The Physics and Psycho-

Acoustics of Surround Recording

David Griesinger
Lexicon
[email protected]
www.world.std.com/~griesngr
 Major Goals

• To show how physics and psycho acoustics

combine to produce absolute standards of quality
in sound recording.
– And for live sound in opera houses and concert halls!
• To show how to know when a recording meet
these standards, and why.
• To show how to create a high quality recording in
practice.
• To play as many musical examples as possible!
 Human sound perception – Separation of the
sound field into foreground streams.
• Engineers are entranced with frequency response, distortion, and time
delay.
– But human perception works differently.
– Human brains evolved to understand speech, not to measure sound
systems.

Third-octave filtered speech.

Blue 500Hz. Red 800Hz

Speech consists of a series of

foreground sound events
separated by periods of
relative silence, in which the
background sound can be
heard.
 The primary function of human hearing
is stream formation
– Foreground sound events (phones or notes) must be separated
from a total sound field containing both foreground and
background sounds (reverberation, noise).
• Foreground events are then assembled into streams of common
direction and/or timbre.
• A set of events from a single source becomes a sound stream, or a
sound object. A stream consists of many sound events.
– Meaning is assigned to the stream through the higher level functions,
including phoneme recognition and the combination of phonemes into
words.
• Stream formation is essential for understanding speech
– When the separation of sound streams is easy, intelligibility is high
• Separation is degraded by noise and reverberation.
• This degradation can be measured by computer analysis of binaural
speech recordings.
Analysis of binaural speech
Reverb
forward

Reverb
backward

Analysis into 1/3 octave bands,

followed by envelope
detection.
Green = envelope
Yellow = edge detection
 Analysis of binaural speech
• We can then plot the syllable onsets as a function of
frequency and time, and count them.

Reverberation forward Reverberation backwards

Note many syllables are detected (~30) Notice hardly ANY are detected (~2)
RASTI will give an identical value for both cases!!
 We also perceive distance and space

• We perceive fluctuations in the level during a sound event

and 50 to 150ms after as a sense of distance and as a sense
of space.
– These fluctuations are usually caused by interference from
reflected energy.
• If the reflections are spatially diffuse (from all directions)
the fluctuations will be different in each ear.
– Fluctuations that occur during the sound event and within 50ms
after the end of the event produce both a sense of distance and the
perception of a space around the source.
• This is Early Spatial Impression (ESI)
• The listener is outside the space – and the sound is not enveloping
• But the sense of distance is natural and pleasant.
– Spatially diffuse reflections later than 50ms after the direct sound
produce a sense of space around the listener.
• This can be perceived as envelopment. (Umgebung)
 Distance Perception

• Reflections during the sound event and up to 150ms after

it ends create the perception of distance
• But there is a price to pay:
– Reflections from 10-50ms do not impair intelligibility.
• The fluctuations they produce are perceived as an acoustic “halo” or
“air”around the original sound stream. (ESI)
– Reflections from 50-150ms contribute to the perception of distance
– but they degrade both timbre and intelligibility, producing the
perception of sonic MUD.
• We will have many examples of mud in this talk!
 Example of reflections in the 50-150ms range

Balloon burst in an
opera house.
Forestage to stalls
row 10.
Note the HUGE burst
of energy about 50ms
after the direct sound.
The 1000Hz 0ctave
band shows the
combined reflections
to be 6dB stronger
than the direct sound.

The sound clip

shows the result of
this impulse
response on speech.

The result (in this case) is a decrease in intelligibility and an increase in distance
 Human Perception – the background sound stream

• We also perceive the background sound in the spaces

between the individual sound.
• The background stream is perceived as continuous, even though it
may be rapidly fluctuating.
• The background stream is perceived at absolute level, not as a ratio to
the foreground sound.
• Perception of background is inhibited for 50ms after the end of a
sound event, and reaches full sensitivity only after 150ms.
 Example of foreground/background
perception (as a cooledit mix)
Series of tone
bursts (with a
slight vibrato)
increasing in
level by 6dB
Reverberation
at constant
level

Mix with
direct
increasing 6dB

Result: backgound tone seems continuous and at constant level

 Example of background loudness as
a function of Reverberation Time
Tone bursts at
constant level,
mixed with
reverberation
switching from 0.7s
RT to 2.0s RT, and
reducing in level
~8dB
Output – perceived
background is
constant! (But the
first half is
perceived as farther
away!)
Note the reverb level in the mix is the same at 150ms and greater. One gets the same
results with speech.
Summary: Perceptions relating to stream
separation
• First is the creation of the foreground stream itself. The major
perception is intelligibility
• Second is the formation of the background sound stream from sounds
which occur mostly 150ms after the direct sound ends. The perception
is reverberance
• Third is the perception of Early Spatial Impression (ESI) from
reflections arriving 10-15ms after the end of the direct sound. The
perception is the presence of distance and acoustic space.
• Fourth is the timbre alteration and reduction of intelligibility due to
reflections from 50 to 150ms after the end of the direct sound event.
The perception is MUD and distance.

• Intelligibility, Reverberance, distance, and mud are of MAJOR

importance in sound recording.
• They are also of HIGHEST importance in Opera Houses.
Binaural Examples in Opera Houses
It is very difficult to study opera acoustics, as the
sound changes drastically depending on:
1. the set design,
2. the position of the singers (actors),
3. the presence of the audience, and
4. the presence of the orchestra.
Binaural recordings made during performances
give us the only clues.
Here is a sound bite from a famous German opera
house: Note the excessive distance of
the singers, and the low intelligibility
And here is an example from another famous
German opera house: Note the increase
in intelligibility and the improvement in
dramatic connection between the singer and
the audience.
 Synthetic Opera House Study
• We can use MC12 Logic 7 to separate the orchestra from the singers on
commercial recordings, and test different theories of balance and reverberation.
• From Elektra – Barenboim. Balance in original is OK by Barenboim.
Original
Orchestra Left&Right

Vocals

Downmix - No reverb
on the singers

Reverb from orchestra

Reveb from singers

Downmix with reverb
on the singers.
 Localization
• Localization is related to stream formation. It depends
strongly on the beginning of sound events.
– IF the rise-time of the sound event is more rapid than the rise-time
of the reverberation
– Then during the rise time the IID (Interaural Intensity Difference)
and the ITD (Interaural Time Difference) are unaffected by
reflections.
• We can detect the direction of the sound source during this brief
interval.
• Once detected, the brain HOLDS the detected direction during the
reverberant part of the sound.
• And gives up the assigned direction very reluctantly.
– The conversion between IID and ITD and the perceived direction
is simple in natural hearing, but complex (and unnatural) when
sound is panned between two loudspeakers.
• Sound panning only works because localization detection is both
robust and resistant to change.
• A sound panned between two loudspeakers is profoundly unnatural.
 Detection of lateral direction through
Interaural Cross Correlation (IACC)
Start with
binaurally recorded
speech from an
opera house,
approximately 10
meters from the
live source.

We can decompose
the waveform into
1/3 octave bands
and look at level
and IACC as a
function of
frequency and
time.
Level ( x = time in ms y=1/3 octave bands 640Hz to 4kHz) IACC
Notice that there is NO information in the IACC below 1000Hz!
Position determination by IACC
We can make a histogram of
the time offset between the
ears during periods of high
IACC.
For the segment of natural
speech in the previous slide, it
is clear that localization is
possible – but somewhat
difficult.
 Position determination by IACC 2

Level displayed in 1/3 octave bands (640Hz to 4kHz) IACC in 1/3 octave bands

We can duplicate the sound of the previous example by adding reverberation to dry
speech, and giving it a 5 sample time offset to localize it to the right.
As can be seen in the picture, the direct sound is stronger in the simulation than in the
original, and the IACCs - plotted as 10*log10(1-(1/IACC)) - are stronger.
 Position determination by IACC 3

Histogram of the time

offset in samples for each
of the IACC peaks
detected, using the
synthetically constructed
speech signal in slide 2.

Not surprisingly, due to the higher direct sound level and the artificially
stable source the lateral direction of the synthetic example is extremely clear
and sharply defined.
 The physics of two-channel panning

The pressure at each ear is the sum of the direct sound pressure from one speaker
and the diffracted sound pressure from the other.
These two signals interfere with each other, producing a highly frequency
dependent signal.
 Consequences of panning physics
• A two channel pan is entirely different from the
localization process in natural hearing.
– Localization always depends on the interaural time delay (ITD)
and the interaural intensity difference (IID).
– In natural hearing the ITD and IID vary due to head shadowing
alone.
• Between 500Hz and 1500Hz the ear relies increasingly on IID rather
than ITD, and the precise phase of the ear signals becomes inaudible.

– In a two channel pan, ITD and IID vary due to INTERFERENCE.

• The interference is entirely PHYSICAL. It occurs in the air at the
entrance to the ear canals, and it happens at all frequencies, even HF.
• The interference can be measured, and it can be calculated.
– The result of the interference is that the ITD and IID become
highly frequency dependent.
• For broadband sources the brain must make a “best guess” about the
true azimuth of the source.
• The localization of narrow band souces can be bizarre.
 The frequency transmission of the pinnae and
middle ear

From: B. C. J. Moore, B. R.
Glasberg and T. Baer, “A model
for the prediction of thresholds,
loudness and partial loudness,” J.
Audio Eng. Soc., vol. 45, pp.
224-240 (1997).

The intensity of nerve firings is concentrated in the frequency range of human

speech signals, about 700Hz to 4kHz. With a broad-band source, the ITD and IID
at these frequencies will dominate the apparent direction.
 The influence of expectation, visual position,
and past history on current position
• We discovered that the apparent position of a sound source is highly
influenced by its visual position, expected position and its past history.
– Localization in a natural environment is almost always dominated by the
visual field, or a memory of a visual field.
– Expectation of azimuth and elevation is particularly important in sound
recording.
• In panning experiments we can alter the bandwidth or the band
frequency of a known source type, like speech. This change of
frequency mix will drastically alter the IDT and IID.
– And yet a source which appears to be located at a particular position with
one mix of frequencies will remain in that position when the frequency
mix is changed.
• Alternating the presentation from left to right by switching the speaker
channels breaks this hysterisis.
– The subject is asked to estimate the width between sound images which
alternate left and right,
– rather than the position of images presented consistently on one side or the
other.
Apparent width of broadband syllabic sources
• Broadband syllabic sources are consistently perceived as
narrow in width.
– Although when they are broken up into 1/3 octave or critical bands
the apparent direction of sound may vary over a wide range.
• The neurological process of separating sound into streams
assigns a “best guess” position to the entire stream, rather
than separating the perception into several streams in
different directions.
– Once the direction of a stream has been assigned, the hearing
apparatus is quite reluctant to change it.
• ESI – which is often present around syllabic sources, is
most often described as “source width” by untrained
listeners.
– On careful listening the source will be found to be narrow, but
surrounded by a local acoustic halo.
 Apparent source width (ASW)

• ASW is consistently cited in the literature as important in

musical acoustics.
• But human physiology insists that the apparent width of all
syllabic sources – speech, music with clear notes, etc – will
be sharp.
• The apparent width of continuous sources – legato string
sections, pink noise, etc. can increase when ITD and IID
become inconsistent or unstable.
– ASW may be a useful concept for such sources, but these sources
are not very interesting acoustically.
 Apparent width of broadband
continuous sources
• Reflection can increase the apparent width of continuous
sources.
• Such sources are often physically wide in any case.
– Examples might include an orchestral string section or a chorus.
• In this case fluctuating or inconsistent ITD and IID cause
the apparent width of the source to broaden.
– In acoustics, a common experimental model consists of a
loudspeaker which plays pink noise or legato strings.
– This example (and the concept of source width itself) is not very
useful.
Experimental Models
Diffraction around the
head can be modeled
with a combination of
delay and a shelving
filter. Typical filter
values can be
determined from
dummy-head data, with
a –3dB point of
1500Hz, and a shelf
depth of 6dB.
 Dummy Head measurement

Modified Neumann KU81 head. Sound source is front (0 degrees) for

the orange curve. Source angle is 45 degrees right for the blue curve.
 Using the model
• Because direction determination is strongly weighted by the pinnae
and middle ear, it is important to duplicate the frequency response at
the ear drum.
– The model assumes a flat frequency transfer between the model outputs
and the eardrum of the listener. The HRTF (head related transfer
function) from the ear drum to the left (or right) speaker position is not
included.
– This means the earphones in the experiment must be matched to the
HRTF’s of the individual listener.
– The best way to do this is by putting a probe microphone on the eardrum,
and matching the earphone response to the measured response of a
loudspeaker at the listening position.
• Results for poorly matched HRTFs will not correspond to natural
hearing.
– But will preserve the relative positions of different sources and are still
quite useful.
• The model allows us to analyze the interference at the entrance to the
ear canals, and relate what we find to the apparent source direction.
 Results from the model – using ITD only

Results assuming an interaural spacing of 21cm: The apparent position of low

frequencies matches the sine law. The sine/cosine law is accurate at 600Hz, and
above 600Hz the sound image moves rapidly to the left.
 Listening experiment results at High
Frequencies (ILD + ITD)
Apparent position of
1/3 octave filtered
speech as a function
of frequency.
Although the
original signal is
panned to a position
of 22 degrees, at
1700Hz the image is
outside the
loudspeaker basis,
(+-45 degrees.)

Note that the position of the broadband source is strongly pulled

outward by the increased width of the frequencies above 1kHz.
 Further observations:

• The model allows us to generate signals that combine

different IDTs and ILDs. Matlab can be used to remove
either the delay component or the amplitude difference
between the two ear signals.
– The resulting signals can then be tested for perception with
earphones.
• We find that, as expected, above 1kHz the direction
determination is dominated by the amplitude difference
(ILD), but that the ITD of the envelope of the signal still
contributes to the apparent position.
• A signal that varies in ITD only is perceived as closer to
the center than a signal that varies in both ITD and ILD.
– And is less reliably localized.
 Conclusions on two channel panning
• The apparent position of narrow band speech or syllabic music
sources in a typical two channel pan is HIGHLY dependent on
frequency
– A source panned half way between center and left with a conventional pan
pot can appear slightly left of center to full left, depending on the selection
of 1/3 octave frequency band.
– At frequencies above 700Hz the apparent position of these sources is
usually further from the center than would be expected from the
sine/cosine law.
• Images from broadband speech and syllabic music sources are
consistently perceived as further left or right from the center than
would be predicted by a sine/cosine pan law.
– A more accurate pan law can be made by simply expanding the scale on
the pan control, so half-left or half-right is closer to center.
• The discrepancy can be explained by the dominance of frequencies
between 700Hz and 4kHz in human hearing.
– These frequencies are consistently perceived as wider than low
frequencies in a pan which is half-way between center and either right or
left.
 Conclusions, continued
• The hearing mechanism appears to simply weight the apparent position
of each frequency band by the intensity of nerve firings in that band
when assigning an azimuth to a particular sound stream.
– Thus frequencies above 500Hz are particularly important
• Syllabic sound sources are perceived as narrow in lateral width
regardless of the positional discrepancy between critical bands.
– This observation seems to be a universal property of human perception.
• Fluctuation in IID or ITD causes an acoustic impression around the source, not
widening of the source itself.
– Continuous sources, such as pink noise or legato string sound, can be
broadened by positional discrepancy between bands, or rapidly fluctuating
IID and ITD.
• The expected position of a sound stream, and the past history of its
position, have a strong influence on perception.
– A sound source is usually prohibited from leaving its visual position.
– A sound source is highly resistant to moving to a new position once its
original position has been established.
 Conclusions on Panning for microphone technique

• Sound localization with amplitude panning is

NOT NATURAL
– It is strongly frequency dependent.
– Broad band sources appear narrow, but are localized with
difficulty.
• Sound localization with time delay panning is even
LESS NATURAL
– Time delay panning is useful only in the sweet spot, and this
restriction is not commercially viable.
– Low frequencies are always in the middle of the head.
– High frequencies also frequency dependent.
 So – how can we evaluate our
recordings?
• Recording standards are NOT arbitrary.
• Experience judging student recordings shows:
– Everyone on the jury panel agrees on the ranking of the submitted
recordings.
– When asked to explain their reasons everyone agrees on the
comments.
– The juries do NOT agree on the best method to achieve good
results. In fact, with regard to microphone type and position they
can be quite closed-minded.
– But when they do not know how the recordings were made, they
agree on their quality. This is very refreshing.
• Recordings stand or fall on their ability to satisfy the needs
of human hearing. This is constant among people, even
among recording engineers.
 Main Message:

• The recording venue is CRITICAL!!!

– Large (>2000 seat) concert halls can make stunningly beautiful
recordings.
• A wide variety of techniques can achieve satisfactory results.
– A technique that works well in a large concert hall will probably
NOT work well in a hall with 1200 seats.
• It is the job of the engineer to make a stunningly beautiful
recordings in the hall that happens to be available.
– Working to a world-class standard in a small space takes both
science and art.
 Spatial Properties of Recordings

• All engineers know how to judge instrumental balance and

tonal balance
– So we will not talk about these perceptions.
• We are going to talk exclusively about
– Intelligibility
• not really a problem in recordings
– Distance
– Localization
– Envelopment
• How can you learn to hear and to assess these perceptions?
• How can you achieve the best results?
 Training to hear localization
• The importance of ignoring the sweet spot
– Most research tests of localization use a single listener, who is strictly
restricted to the sweet spot.
– Your customers will not listen this way!
• And neither will the jury for your student submission. There are always at lest
three jurors, and they move around.
– A recording that only localizes well in the sweet spot will not make it past
the first round of judging!
• How do you know if the recording will pass this test?
– Move laterally in front of the loudspeakers. Does the sound image stay
wide and fixed to the loudspeakers, or does it follow you?
– Do the soloists in the center follow you left or right? If they do they are
recorded with too much phantom center.
• Since most 5 channel recording methods are derived from stereo
techniques almost all have too much phantom center.
• A center image that follows a listener who moves laterally out of
the sweet spot is the most common failing of even the best five
channel recordings.
» Play example
 Example: Time delay panning outside
the sweet spot.

Record the orchestra with a

“Decca Tree” - three omni
microphones separated by one
meter. A source on the left will On playback, a listener on the far right
give three outputs identical in will hear this instrument coming from the
level and differing by time delay. right loudspeaker. This listener will hear
every instrument coming from the right.
 Amplitude panning outside the sweet
spot.

If you record with three widely spaced A listener on the far right will hear the
microphones, an instrument on the left instrument on the left. Now the
will have high amplitude and time orchestra spreads out across the entire
differences in the output signals. loudspeaker basis, even when the
listener is not in the sweet spot.
 Training to hear distance

• Closely miked sources often sound in front of the

loudspeaker.
• They seem unnaturally forward and dry.
• Adding diffuse early reflections through all four lateral
speakers puts them behind the loudspeakers and in a
unified acoustical space of no perceivable size or shape.

» Play examples
Boston Cantata Singers in Jordan Hall
 Major Characteristics
• Chorus is deep in an enclosing stage-house with
significant reverberation.
• Small distances between microphones results in
unwanted leakage.
• Microphones pointed into the stage house increase
the amount of undesirable reverberation.
– Thus the chorus mikes, which must face the chorus, are
supercardiod to minimize reverberation pick-up.
– And the orchestra mikes face the hall, not the stage
house.
• Microphones in front do not pick up enough direct
sound from the chorus to supply the sense of
distance without also getting considerable mud.
Jordan Hall Setup
 Solutions
• Add distance to the chorus at the mixing
stage with controlled early reflections
• Minimize stage-house pickup wherever
possible
 Audio Demos
• Early reflections
• Late reverberation
 Training to hear MUD
• It is relatively easy to train yourself to hear mud, but it is
often very hard to avoid it.
• Mud occurs when the reverberant decay of the recording
venue has too much reflected energy in the 50-150ms
region of the decay curve.
– This is true of nearly all sound stages, small auditoria, and
churches.
• If you are recording in such a space with a relatively large
ensemble, you are in trouble.
 Example: John Eargle at Skywalker
ranch
• John Eargle made a wonderful 5.1 channel DVD audio
recording at the Skywalker ranch in Los Angeles.
• Skywalker is a large sound stage with controllable
acoustics. It is not a concert hall.
• As a consequence the reverberation radius is relatively
short. By my estimate (without having seen it) the radius
is about 3.5 meters.
• It is very easy to record mud in such a space.
– Many instruments are beyond the reverb radius.
– Adding more microphones only increases the reverberant pickup.
Example: Revels Chorus in the Sonic Temple
 Characteristics
• Main problem here was excessive reverberation level.
– Solution was to add blankets – a LOT of them. 648ft^2
– Here we list the measured reverberation times
– Hz blankets empty
– 8000 0.6 0.9
– 4000 0.8 1.2
– 2000 0.9 1.4
– 1000 0.9 1.4
– 500 1.0 1.3
– 250 0.9 1.3
– 125 1.1 1.4
– 63 1.0 1.5
– Reverb radius before the blankets: ~6 feet (2 meters)
– Reverb radius after the blankets: ~8 feet (2.7 meters)
 After the blankets
• Reverberation time drops below 1 second, the
magic number for the early decay in Boston
Symphony Hall
– Recording the band is easy, as we can mike them all
quite closely.
– Recording the chorus is hard, as there are >20 singers,
and we cannot get the microphones close enough to
each.
• Adding more microphones simply results in picking up more
reverberation!
– With the blankets we can record with adequate clarity
using only four supercardioid microphones.
• Once again we augment the early reflections in all outer
channels using the Lexicon.
• Late reverberation is also created using Lexicon late
reverberation.
 Reverberation Radius
• The reverberation radius changed from 6’ to 8’
when we added blankets. ~2dB.
– This is not a large enough change to account for the
perceived difference in sound.
• But the change in the total reflected energy in the
time range of 50-150ms (the undesirable time
range) is much larger: 4.5dB.
– This is a highly significant and desirable decrease!
• The decrease in the late reverberation (150ms and
greater) is 6dB.
– But we make this back up with the Lexicon.
 Training to hear envelopment
• Here it is essential that you move around the room, and
that you face different directions!

• You must fill the WHOLE room with the sound of the
original recording, and it must work when you face all
directions.

• Reproducing space only in front, or only in the rear, will

not get you the prize.
 3/0 versus 3/2
• OK, perhaps we need three speakers in the front, and amplitude
panning in the front.
• Why do we need two additional speakers and channels?

With decorrelated
Mono sounds poor because it We need at least four speakers
reverberation a few spatial
does not reproduce the spatial to reproduce a two
properties come through, but
properties of the original dimensional spatial sensation
only if the listener faces
recording space. that is uniform through the
forward. And the sense of
room.
space is stronger in the front.
Boston Symphony Hall
 Boston Symphony Hall
• 2631 seats, 662,000ft^3, 18700m^3, RT 1.9s
– It’s enormous!
– One of the greatest concert halls in the world – maybe the
best.
– Recording here is almost too easy!
– Working here is a rare privilege
• Sufficiently rare I do not do it. (It’s a union shop.)
– The recording in this talk is courtesy of Alan McClellan of
WGBH Boston. (Mixed from 16 tracks by the presenter)
– Reverb Radius is >20’ (>6.6m) even on stage.
– The stage house is enormous. With the orchestra in place,
stage house RT ~1 sec
Boston Symphony Hall, occupied, stage
to front of balcony, 1000Hz
 Why is the impulse response
relevant?
• Because the early decay (from the stage) is
short enough to get out of the way before it
muddies he sound.
• And the late decay (from the hall) is long
enough to provide envelopment.
Boston Symphony Orchestra in Symphony Hall
Boston Cantata Singers in Symphony Hall. March 17, 2002
 How can we reproduce
envelopment in a small room?
• The reverberant field of a LARGE room can be
reproduced in a SMALL room if:
– We can excite a fluctuating sound VELOCITY across
the listener’s head that mimics the fluctuating velocity
in the original space.
– To do this we MUST have at least two LF drivers on
opposite sides of the listener.
– If the listener is allowed to turn the head, we must have
at least 3 independent drivers, and four is better!
– All the LF drivers must be driven by independent
(uncorrelated) reverberation signals, derived from a
large, non-steady-state room.
 Low frequencies are particularly
important!
• In our concert hall and opera work it is frequencies below
300Hz where the major benefit is achieved.
– The result is “inaudible” but highly effective in increasing the
emotional power of the music.
• It is commonly believed that because we “cannot localize”
low frequencies in a playback room we need only one LF
driver
– We can however easily hear the difference on reverberation.
• It is often the case that using a shelf filter on the rear
channels can greatly improve the surround impression.
 Shelf filter for rear channels

Applying a shelf filter to the rear channels increases subjective

envelopment dramatically without drawing attention to the rear
speakers.
 Correlation in the omni “Hauptmicrophone”
two omnis just behind the conductor.

___ = measured correlation; - - - = calculated, assuming d=25cm

 Let’s build the hall sound
• We need decorrelated reverberation in both the
front and the rear with equal level
• Test just the hall microphones to see if the
reverberation is enveloping and uniform.
• Then add the microphones for the direct sound.
• Here there is too little chorus in the reverberation!!
• So we add hall (equally in all four outer speakers)
from the Lexicon surround reverberator.
Oriana Consort in Swedenborg Chapel
 Major Characteristics
• Hall has relatively low volume of 1450m^3 at the
same time as medium RT ~1.5s
– Low Volume and high RT means the reverb LEVEL
will be very high!
• We will have to keep the microphones close
– Reverb time is a bit to short for this type of music.
– With a small group it might be possible to use a
microphone pair for a two channel recording.
• But it might sound better if you did not.
Oriana Setup
 Surround Recording
• The recording is created using the
multimicrophone front array, (equalized)
– Augmented with an early reflection pattern
from Lexicon in all four outer speakers.
• The surround environment is created using
the rear microphones (equalized for the bass
roll-off) for the rear channels.
– And Lexicon late reverberation for the front,
• And some in the rear also.
 The Microphone Pair

A venerable pair of multi-pattern microphones

Another possibility
 Pressure Gradient Microphones
• Pressure gradient microphones are a
combination of an omni and a figure of
eight.
• When the two are mixed with equal on-axis
sensitivity, a cardioid results.
– Reduce the gain of the omni by 6dB and you
have a Supercardioid.
– Reduce the gain of the omni by 10dB and you
have a Hypercardioid.
 Problem:
• The figure of eight in nearly all available
microphones has a bass roll-off, typically at about
120Hz. (Depends on diaphragm size.)
– When we combine this with the omni – which (may) be
inherently flat at LF, and:
• The overall sensitivity decreases at LF
– The mike sounds weak in the bass compared to an omni
• The pattern may become omni directional at low frequencies
– This is particularly true for large dual-diaphragm mikes.
 Solution
• One solution to this problem is to “correct” it by measuring
the microphone at a distance of 1M from the sound source!
– A spherical sound wave increases the LF velocity of the sound at
6dB/octave when the distance to the sound source approaches ½
the wavelength.
– A one meter measurement distance exactly compensates for the
inherent roll-off of the velocity transducer, and an apparently
perfect microphone results.
• A more satisfactory solution would be to equalize the
figure of eight pattern electronically before combining it
with the omni.
– The “Soundfield” microphone does this.
• One can also roll off the omni response (electronically or
mechanically) to match the figure of eight.
– Mr. Wuttke (Schoeps) takes this approach.
 Consequences
• Nearly all available directional microphones either roll off
the bass response,
– Which can be compensated for at the mixing desk
• Or they become omnidirectional at low frequencies,
– Which usually cannot be compensated.
• Or they do both.
– The venerable microphones shown earlier do both.
• The consequence for a ORTF – style pair is
– The low frequencies will be generally weak
• Which can be compensated.
– The low frequencies may be monaural
• Which is more difficult to compensate.
• But which can be fixed with a Blumlein shuffler circuit
• Be sure to equalize the LF when you use directional
microphones!
 Correlation of reverberation
• Remember we are (desperately) trying to
keep the reverberation decorrelated.
– We can do this with a coincident pair if we
choose the right pattern and the right angle
 Ideal angle as a function of microphone pattern for
decorrelated reverberation in a coincident pair.

• It is NOT possible to achieve decorrelation with cardioid

microphones!
 Correlation through distance
• Normal ORTF technique with cardioid
microphones reduces the correlation at HF
by adding distance.
• But the trick does NOT work at LF,
• And LF correlation is exceedingly audible
 Correlation of two omnidirectional microphones in a
reverberant field as a function of microphone separation.

• Notice high correlation below 300Hz, and negative

correlation at 800Hz.
• Frequency and distance are inversely proportional.
 Audio Demos
• Omni pair
– Slavery Documents
• Cardioid Pair
– AKG large diaphragm mikes with Oriana
• SuperCaridoid pair.
– AKG large diaphragm mikes with Oriana
• Multimiked front image.
– Oriana with four Schoeps Supercardioids
 We can use a goniometer

X-Y plot of the omni front pair in Slavery Documents. Red trace
is Low pass filtered at 200Hz, Yellow trace LP filtered at 100Hz.
 Goniometer with AKG pair

X-Y plot of Oriana front pair with The same data, filtered at 100Hz.
a 200Hz LP filter. Red is Note that now the supercardioid
Cardioid, Yellow is Supercardioid is behaving like an omni.
 Measure the decorrelation in the playback room

Hall sound from Slavery

Documents. All speakers
decorrelated.

We can make an x-y plot of the left ear and right ear
signals, after boosting the L-R component ~10dB
below 200Hz. These plots cover the frequency
range of 20-100Hz.
Same, with a signal correlation
The change in the ear correlation is quite audible. of 30%
Final Mix
 Conclusions
• Recording is a lot of fun!!!
• It is a great pleasure, and is often useful, to
understand some of the science behind the
microphones.
• Although simple techniques using microphone
pairs or arrays can be seductive, a world-class
sound usually requires many microphones, a lot of
work, and artificial acoustic augmentation.
– Time delay panning is undemocratic. Avoid it.
• Make SURE your reverberation is decorrelated,
particularly at low frequencies.
THE END