1.16 1.34 1.58 1.93 2.49 3.5 KAR_MMUTE
control Music channel mute
(tone)
KEYVAL 10 11 12 13 14 15 16 17 18 19 7 6 5 4 3 2 1 0
UE
Reserved MUTE
(decimal)
Address: 0x87
KAR_VCANCEL
Type: R/W
Voice cancellation ON/OFF
Reset value:0
7 6 5 4 3 2 1 0
Reserved VCANCEL
Description:
This register mutes the music channel.
Address: 0x85
Type: R/W Bitfield Description
Reset value: 0 MUTE 0: not muted, 1: muted
Description:
KAR_VCh0VOL
Voice channel 0 (L) volume
VCANCELL 0: voice cancellation off,
1: voice cancellation on 7 6 5 4 3 2 1 0
Value
KAR_VVALUE
Address: 0x88
Degree of voice cancellation
Type: R/W
7 6 5 4 3 2 1 0 Reset value: 0xFF
Reserved LEVEL[2:0]
Description:
Address: 0x86
This register has the same function as
Type: R/W KAR_MCh0VOL for the left voice channel instead of
Reset value: 0x0 the left music channel.
Description: KAR_VCh1VOL
When the voice cancellation is enabled by the Voice channel 1 (R) volume
KAR_VCANCEL register, KAR_VVALUE specifies
the extent of the voice cancellation according to the 7 6 5 4 3 2 1 0
following table: Value
Address: 0x89
LEVEL[2:0] 0: cut-band filter with 40dB attenuation at
700Hz Type: R/W
1: cut-band filter with 35dB attenuation at Reset value: 0xFF
700Hz
2: cut-band filter with 32dB attenuation at
700Hz Description:
3: cut-band filter with 27dB attenuation at
700Hz This register has the same function as
4: cut-band filter with 23dB attenuation at KAR_MCh0VOL for the right voice channel instead of
700Hz the left music channel.
73/90
STA310
KAR_DUET Description:
Duet ON/OFF switch Bitfield Description
7 6 5 4 3 2 1 0 VOICEEFF Selecte the voice effects:
0: No effect is applied to the voice input
Reserved DUET 1: Echo is applied to the voice inputs,
tuned by registers KAR_VDELAY and
Address: 0x8A KAR_VBAL
2: Chorus is applied to the voice inputs,
Type: R/W tuned by registers KAR_VDELAY and
Reset value: 0 KAR_VBAL
3: Reverb is applied to the voice inputs,
tuned by register KAR_VDELAY.
Description: MIX Voice channel mixing:
The value in this register sets the duet function on or 0: No mix, voice is output on centre
off. When selected, the duet function is configured by channelt
1: Mix music and voice channels into music
register KAR_DUETTHRESH.
channel.
DUET 0: duet off and 1: duet on
KAR_VDELAY
Programmable delay/decay music effects
KAR_DUETTHRESH
7 6 5 4 3 2 1 0
Duet threshold control
Value
7 6 5 4 3 2 1 0
DUETTHRESHOLD[7:0] Address: 0x8D
Type: R/W
Address + 0x8B Reset value: 0x0
Type: R/W
Reset delay: 0 Description:
The value in this register specifies the delay used for
Description: voice input effects. The delay can be set in the range
When the Duet function is enabled by the from 0 to 2048/Fs seconds (where Fs is the sampling
KAR_DUET register, this register specifies the ampli- frequency in KHz).
tude of the voice line below which the voice is can- ‘desired time delay’ = ( 2048 / Fs) * ( Value / 256)
celled. If the amplitude of the voice line is below this which gives:
threshold, the recorded voice is played instead. The
value of DUETTHRESHOLD ranges from 0 to 255, KAR_VDELAY value = (Fs / 8) * ’desired time delay
full scale signal. For reverberation effects, this register gives the de-
cay factor,which can vary within the range 0 to 1.0.
KAR_VOICE
Selection of voice effects KAR_VBAL
Programmable mix for echo and chorus effects
7 6 5 4 3 2 1 0
7 6 5 4 3 2 1 0
Reserved MIX VOICEEFF[1:0
] BALANCE[7:0]
Address: 0x8C
Address: 0x8E
Type: R/W
Type: R/W
Reset delay: 0x0
Reset value: 0x3F
74/90
STA310
Description: KAR_MODE
This register sets the balance between the original Operating mode selection
sound and its delayed version for the echo and cho-
rus effects according to the formula. 7 6 5 4 3 2 1 0
echo (or chorus) output = original_sound * (1 - bal-
Reserved KAR_MODE[1:0]
ance) + delayed_sound * balance
where balance = Balance[7:0] / 255
Address + 0x91
where balance can vary in the range of 0 to 1. A bal-
ance limit of 0.5 is recommended. BALANCE[7:0] = Type: R/W
balance * 255. Reset value: 0x01
KAR_VMUTE Description:
Voice channel mute This register specifies the working mode of the
Karaoke module.
7 6 5 4 3 2 1 0
Reserved MUTE Bitfield Description
KAR_MODE 00: Karaoke processor in waiting mode.

Address: 0x8F [1:0] This is the default mode after a hardware
reset, total or partial software reset. This
Type: R/W mode is used to programme all the
Reset value: 0x00 registers at first initialization.
01: Karaoke processor running.

Description:
10: Partial Software reset. This resets the
This register mutes the voice channel: ’0’ means not internal DSP program but keep the
muted and ’1’ means muted. register configuration as it was before the
partial reset. When the partial reset is
finished, KAR_MODE[1:0] is automatically
KAR_PLAY set to ’01’.
Mute of voice and music
11: Total Software reset. The program is
reset and the registers values are
7 6 5 4 3 2 1 0
changed back to their reset default values.
Reserved PLAY
Address: 0x90 KAR_DIN_CTL

Type: R/W Control of voice channel
Reset value: 0x01 7 6 5 4 3 2 1 0
Reserv JUSTIF DELAY WS_P CLK_P WS[1:0] DINEN

Description: ed OL OL
This register mutes the voice and the music channels
simultaneously: PLAY = ’0’ means muted and ’1’ Address: 0x92
means playing. The registers and KAR_VMUTE Type: R/W
have priority for muting.
Reset Value: 0x00
Description:
This register specifies the input format for configuring
75/90
STA310
the handling of the second input. Description:

Bitfield Description This register sets the sampling frequency, Fs, of the
incoming PCM stream.
DINEN DIN enable: 0: disabled, 1: enabled
WS[1:0] PCM precision: 00: 16-bit mode, 01: 18-bit Fs (KHz) 48 44.1 32 - 96 88.2 64 - 24 22.05
mode, 10: 20-bit mode, 11: 24-bit mode Value 0 1 2 3 4 5 6 7 8 9
CLK_POL 0: Data and WS change on clockalling (decimal)
edge
1: Data and WS change on clockising
edge
WS_POL 0: Left data word = WS low, right data word Fs (KHz) 16 - 12 11.02 8 - 192 176.4 128 -
= WS high 5
1: Left data word = WS high, right data Value 10 11 12 13 14 15 16 17 18 19
word = WS low (decimal)
DELAY 0: First bit of data occurs on transition of
WS
1: First bit of data occurs with 1 clock cycle CANINPUT_MODE
delay (I2S compatible)
Selection of input data format
JUSTIF 0: Left padded, 1: Right padded
7 6 5 4 3 2 1 0
KAR_UPDATE SWAP MODE[6;0]
Change active Karaoke functions

Address: 0x95
7 6 5 4 3 2 1 0 Type: R/W
Reserved UPDATE Reset Value: 0x00
Address: 0x93 Description;

Type: R/W This register specifies the input format for configuring
Reset Value: 0 the handling of the second input.
Description: 0: 16 slots mode
1: 16 slots mode, LSB first
This register loads the new Karaoke configuration 2: 32 slots mode, left aligned
into the internal registers when UPDATE is set to ’1’. MODE 3: 32 slots mode, right aligned
When the bit is reset to ’0’ the system continues in the [6:0] 4: 32 slots mode, I2S mode
configuration last loaded. 5: 32 slots mode, sign extended
6: 32 slots mode, 8-bit data
7: 32 slots mode16-bit data
Channel swap: 0: Left channel first, 1: Right
SWAP
channel firs
9.21 Second serial input registers
SFREQ2 9.22 Linear PCM (DVD audio) registers
Sampling frequency of voice channel LPCMA_DOWNMIX
Downmix
7 6 5 4 3 2 1 0
7 6 5 4 3 2 1 0
Reserved Value
Reserved Value
Address: 0x94
Type: RO Address: 0x6F
Reset Value: 0 Type: R/W
Software Reset: NC
76/90
STA310
Hardware Reset: UND

Description:
Description;:. This register sets the phase coefficients for channels
mixing to Lmix. The input signal is inverted when
Value Description PH_xL = ’0’ and non-inverted when ’1’.
0 Downmix not applied
1 Force downmix 2/0 LPCMA_DM_COEFT_1
Downmix 2/0 is applied if the flags Downmix phase coefficients 1
down_mix_code_validity AND
2
stereo_playback_mode are both ’0’ in the 7 6 5 4 3 2 1 0
bitstream
PH_0R 0 PH_2R PH_3R PH_4R PH_5R Reserved
The notation, 2/0 represents 2 front speakers and no
surround speakers. Address: 0x98
Type: R/W
LPCMA_FORCE_DWS Software Reset: NC
Downsampling 192 to 96KHz Hardware Reset: UND
7 6 5 4 3 2 1 0
Description:
Reserved Value
This register sets the phase coefficients for channels
Address : 0x70 mixing to Rmix. The input signal is inverted when
PH_xR = ’0’ and non-inverted when ’1’.
Type: R/W
Software Reset: NC
LPCMA_DM_COEFT_2
Hardware Reset: UND
Downmix gain coefficients 2
Description: 7 6 5 4 3 2 1 0
This register selects whether downsampling is used COEF_0L
for input streams requiring sampling frequencies of
192KHz or 176.4KHz. When ’automatic’ is selected, Address: 0x99
register is automatically updated to correspond to the
new output frequency. Type: R/W
Software Reset: NC
Hardware Reset: UND
Value 00: Automatic (if Fs = 192KHz or 176.4KHz)
01: Automatic (if Fs = 192KHz or 176.4KHz)
10: No downsampling Description:
This register sets the mixing gain for Lf to Lmix. See
the note after register
LPCMA_DM_COEFT_0
Downmix phase coefficients 0 LPCMA_DM_COEFT_3
7 6 5 4 3 2 1 0 Downmix gain coefficients 3
0 PH_1L PH_2L PH_3L PH_4L PH_5L Reserved
7 6 5 4 3 2 1 0
COEF_0R
Address : 0x97
Type: R/W Address : 0x9A
Software Reset: NC Type: R/W
Hardware Reset: UND Software Reset: NC
77/90
STA310

Hardware Reset: UND
Description:
This register sets the mixing gain for Lf to Rmix. See Description:
the note after register This register sets the mixing gain for C to Lmix. See
the note after register .
LPCMA_DM_COEFT_4
Downmix gain coefficients 4 LPCMA_DM_COEFT_7
7 6 5 4 3 2 1 0 Downmix gain coefficients 7
COEF_1L
7 6 5 4 3 2 1 0
Address: 0x9B COEF_2R
Type: R/W
Address: 0x9E
Software Reset: NC
Type: R/W
Hardware Reset: UND
Software Reset: NC
Hardware Reset: UND
Description:
This register sets the mixing gain for Rf to Lmix. See
the note after register Description:
<BlueHT>LPCMA_DM_COEFT_13. This register sets the mixing gain for C to Rmix. See
the note after register
LPCMA_DM_COEFT_5
Downmix gain coefficients 5 LPCMA_DM_COEFT_8
7 6 5 4 3 2 1 0
7 6 5 4 3 2 1 0
COEF_1R
COEF_3L
Address: 0x9C
Type: R/W Address: 0x9F
Software Reset: NC Type: R/W
Hardware Reset: UND
Description:
This register sets the mixing gain for Rf to Rmix. Description:
See the note after register This register sets the mixing gain for Ls / S to Lmix.
See the note after register
<BlueHT> LPCMA_DM_COEFT_13.
LPCMA_DM_COEFT_9
LPCMA_DM_COEFT_6 Downmix gain coefficients 9
7 6 5 4 3 2 1 0
7 6 5 4 3 2 1 0
COEF_3R
COEF_2L
Address : 0xA0
Address: 0x9D Type: R/W
Type: R/W Software Reset: NC
78/90
STA310
Hardware Reset: UND LPCMA_DM_COEFT_13

Description:
7 6 5 4 3 2 1 0
This register sets the mixing gain for Ls / S to Rmix.
See the note after register COEF_5R
Address: 0xA4
LPCMA_DM_COEFT_10
Type: R/W
Software Reset: NC
COEF_4L
Description:
Address : 0xA1
This register sets the mixing gain for LFE to Rmix
Type: R/W
Note: For DVD audio, the real coefficient value, alpha[x], applied to
Software Reset: NC channel x is calculated with the following formulae:
Hardware Reset: UND
alpha[x] = 2-(COEF_xL/30) 0<COEF_xL 199
Description: alpha[x] = 2-((COEF_xL - 100)/30) 200<COEF_xL 254
alpha[x] = 0 COEF_xL 255
This register sets the mixing gain for Rs to Lmix. See
the note after register .
LPCMA_STATUS0
LPCMA_DM_COEFT_11 Linear PCM (DVD audio) status register
7 6 5 4 3 2 1 0
7 6 5 4 3 2 1 0 EMPH Reser STERE DWNMX DOWN_MIX_CODE
ASIS ved O_PB _VALID [3:0]
COEF_4R
Address: 0x76
Address: 0xA2
Type: RO
Type: R/W
Software Reset: NC
Software Reset: NC
Hardware Reset: UND
Hardware Reset: UND
Description:
Description: ed from the stream.
This register sets the mixing gain for Rs to Rmix. See
the note after register Bitfield Description
LPCMA_DM_COEFT_12 DOWN_MIX_CODE Identifying code
Downmix gain coefficients 12 DWNMX_VALID DOWN_MIX_CODE valid
7 6 5 4 3 2 1 0 STEREO_PB Stereo playback mode
COEF_5L
EMPHASIS Emphasis flag
Address: 0xA3
Type: R/W
LPCMA_STATUS1
Software Reset: NC
Linear PCM (DVD audio) status register
Hardware Reset: UND
7 6 5 4 3 2 1 0
QUANTIZATION_WORD_L QUANTIZATION_WORD_L
This register sets the mixing gain for LFE to Lmix.
ENGTH_1[3:0] ENGTH_2[3:0]
See the note after register .
79/90
STA310
Address: 0x77 Description:

Type: RO This register contains bit stream information extract-
Software Reset: NC ed from the stream.
Hardware Reset: UND Bitfield Description
MULTI_CHANNEL_TYPE[3:0]
Description:
ed from the stream. LPCMA_STATUS4
Bitfield Description Linear PCM (DVD audio) status register
QUANTIZATION_WORD Quantization word length for 7 6 5 4 3 2 1 0
_LENGTH_2[3:0] group 2
BIT_SHIFT_OF_CHANNEL CHANNEL_ASSIGNMENT
QUANTIZATION_WORD Quantization word length for _GR2[3:0] [3:0]
_LENGTH_1[3:0] group 1
Address: 0x7A
Type: RO
LPCMA_STATUS2
Software Reset: NC
Linear PCM (DVD audio) status register
Hardware Reset: UND
7 6 5 4 3 2 1 0
SAMPLING_FREQUENCY SAMPLING_FREQUENCY_
_1[3:0] 2[3:0]
Description:
Address: 0x78 ed from the stream.
Software Reset: NC
Hardware Reset: UND
CHANNEL_ASSIGNMENT[3:0]
Description: BIT_SHIFT_OF_CHANNEL_GR2[3:0]
ed from the stream.
Bitfield Description LPCMA_STATUS5
SAMPLING_FREQU Sampling frequency for Linear PCM (DVD audio) status register
ENCY_2[3:0] group 2
SAMPLING_FREQU Sampling frequency for 7 6 5 4 3 2 1 0
ENCY_1[3:0] group 1 DYNAMIC_RANGE_CONTROL[7:0]
Address: 0x7B
LPCMA_STATUS3 Type: RO
Linear PCM (DVD audio) status register Software Reset: NC
Reserved MULTI_CHANNEL_TYPE[3:0]
Description:
Address: 0x79 This register contains information, extracted from the
stream, for the dynamic range control.
Type: RO
Software Reset: NC
Hardware Reset: UND
80/90
STA310
9.23 Linear PCM (DVD video and PCM) LPCMV_DM_COEFT_0

registers Downmix phase coefficients 0
LPCMV_DOWNMIX
7 6 5 4 3 2 1 0
Downmix
PH_0L PH_1L PH_2L PH_3L PH_4L PH_5L Reserved
7 6 5 4 3 2 1 0
Address: 0x97
Reserved Value
Type: R/W
Address : 0x6F Software Reset:NC
Type: R/W Hardware Reset: UND
Software Reset: NC
Hardware Reset: UND Description:
This register sets the phase coefficients for channels
Description;. mixing to Lmix. The input signal is inverted when
PH_xL = ’0’ and non-inverted when ’1’.
Value Description
0 Downmix not applied LPCMV_DM_COEFT_1

Downmix phase coefficients 1
1 Force downmix 2/0
7 6 5 4 3 2 1 0
The notation, 2/0 represents 2 front speakers and no
surround speakers. PH_0R PH_1R PH_2R PH_3R PH_4R PH_5R Reserved
Address: x98
LPCMV_FORCE_DWS Type: R/W
Downsampling 96 to 48KHz Software Reset:NC
Reserved Value
Description:
Address: 0x70 This register sets the phase coefficients for channels
Type: R/W mixing to Rmix. The input signal is inverted when
PH_xR = ’0’ and non-inverted when ’1’.
Software Reset: NC
Hardware Reset: UND
LPCMV_DM_COEFT_2
Description:
For details see register 0x99
This register selects whether downsampling is used
for input streams requiring a sampling frequency of
96KHz.
When ’automatic’ is selected, register is automatical- LPCMV_DM_COEFT_3
ly updated to correspond to the new output frequen- Downmix gain coefficients 3
cy. For details see register 0x9A
Value 00: Automatic (if Fs = 96KHz)

01: Automatic (if Fs = 96KHz) LPCMV_DM_COEFT_4
10: No downsampling Downmix gain coefficients 4
For details see register 0x9B
81/90
STA310
LPCMV_DM_COEFT_5 Note: For DVD video & PCM, the real coefficient value, alpha[x], ap-
Downmix gain coefficients 5 plied to channel x is calculated with the following formulae:
alpha[x] = 2-(x+(y/30)) 0<Y,29 0<X,7
For details see register 0x9C
COEF_xL = register bits [b7,b6,b5,b4,b3,b2,b1,b0]
X = register bits[b7,b6,b5]
Y = register bits[b4,b3,b2,b1,b0
PCMV_DM_COEFT_6
Downmix gain coefficients 6 LPCMV_STATUS0
For details see register 0x9D
7 6 5 4 3 2 1 0
EMPH_ MUTE_ Reserved FRAME_NUM
FLAG FLAG
LPCMV_DM_COEFT_7
Downmix gain coefficients 7 Address: 0x76
For details see register 0x9E Type: R/W
Reset Value: UND
LPCMV_DM_COEFT_8 Description:
For details see register 0x9F
FRAME_NU frame number of the first access unit in
M the group of audio frames
Reserved Set to 0
LPCMV_DM_COEFT_9 MUTE_FLAG 0: mute off, 1: mute on
Downmix gain coefficients 9 EMPH_FLAG Emphasis status after the first access
For details see register 0xA0 unit: 0: emphasis off; 1: emphasis on
LPCMV_STATUS1
PCMV_DM_COEFT_10
Downmix gain coefficients 10 7 6 5 4 3 2 1 0
For details see register 0xA1 Word_Length Samp_Freq Reserved Channels
Address :0x77
LPCMV_DM_COEFT_11 Type: R/W
Downmix gain coefficients 11 Reset Value: UND
For details see register 0xA2
Description:
LPCMV_DM_COEFT_12 Channels Number of audio channels:000=1
Downmix gain coefficients 12 channel (mono), 001=2 channels
(stereo), 010=3 channels,
For details see register 0xA3 011=4 channels, 100=5
channels,101=6 channels, 110=7
LPCMV_DM_COEFT_13 channels, 111=8 channels
Downmix gain coefficients 13 Reserved Set to 0
For details see register 0xA4 Samp_Freq Sampling frequency: 00=48kHz,

01=96kHz, 10=reserved, 11=reserved
Word_Length Audio sample length: 00=16 bits,
01=20 bits, 10=24 bits, 11=reserved
82/90
STA310
PCMV_STATUS2 Description:
7 6 5 4 3 2 1 0 Value Description
Dyn_Range_Control Value (decimal) This register configures the multi
channel structure for the output
Address : 0x78 channels:
Type: RW 0: Stereo
1: Multi channels
Reset Value: UND
Description:
This register sets the dynamic range compression
from the first access unit. For the hexadecimal value 9.24 MLP registers
0x80, dynamic range control is not set. For all other MLP_CRC
values, the dynamic range control is CRC check
(24.082 - 6.0206 * X - 0.2007 * Y)dB, where
X = dynamic_range_control[7..5] and 7 6 5 4 3 2 1 0
Y = dynamic_range_control[4..0].
reserved SU_P MA_S MS_C RH_C
LPCMV_CH_ASSIGN Address: 0x6C

Channel assignment Type: R/W
Software Reset: NC
7 6 5 4 3 2 1 0
Hardware Reset: UND
Reserved Value
Address: 0xA8 Description:

Type: R/W This register controls the four different CRCs in MLP.
Software Reset: NC If the check is false, an error is returned (error num-
Hardware Reset: UND bers 80-83 in register and the outputs of all 6 chan-
nels are muted
Description: Bitfield Description
RH_C 1: Check of ’Restart_Header_CRC’ enable

Value Description
MS_C 1: Check of ’Major_Sync_CRC’ enable
Value This register configures the audio channels:
(decimal) See "DVD Specifications for Read-Only MA_C 1: Check of ’Max_Shift’ enable
Disc", Part 4 AUDIO SPECIFICATIONS,
Version 1.0, March 1999, Table C.1-2. SU_P 1: Check of ’Substream_Parity’ enable
PCMV_MULTI_CHS MLP_DOWNMIX
Multi channels Downmix
7 6 5 4 3 2 1 0 7 6 5 4 3 2 1 0
Reserved Value DWNMIX[7:0]
Address: 0xA9 Address: 0x6F

Type: R/W Type: R/W
Hardware Reset: UND Hardware Reset: UND
83/90
STA310
Description: new output frequency.

This register controls the MLP downmix. Bitfield Description
Bitfield Description Value 00: Automatic (if Fs = 192KHz or 176.4KHz)
01: Automatic (if Fs = 192KHz or 176.4KHz)
DWNMIX 0x00: 2/0 (L / R)
10: No downsampling
[7:0] 0x01: 2/0(a) (Lo / Ro) (according to
bitstream0)
0x02: 3/0 (L, R, C)
0x03: 2/1 (L, R, S) MLP_LFE
0x04: 3/1 (L, C, R, S) Decode LFE
0x05: 2/2 (L, R, Ls, Rs)
0x06: 3/3 (L, C, R, Ls, Rs)
7 6 5 4 3 2 1 0
For all other values there is no downmix.
LFE[7:0]
(a) Downmix 1/0 (one channel only) is
forbidden in DVD audio
Address: 0x68
Type: R/W
MLP_DRC Software Reset: NC
Dynamic range control Hardware Reset: UND
7 6 5 4 3 2 1 0
Description:
DRC[7:0]
When this register = 0x00, LFE is not decoded and
Address : 0x6A when 0x01, LFE is decoded.
Type: R/W
Software Reset: NC MLP_STATUS0
Hardware Reset:: UND MLP status 0 register
7 6 5 4 3 2 1 0
Description: Reserved FS_CODE[4:0]
When this register = 0x00, the dynamic range control
is disabled. When 0x01, the dynamic range control is Address: 0x76
enabled and the DRC information in the MLP stream Type: R/W
is used.
Software Reset: NC
Hardware Reset: UND
MLP_FORCE_DWS
Downsampling 192 to 96kHz or 176.4 to 88.2kHz
Type:
7 6 5 4 3 2 1 0 This status register contains the sampling frequency
codes..
Reserved Value
Address: 0x70
Type: R/W FS_COD This list gives the codes and the
E[4:0] corresponding sampling frequency.
Software Reset: NC 0x09: 44.1KHz
Hardware Reset: UND 0x0A: 48KHz
0x0D: 88.2KHz
0x0E: 96KHz
Description: 0x11: 176.4KHz
0x12: 192KHz
This register selects whether downsampling is used 0x1F: Undefined
for input streams requiring sampling frequencies of The remaining codes are reserved.
192KHz or 176.4KHz. When ’automatic’ is selected,
register is automatically updated to correspond to the
84/90
STA310
MLP_STATUS1 Description:
MLP status 1 register This status register contains the sub-stream informa-
tion codes..
7 6 5 4 3 2 1 0
Reserved CH_ASSIGN[4:0] SUBSTR_CODE 2-channel decoder:
[3:0] bit0 = ’1’: sub-stream 0 is decoded
Address : 0x77 bit1 = ’1’: a simplified decoder can
be used for sub-stream 0
Type: RO 6-channel decoder:
Software Reset: NC bit2 = ’1’: sub-stream 0 is decoded
bit3 = ’1’: sub-stream 1 is decoded
Hardware Reset: UND
Description:
This status register contains the channel assignment. 9.25 De-emphasis register
Bitfield Description DEEMPH
CH_ASS This gives the channel assignment: De-emphasis
IGN[4:0] See "DVD Specifications for Read-Only
Disc", Part 4 AUDIO SPECIFICATIONS, 7 6 5 4 3 2 1 0
Version 1.0, March 1999, Table C.1-1.
reserved D[1:0]
MLP_STATUS2 Address: 0xB5

MLP status 2 register Type: R/WS
Software Reset: NC
7 6 5 4 3 2 1 0
Hardware Reset: UND
Reserved NSUBSTR[3:0]
Address: 0x78 Description:

Type: RO This register is used in MPEG, DVD_LPCM or CDDA
modes; it is not supported in Dolby Digital.
Software Reset: NC
In MPEG and DVD_LPCM modes, its register value
Hardware Reset: UND is extracted from the bitstream. When the emphasis
status changes (by setting bit DEM of the register),
Description: an interrupt is generated. In CDDA mode, the register
value must be updated by the application.
This status register contains the number of sub-
streams present in the audio frame. The de-emphasis filter specified here is applied only
if bit DEM of the register is set.
D[1:0] 00: none, 01: 50/15µs, 10: reserved, 11:
MLP_STATUS3
CCITT J.17
MLP status 3 register
7 6 5 4 3 2 1 0
Reserved SUBSTR_CODE[3:0]
9.26 Auxilliary outputs registers
Address:: 0x79 VCR_MIX
Type: RO VCR outputs
Software Reset: NC
7 6 5 4 3 2 1 0
Hardware Reset: UND
reserved STEREO PRL reserved COPY 3D_VCR
85/90
STA310
Address: 0xAE VCR_RDLY

Type: R/WS?? VCR right channel delay
Software Reset: NC 7 6 5 4 3 2 1 0
Hardware Reset: 0 RIGHT_VCR_DELAY
Description: Address: 0xAF

Type: R/WS??
Software Reset: NC
3D_VCR This bit selects "3-D sound" on the VCR Hardware Reset: 0
channels using SRS processing
(depending on the PDEC registers and ):
0: Standard sound (disable "3-D sound"), Description:
1: Enable "3-D sound".
This register contains the VCR right channel delay
COPY This bit is used to copy "Left/Right" value. The values for LEFT_VCR_DELAY and
channels to "VCR" channels: RIGHT_VCR_DELAY are taken into account only
0: no copy, when register.bit .DLY = ’1’.
1: copy enable.
PRL This bit enables a "ProLogic downmix" on

the "VCR" channels:
0: Disable,
1: Enable.
9.27 Miscellaneous
BREAKPOINT
STEREO This bit enables a "2/0 downmix" on the
"VCR" channels:
To be defined
0: Disable,
7 6 5 4 3 2 1 0
1: Enable.
Reserved
Note: 1. To have both "3-D sound" on the "VCR" and "Left/Right"
channels, the setup is: Address + 0x2B
VCR_MIX = 0x02 and PDEC = 0x40 for SRS process-
ing,
Type: R/W
Software Reset: NC
Hardware Reset: 0
VCR_LDLY
VCR left channel delay Description:
7 6 5 4 3 2 1 0 This register must be set to 0x08.
LEFT_VCR_DELAY
CLOCKCMD
Address: 0xAF To be defined
Type: R/WS??
7 6 5 4 3 2 1 0
Software Reset: NC
Reserved
Hardware Reset: 0
Address: 0x3A
Description: Type: R/W
This register contains the VCR left channel delay val- Software Reset: NC
ue. See note after next register description.
Hardware Reset: 0
Description:
This register must be set to 0x00.
86/90
STA310
INIT_RAM
RAM initialization
7 6 5 4 3 2 1 0
Reserved RAM_INIT
Address : 0xFF
Type: RO
Software Reset: 1
Hardware Reset: 0
Description:
The register is used to signal when the STA310 has finished to boot.
After a soft reset or a hardware reset, or a hardware reset, the host processor must wait until INIT_RAM hold
the value “1”.
the host can then start to configure the STA310 according to its application
APPENDIX A OVERVIEW OF THE CHIP

This STA310 is based on a very high performances low power general purpose DSP core, MMDSP+, and a set
of dedicated peripherals. Internal audio and system PLL allows to configure the chip for a wide range of audio
frequencies and DSP processing power (1 to 100 Mips).
A.1 Architectural Block Diagram
8 channels Pcm output

Sys PLL
CNA interf 4x2
slave Audio PLL I2S/Sony

768words out
I2C
Ram
256x8
S/Pdiff out
host S/Pdiff
Control
Host registers out

Parallel 32 Kinstr
Rom
Emulation Ybus
ST ASDSP
Rs232 line
RS232 18Kx24 256x24 MMDSP+ 15Kx24 4Kx16 1WS Mem DMA

in/out Rom Ram core Ram Ram 64Kx8 pcm
Xbus
S/Pdif
000 DMA x 3
Two
000
in CRC
checkers
Data in
I2S Packet Audio Frame Buffer Dbit/Nbit

in Parser Parser 32k x 8
2048 bits
Parallel input FIFO I2Sin
Data in
Second Serial Input
87/90
STA310
A.2 Description of the architecture

The MMDSP+ DSP core can access 5 banks of RAM/ROM memories:
- the 32K instruction ROM,
- the 768 words instruction development RAM,
- X_memory 19K x 24 RAM,
- Y_memory 18 K x 24 ROM,
- Y_memory 1K x 24 RAM.
The DSP core can also access some dedicated and general purpose peripherals. These peripherals (called
MMIO peripherals) are mapped as memory locations of the X memory space of the MMDSP+ DSP core. On top
of the front-end dedicated ones, the list of the peripherals is the following:
- Four PCM out I2S/Sony (16,18,20,24 bits) serial output interfaces are provided to connect, typically,
to external DACs. This interface and the Audio PLL provide the Oversampling clocks and the serial
clocks necessary to interface the DACs .This interface provides up to 8 independent audio channels.
A “DMA PCM” MMIO block makes the link between the X data memory of the DSP core (which can
store the audio samples) and the I2S/Sony serial interfaces. This MMIO block is a DMA (Direct Mem-
ory Access) and handles automatically the transfer of data by blocks. This peripheral implements also
an hardware mechanism to support delayed channels. Each channel can de delayed (resolution 1
sample) by a programmable number of data samples. This function is totally transparent to the user.
- A 256 x 8 address space is shared between the MMDSP+ core (as MMIO peripheral) and the external
world of the STA310 through the I2C Slave interface or the Host parallel interface. This area is divided
mainly in 2 parts:
a 192 x 8 general purpose RAM area,
a 64 x (1 to 8 bits) area of specific registers.
- The two PLLs (Audio PLL and System PLL) can be controlled by the DSP itself (thru the MMIO bus)
or by the external world of the STA310 (thru the I2C Slave I/F or the Host parallel I/F).
88/90
STA310
mm inch
DIM. OUTLINE AND
MIN. TYP. MAX. MIN. TYP. MAX. MECHANICAL DATA
A 1.60 0.063
A1 0.05 0.15 0.002 0.006
A2 1.35 1.40 1.45 0.053 0.055 0.057
B 0.22 0.32 0.38 0.009 0.013 0.015
C 0.09 0.20 0.003 0.008
D 16.00 0.630
D1 14.00 0.551
D3 12.35 0.295
e 0.65 0.0256
E 16.00 0.630
E1 14.00 0.551
E3 12.35 0.486
L 0.45 0.60 0.75 0.018 0.024 0.030
L1 1.00 0.0393 TQFP80

K 3.5°(min.), 7°(max.) (14x14x1.40mm)
D
A
D1
A2
D3
A1
60 41
61 40
0.10mm
.004
Seating Plane
E3 E1 E
Gage plane
PIN 1
0.25mm
IDENTIFICATION
80 21
1 20
K C
TQFP80L
L
L1
89/90
STA310
Information furnished is believed to be accurate and reliable. However, STMicroelectronics assumes no responsibility for the consequences
of use of such information nor for any infringement of patents or other rights of third parties which may result from its use. No license is granted
by implication or otherwise under any patent or patent rights of STMicroelectronics. Specifications mentioned in this publication are subject
to change without notice. This publication supersedes and replaces all information previously supplied. STMicroelectronics products are not
authorized for use as critical components in life support devices or systems without express written approval of STMicroelectronics.
The ST logo is a registered trademark of STMicroelectronics

 2003 STMicroelectronics - All Rights Reserved
STMicroelectronics GROUP OF COMPANIES
Australia - Brazil - Canada - China - Finland - France - Germany - Hong Kong - India - Israel - Italy - Japan -Malaysia - Malta - Morocco -
Singapore - Spain - Sweden - Switzerland - United Kingdom - United States.
http://www.st.com
90/90

Chip Dolby Decoder

Uploaded by

Copyright:

Available Formats

Chip Dolby Decoder

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Chip Dolby Decoder

Uploaded by

Copyright:

Available Formats

STA310

6+2-CH. MULTISTANDARD AUDIO DECODER

■ SPDIF and IEC-61937 input interface. ■ Multimedia PC.

■ SPDIF and IEC-61937 output interface.

June 2003 1/90

2 STA310 AUDIO DECODER PIN DESCRIPTION

Pin Number Name Type Function

48 IRQB O (1) Interrupt Signal (level), active low

Parallel Control Interface

78 - 79 - 80 - 1 D0 - D1 - D2 - D3 I/O Host Data

22 R/W I Read/Write Selection: read access when high, write access

DATA INPUT INTERFACE

First Serial Data Interface (I2S)

41 SIN I Serial Input Data

40 LRCLKIN I Word Clock for the Input

42 REQ O Handshake for the Data Transfer, aconfigurable by the

Second Serial Data Interface (I2S)

60 SIN2 I Serial Input Data

61 LRCLKIN2 I Word Clock for the Input

63 REQ2 O Handshake for the Data Transfer, active low

DATA OUTPUT INTERFACES

67 SCLK O Bit Clock for the DAC

2 STA310 AUDIO DECODER PIN DESCRIPTION (continued)

Pin Number Name Type Function

68 LRCLK O Word Clock for the DAC

72 PCM_OUT0 O Data from a Prologic downmix (VCR_L/VCR_R)

73 PCM_OUT1 O Data for the first DAC (Left/Right)

76 PCM_OUT2 O Data for the second DAC (Centre/Sub)

77 PCM_OUT3 O Data for the third DAC (LeftSur/RightSur)

IEC958 Interface (S/PDIF) - One Output Port., One Input Ports.

58 I958OUT O S/PDIF Signal

25 SPDP I First differential input of S/P DIF port

24 SPDN I Second differential input of S/P DIF port

26 SPDF I External Filter

28 VDDA I Analog VDD for S/P DIF Input port

29 GNDA I Analog GND for S/P DIF Input port

PCM Related Information

Audio Video Synchronization

59 PTSB O Indicates that a PTS has been detected, active low.

31 CLK I Master Clock Input Signal.

36 RESET I(2) Reset signal input, active low.

52 TESTB I(2) Reserved pin: to be connected to VDD

49 SMODE I Reserved pin : to be connected to GND

64 CLKOUT O System clock output with programmable division ratio

27 PLLAF I External Filter For Audio PLL.

2 STA310 AUDIO DECODER PIN DESCRIPTION (continued)

Pin Number Name Type Function

30 PLLSF I External Filter For System PLL.

Power and Ground

5 - 11 - 23 - 33 - GND GND Ground

Notes (1) Open Drain

PIN CONNECTION (Top view)

ABSOLUTE MAXIMUM RATINGS

Vil Low level input voltage 0.8 V 1

Vih High level input voltage 2.0 V 1

Vilhyst Low level threshold input falling 0.8 1.35 V 1

Vihhyst High level threshold input rising 1.3 2.0 V 1

Vhyst Schmitt trigger hysteresis 0.3 0.8 V 1

ELECTRICAL CHARACTERISTICS (continued)

Ipu Pull-up current Vi = 0V -66 0.8 µA 1