Skype Audio Lab

Setup Guide

Disclaimer: This document is provided “as-is”. Information and views expressed in this document, including URL and
other Internet Web site references, may change without notice. You bear the risk of using it.

This document does not provide you with any legal rights to any intellectual property in any Microsoft product. You
may copy and use this document for your internal, reference purposes.

Feedback: You have no obligation to give Microsoft any suggestions, comments or other feedback ("Feedback")
relating to the materials herein (“Materials”). However, any Feedback you voluntarily provide may be used in Microsoft
products and related specifications or other documentation (collectively, "Microsoft Offerings") which in turn may be
relied upon by other third parties to develop their own products. Accordingly, if you do give Microsoft Feedback on any
version of these Materials or the Microsoft Offerings to which they apply, you agree: (a) Microsoft may freely use,
reproduce, license, distribute, and otherwise commercialize your Feedback in any Microsoft Offering; (b) you also grant
third parties, without charge, only those patent rights necessary to enable other products to use or interface with any
specific parts of a Microsoft product that incorporate your Feedback; and (c) you will not give Microsoft any Feedback
(i) that you have reason to believe is subject to any patent, copyright or other intellectual property claim or right of any
third party; or (ii) subject to license terms which seek to require any Microsoft Offering incorporating or derived from
such Feedback, or other Microsoft intellectual property, to be licensed to or otherwise shared with any third party.

© 2015 Microsoft Corporation. All rights reserved.

 Skype Audio Lab Setup Guide

Table of Contents
Revision History ............................................................................................................................................ 4
1.0 Introduction ...................................................................................................................................... 5
2.0 Definitions and abbreviations ........................................................................................................... 6
3.0 References ........................................................................................................................................ 7
4.0 Overview ........................................................................................................................................... 8
5.0 Lab Setup........................................................................................................................................... 9
5.1 Lab facilities................................................................................................................................... 9
5.1.1 Anechoic test environment................................................................................................. 10
5.1.2 Reverberant test environment ........................................................................................... 12
5.2 Equipment ................................................................................................................................... 14
5.2.1 Software .............................................................................................................................. 14
5.2.2 Measurement equipment ................................................................................................... 14
5.2.3 PCs and add-ons .................................................................................................................. 15
5.3 Preparing the PCs ........................................................................................................................ 16
5.3.1 Setting up ACQUA PC .......................................................................................................... 16
5.3.2 Setting up REF PC ................................................................................................................ 17
5.3.3 Setting up the DUT PC ......................................................................................................... 17
5.3.4 Setting up HAE-BGN PC ....................................................................................................... 17
6.0 Preparing the lab for testing ........................................................................................................... 18
6.1 Setting up MFE VI ........................................................................................................................ 18
6.1.1 MFE VI.I introduction .......................................................................................................... 18
6.1.2 Naming convention for measurement equipment ............................................................. 19
6.1.3 Calibrating inputs ................................................................................................................ 19
6.1.4 Calibrating artificial mouths ................................................................................................ 19
6.2 Setting up the HAE-BGN system ................................................................................................. 20
6.2.1 Calibrating the HAE-BGN system ........................................................................................ 20
6.2.2 Connecting the HAE-BGN system to ACQUA PC ................................................................. 20
7.0 Setting up MFE VI.I for testing ........................................................................................................ 22
7.1.1 Naming conventions for test setups ................................................................................... 22
7.1.2 Static settings in ACQUA (for Skype measurements).......................................................... 22
7.1.3 Measurement Settings in ACQUA ....................................................................................... 23

Page 2 of 53
 Skype Audio Lab Setup Guide

7.1.4 Calibration Assignments in ACQUA ..................................................................................... 24

7.1.5 MFE VI.I “Skype” settings in an anechoic test environment ............................................... 26
7.1.6 MFE VI.I “Skype” settings in a reverberant test environment ............................................ 27
8.0 Setting up the DUT .......................................................................................................................... 29
8.1 Controlling the audio processing ................................................................................................ 29
8.1.1 Using DUT Editor to test raw mode .................................................................................... 29
8.1.2 Using XML keys to control the audio processing ................................................................ 29
8.2 Executing the test sequence in ACQUA ...................................................................................... 29
8.2.1 Choosing correct specification and ACQUA settings .......................................................... 29
8.2.2 Preparation sequence ......................................................................................................... 30
8.2.3 Choosing the correct tests .................................................................................................. 30
9.0 Annex-A: How to import a database............................................................................................... 35
10.0 Annex-B: Step by step guide how to recalibrate a G.R.A.S. ¼” measurement microphone using
B&K pistonphone ........................................................................................................................................ 36
11.0 Annex-C: Step-by-step guide for how to recalibrate G.R.A.S. mouth ............................................. 39
12.0 Annex-D: How to position DUT in an anechoic test environment .................................................. 43
12.1.1 Device with headset audio UI in an anechoic test environment ........................................ 43
12.1.2 Device with handset audio UI in an anechoic test environment ........................................ 43
12.1.3 Device with handheld speakerphone audio UI in an anechoic test environment .............. 44
12.1.4 Device with personal speakerphone audio UI in an anechoic test environment ............... 45
12.1.5 Device with group speakerphone audio UI in an anechoic test environment ................... 46
12.1.6 Device with long-range speakerphone audio UI in an anechoic test environment............ 47
13.0 Annex-E: How to position DUT in a reverberant test environment................................................ 48
13.1.1 Device with headset or handset audio UI in a reverberant test environment ................... 48
13.1.2 Device with handheld speakerphone audio UI in a reverberant test environment ........... 49
13.1.3 Device with personal speakerphone audio UI in a reverberant test environment ............ 50
13.1.4 Device with group speakerphone audio UI in a reverberant test environment ................. 51
13.1.5 Device with long-range speakerphone audio UI in a reverberant test environment ......... 52

Page 3 of 53
 Skype Audio Lab Setup Guide

Revision History
Revision Date Description

1.0 07/09/14 First release

Page 4 of 53
 Skype Audio Lab Setup Guide

1.0 Introduction
The Skype & Lync Audio Test Specification includes performance requirements and testing procedures
for audio device certification. This guideline provides additional information intended to be used to aid
replicating the audio test environment. It is provided for guidance and best practices only.

This document was developed with the intent not to repeat or copy information from the Skype & Lync
Audio Specification document, but rather to provide an example of a lab setup and to share experience
from setting up multiple labs.

In the event this document presents information that conflicts with the Skype & Lync Audio Test
Specification, then the most current document takes precedence.

Page 5 of 53
 Skype Audio Lab Setup Guide

2.0 Definitions and abbreviations

Term Definition
ACQUA ACQUA is a dual-channel analysis system developed by HEAD acoustics. The
acronym stands for Advanced Communication Quality Analysis.
http://www.head-acoustics.de/eng/telecom_acqua.htm
AEC Abbreviation of the words “Acoustic Echo Canceler.”
Artificial ear A device used to measure the acoustic output of earphones. This acoustic ear
incorporates an acoustic coupler and a calibrated microphone to measure sound
pressure. The resulting combination has an overall acoustic impedance similar to
that of the average human ear over a given frequency band.
In Skype & Lync Audio Test Specification context an artificial ear Type 3.3 is used.
Artificial mouth A device consisting of a loudspeaker mounted in an enclosure and having a
directivity and radiation pattern similar to those of the average human mouth.
The frequency response of artificial mouth is compensated such that it provides a
flat frequency response at MRP from 80 Hz to 11 kHz by equalizing the signal
played back.
Audio UI Categorization based on Skype & Lync Audio Test Specification.
DUT Abbreviation of the words “device under testing.”
HATS Abbreviation of the words “Head and Torso Simulator.” The HATS is usually
equipped with an artificial mouth and two artificial ears. For handset testing, the
handset positioner option module is needed as well.
MRP Abbreviation of the words “Mouth Reference Point.”
REF Abbreviation of the word “reference.” In this document, context used in relation
with the PC that is set up to accept calls from DUT for testing purposes.

Page 6 of 53
 Skype Audio Lab Setup Guide

3.0 References
Skype & Lync Audio Test Specifications

ITU-T P.341

ITU-T P.64

ETSI EG 202 396

ACQUA 3 Help System

ACQUA Database Administrator Help

HEAD acoustics webpages

Page 7 of 53
 Skype Audio Lab Setup Guide

4.0 Overview
The following areas are covered in this guideline:

1. Suggestions about what to consider when choosing facilities

2. Equipment used in the audio labs
3. Setting up the equipment
4. Calibration
5. Test setups

The goal of audio testing is to complete the test report against the specification with all required
information and test data.

Page 8 of 53
 Skype Audio Lab Setup Guide

5.0 Lab Setup

The objective of this section is to suggest how to equip a lab for performing testing according to the
Skype & Lync Audio Test Specification. There are many possible alternatives to the items described
herein, and the intent is not to restrict the lab operator to any particular setup.

5.1 Lab facilities

When planning the facilities, it is possible to optimize based on the solutions to be tested in the lab.

The Skype & Lync Audio Test Specification separates devices based on audio processing:

A. Devices using all Skype’s audio preprocessing (RAW capture pin)

 All test cases are run in anechoic or semi-anechoic room
B. Devices using custom software or hardware audio preprocessing (Processed capture pin)
 Most test cases are run in anechoic or semi-anechoic room
 Some extended test cases are run in reverberant room with background noise simulation
setup

The Skype & Lync Audio Test Specification separates devices based on audio UI:

 Headset (monaural, binaural)

 Handset (in handset use-case mode)
 Speakerphone

Testing devices falling under Option A specification

The devices falling under Option A do not require the reverberant room and the background noise
simulation system. Thus, the requirement for test environment is more relaxed.

Headset or handset audio UI under Option A can be tested in an anechoic room or a semi-anechoic
room, or in a quiet, low-reverb room where the closest objects are at least 2 meters (m) from the HATS.
Because the microphone and earpiece are very close to the HATS mouth/ear, the room reverb and
reflections are less likely to interfere with the meaurements.

For speakerphone audio UI devices, an anechoic or semi-anechoic room is strongly recommended.

Because the microphone/speaker is farther from the artificial mouth/microphone, the test room
acoustic reflection and room modes will show up, especially in frequency response test results and
POLQA objective MOS results.

Testing devices falling under Option B specification

Devices falling under Option B do require the reverberant room with the background noise simulation
system.

Headset or handset audio UI under Option B can be tested in an anechoic room or a semi-anechoic
room, or in a quiet low-reverb room where the closest objects are at least 1.5 m from the HATS. Because
the microphone and earpiece are very close to the HATS mouth/ear, the room reverb and reflections are

Page 9 of 53
 Skype Audio Lab Setup Guide

less likely to interfere with the meaurements. The Option B specification adds test cases for AEC and
noise suppression in a reverberant room.

For speakerphone audio UI devices, an anechoic room or semi-anechoic room is strongly recommended.
Because the microphone/speaker is farther from artificial mouth/microphone, the test room acoustic
reflection and room modes will show up, especially in frequency response test results and POLQA
objective MOS results. The Option B specification adds test cases for AEC and noise suppression in the
reverberant room.

5.1.1 Anechoic test environment

For testing speakerphone audio UI devices, the anechoic chamber or semi-anechoic chamber used
should fulfill the acoustical requirements for high-quality, wideband VoIP and telecom product testing.
Please refer to ITU-T P.341 section 4.1 for the recommended parameters for the test room and methods
to validate the room performance.

Figure 1: Anechoic room in Skype labs

For headset and handset audio UI devices, a similar room is recommended. However, the bare minimum
requirements are the following:
 There should be no big items closer than 1.5 m from the artificial mouth in any direction.
 The average room noise level should be below 29 dBSPL(A).
 The room reverberation time (RT60) should be below 0.3 seconds at frequencies between
150 Hz to 12 kHz.

When building a lab to measure devices with speakerphone audio UI, consideration of the following
practicalities is recommended:

 Ergonomics of frequency response measurements – the position of the artificial mouth and
measurement microphone needs to be adjusted for send and receive frequency response
measurements as shown in Figure 2.

Page 10 of 53
 Skype Audio Lab Setup Guide

Figure 2: Measurement positions for -7.5 deg, 0 deg, +15 deg

measurement at 30 cm, 50 cm, and 100 cm

The first option to achieve this in smaller anechoic booths without a hard floor is by using a sliding
holder on a horizontal pole that allows moving the mouth and microphone left and right by the
necessary distance. A suitable product will be available for purchase from HEAD acoustics.

Figure 3: Artificial mouth and ½-inchmeasurement microphone

mounting rig

The second option in case of a semi-anechoic room or a bigger anechoic room is to physically move the
stand on the floor left and right, respectively.

 Means to simulate back wall if a TV is tested. If other solutions also are to be tested, making is
removable is recommended. In case of a small room, it might be worth considering a solution to
simplify sliding the TV onto a table close to the wall.

Page 11 of 53
 Skype Audio Lab Setup Guide

Figure 4: Simulated back wall with slide rails for small anechoic booths

5.1.2 Reverberant test environment

The reverberant test room follows the guidelines in ETSI EG 202 396 with the following
recommendations based on audio UIs:

 For testing headset and handset audio UI devices any room compliant with EG 202 396 Part 1 is
suitable for testing.
 For testing speakerphone audio UI devices in addition to EG 202 396 Part 1 Section 6.1, the
additional requirements apply as follows:
 Room size – the room size should be in range between 2.7 m x 3.7 m and 3.5 m x 4.4 m.
Room height, 2.2 m to 3.25 m.
 Treatment of the room – the reverberation time of the room should be less than 0.7
sec, but higher than 0.4 sec in a frequency range between 100 Hz and 8 kHz. The reverb
time should be declining toward high frequencies, but should not have dips or peaks in
some octave bands that deviate more than 0.2 sec compared to the adjacent octave
bands on either side. Such a declining trend of reverb time versus frequency represents
a common meeting room or living room acoustics.
 Noise floor – to avoid room noise influencing the test results, the average noise floor in
the room should be below 29 dBSPL(A).

Page 12 of 53
 Skype Audio Lab Setup Guide

Figure 5: Recommended shape of the room’s reverb time in octave bands – follows
the average reverb time shape of meeting rooms and living rooms.

To be able to fine-tune the room reverb time, it is wise to specify some room surfaces to have variable
absorption, either by replacing wall panels or by covering the absorptive parts with hard surfaces. See
the sample in Figure 6.

Figure 6: Absorptive part of walls can be opened or covered with hard panels to
vary room reverb time

Page 13 of 53
 Skype Audio Lab Setup Guide

5.2 Equipment
This section describes the main hardware pieces used in the audio lab. It is separated into the following
based on function and severity:
 Software:
 To test devices using all Skype audio preprocessing (RAW capture pin)
 Additional licenses needed for testing devices using custom software or hardware audio
preprocessing (Processed capture pin)
 Measurement equipment:
 Must have articles from HEAD acoustics
 Items HEAD acoustics can provide, but alternative solutions and sources can be used
 PCs:
 Requirements for PCs to be used
 Add-ons for PCs

5.2.1 Software
Necessary licenses from HEAD acoustics to enable testing if Skype uses raw audio:
 ACQUA base license
 ACOPT 02 – Analysis package
 D60026 – Skype-Lync Unified test standard
 ACOPT 29 – Option EQUEST
 ACOPT 30 – Option POLQA
To enable testing for the case where Skype is at least partially relying on custom processing, all licenses
listed above are needed. So are the following:
 ACOPT 21 – Option 3QUEST
 ACOPT 32 – Option SpeechBased DT

5.2.2 Measurement equipment

Must-have articles from HEAD acoustics:
 MFE VI.I
 HEAD acoustics background noise simulation system (4.1 channel system)
This item is a system of multiple pieces. Consulting with HEAD acoustics is strongly
recommended.

Items needed that can be bought through HEAD acoustics. Nevertheless, the source can be chosen:
 Head and Torso Simulator (HATS)
Both HEAD acoustics and B&K products can be used for headset/handset testing:
 B&K model 4128C (with handset positioner if testing handsets)
 HEAD acoustics HMS II.3-33 (with handset positioner HHP III.1 if testing handsets).
 Artificial Ear pinna type 3.3
 ¼” pressure measurement microphone together with a pre-amplifier with a LEMO connector
This microphone is used to calibrate the HATS mouth and also is used in a reverberant room as

Page 14 of 53
 Skype Audio Lab Setup Guide

reference microphone for 3QUEST tests.

In Skype labs: G.R.A.S. 40BP + G.R.A.S. 26AC
 ½” free field measurement microphone and pre-amplifier with a LEMO connector
This is used together with the artificial mouth in an anechoic test environment when testing
products with speakerphone audio UI.
 Microphone calibrator together with adaptors enabling the capability to calibrate HATS ear and
¼” and ½” measurement microphones
In Skype labs: G.R.A.S. 42AB
 LEMO microphone cable(s)
It is possible to get all the testing done using only one cable, but considering Ergonomics we
recommend at least two. These cables are available in some predefined lengths as well as
custom length. Optimal length should be decided based on lab layout and cabling routes.
In Skype labs: one pair of cables going to a reverberant environment and one cable for an
anechoic room. In a reverberant test environment, one cable is connected to the HATS ear and
another to the reference microphone used for 3QUEST tests.
 Loudspeakers, subwoofer, and power amplifiers for HAE-BGN system.
It is strongly recommended to consult with HEAD acoustics when choosing.
In Skype labs: Canton Plus X.3 loudspeakers, HEAD Subwoofer for Sound Card, and KMT DC3
power amplifiers.
 A mounting pole to enable simple mounting and positioning of the ½” measurement
microphone and artificial mouth for speakerphone measurements.
HEAD acoustics will offer a solution for purchase.

5.2.3 PCs and add-ons

Add-ons:
 Sound card in Reference Skype Client PC
 RME HDSP9632 (PCI socket)
 HDSPe AIO (PCI express socket alternative)
Consult Head Acoustics documentation for a correct driver version for RME sound card.

If ordering separately, make sure the sound card has the 9-pin D-sub to 2 x RCA Digital, 2 x XLR
Digital cable included (BO968).

 Wireless router with DHCP server and Wi-Fi access point. (The access point can be separate if
the device is used in replicator mode and the DUT device remain in same IP subnet.)
It is needed to minimize the network impact to measurement results. Connecting the DUT and
REF PC to the same router helps to ensure a local DHCP server is used and IP addresses in the
same subnet are assigned to both REF PC and the DUT device. This will make sure that the Skype
call is UDP-UDP and no extra latency or bandwidth limitations occur.

Recommendations and requirements regarding PCs needed:

 ACQUA PC:
 Operating system: Microsoft Windows 8.1 64-bit

Page 15 of 53
 Skype Audio Lab Setup Guide

 Processor: Intel Core i5 or equivalent ≥ 2.6 GHz

 Installed Physical Memory: ≥ 4 GB
 Hard Drive: ≥ 1 TB
 VGA card: supports two monitors
 REF PC:
 Operating system: Microsoft Windows 8.1 64-bit
 Processor: Intel Core i5 or equivalent ≥ 2.6 GHz
 Installed Physical Memory (RAM) ≥ 4 GB
 Sound card: as described above under add-ons
 Webcam: capable to send 1080p video over Skype
 Optional: a good active monitor speaker to enable listening DUT send signal from REF PC
sound card output. (Skype playback signal can be routed to both the AES/EBU and the
headphone analog output connector simultaneously.)
Monitoring DUT send signal during testing makes the failing send tests and echo
cancellation performance fails more intuitive to understand.
In Skype labs: Genelec 8020A
 HAE-BGN PC
 Processor: Pentium IV compatible, ≥ 2.0 GHz
 Installed Physical Memory (RAM) ≥ 1 GB
 Hard Drive: ≥ 40 GB
 5 USB ports
 RME HDSP9632 (PCI socket) or HDSPe AIO (PCI express socket alternative)
 Considerations points regarding the DUT PC:
 If only integrated solutions are to be tested, then there is no need for a DUT PC.
 If the product category to be tested does not have audio playback solution (for example,
external webcam with a microphone), then some loudspeakers should be at hand to
simulate a realistic playback solution. Suggested model is Logitech LS11.
 Operating system: Microsoft Windows 8.1 64-bit
 Processor: Intel Core i5 or equivalent ≥ 2.6 GHz
 Installed Physical Memory (RAM) ≥ 4 GB

5.3 Preparing the PCs

5.3.1 Setting up ACQUA PC
 Install ACQUA according to the guidelines provided by HEAD acoustics.
There are dedicated guidelines provided for the first time installation in order to assist with pre-
conditions.
 Physically connect the MFE VI.I and required dongles.
 Install ACQUA databases using the ACQUA Database Administrator.
If the database DC/DVD is not available, then download the standards as instructed in emails
from HEAD License-Management.
Custom made databases can be installed as described in Annex-A: How to import a database.

Page 16 of 53
 Skype Audio Lab Setup Guide

5.3.2 Setting up REF PC

1. Set up sound card:
a. Install the sound card into the REF PC.
b. Install the sound card drivers recommended by HEAD acoustics.
c. Connect AES/EBU cables as following:
i. Sound card output should be connected to MFE VI.I AES/EBU input socket at the
back of MFE.
ii. Sound card input should be connected to MFE VI.I AES/EBU output socket at the
back of MFE.
2. Set up the HEAD acoustics dongle for REF PC.
a. Connect the dongle. If not recognized, then install driver.
3. Set up the Skype client bundled with the standard:
a. Locate and install the Skype client.
b. Create log on credentials that will be used only in the REF PC.
c. Log on and adjust the settings to automatically accept incoming calls and send video. If
there are only a few known accounts used with DUT, add those accounts to contacts
and adjust privacy options, respectively.
4. Install REF/DUT editor.
5. Configure the REF PC to “REF PC mode” with the REF/DUT editor.
6. Fully close and restart Skype client (action only needed once to make sure settings take effect).

5.3.3 Setting up the DUT PC

1. Set up a recommended versions of Skype. In case of doubts about the version, consult HEAD
acoustics or the Skype Media Hardware Certification team.
a. Create log on credentials that will be used only with DUT or DUT PC.
b. Log on once.
c. Add the Skype account used on REF PC to the Skype contacts list.
2. Copy and run the DUT Agent in DUT PC (only for RAW capture specification tests).

5.3.4 Setting up HAE-BGN PC

1. Set up the sound card:
a. Install the sound card in the HAE-BGN PC.
b. Install the sound card drivers recommended by HEAD acoustics.
2. Install HAE-BGN software and drivers.
3. Install the Remote control HAE USB male <-> USB male cable (CUU I) between the HAE-BGN PC
and the ACQUA PC.
4. Install the AES/EBU and USB cables as instructed in the HAE-BGN user manual.

Page 17 of 53
 Skype Audio Lab Setup Guide

6.0 Preparing the lab for testing

Typically, HEAD Acoustics representatives help with setting up and calibrating the lab for first-time
setup. This setup guide is meant to support the test engineer in case the equipment needs to be moved
to another location or needs to be modified for Skype usage.

6.1 Setting up MFE VI

6.1.1 MFE VI.I introduction
The MFE VI.I has two input and output channels. However, there are multiple connectors available per
channel. The channel and connection have to be matched in MFE VI.I Settings in ACQUA software.
Figures 7 and 8 show the settings window in ACQUA and physical connectors on the MFE VI.I.

Figure 7: MFE VI.I Settings

Figure 8: Front and back side of the HEAD acoustics MFE VI.I

Two channels both ways enable testing over a Skype call – the MFE VI.I to make electric connection to
REF PC and bring acoustic input/output measurement equipment at DUT.

Figure 9: Testing over Skype call using MFE VI

Page 18 of 53
 Skype Audio Lab Setup Guide

6.1.2 Naming convention for measurement equipment

It would simplify the testing if the naming convention of measurement equipment is thought through
before starting with calibrations.
It might be useful to use labels with the assigned names on the equipment. That is especially relevant in
case some articles are duplicated.

In Skype labs, the following naming conventions are used in ACQUA:

 Input devices: vendor identifier, type, (end of pre-amp serial number, if applicable)
Examples: GRAS half 408, B&K RightEar
 Mouths: type, vendor, (end of serial)
Examples: Mouth GRAS 434, HATS B&K
 Predefined settings: Environment, DUT audio UI, input-output pair identifier
Examples: Anechoic – Headset (HEAD HATS), ETSI – Speakerphone (B&K HATS)

6.1.3 Calibrating inputs

On the ACQUA Preparations drop-down menu, choose Input calibration.

Calibrating measurement microphones is straightforward, but with HATS the following has to be noted:
 Nominal sound pressure level on calibrator might need a correction factor when using an
adapter for HATS ear.
 It is recommended that two persons calibrate the HATS ears because the calibrator needs to be
held against the HATS ear from the moment calibration is initiated.
 Both HATS ears (if available) should be calibrated, although in Skype testing only the HATS right
ear is used. Both ears are needed when setting up the HAE-BGN system.

A step-by-step example of calibrating a measurement microphone with ACQUA is given in Annex-B: Step
by step guide how to recalibrate a G.R.A.S. ¼” measurement microphone.

6.1.4 Calibrating artificial mouths

On the ACQUA Preparations drop-down menu, choose Mouth/Loudspeaker Equalization.

It is important to consult with the vendor’s manual to learn if any additional gadgets are needed and
how the microphone has to be mounted.

For example, in B&K HATS case the ¼’’ microphone with ¼’’ preamp can be used directly. In case of
HEAD acoustics HATS, a combination of ½’’ preamp to ½’’-> ¼’’ adapter and ¼’’ pressure capsule
combination is needed in order to use the supplied microphone holder.

After the calibration, check the necessary compensation curve gain. If it goes to high values of 20+ dB,
that is not recommended and a narrower band should be used to avoid this.

A step-by-step example of calibrating the G.R.A.S. artificial mouth (44AB in ACQUA) is given in Annex-C:
Step-by-step guide for how to recalibrate G.R.A.S. mouth.

Page 19 of 53
 Skype Audio Lab Setup Guide

6.2 Setting up the HAE-BGN system

All information in this section is covered in the manual provided by HEAD acoustics, but below are some
steps to simplify reading it. The HAE-BGN manual covers two usage modes: 1) Car and 2) ETSI room. The
sections for HAE-car can be skipped.

6.2.1 Calibrating the HAE-BGN system

Thoroughly covered in the manual, there are a few things worth highlighting separately:
 The setup and wiring for calibration is different to the setup and wiring in later noise playback
usage. For calibration, the MFE VI.I is connected to HAE-BGN PC USB port and the AES/EBU cable
is routed from MFE VI.I output to the HAE-BGN PC AES/EBU input.
After successful calibration the measurements the MFE VI.I is reconnected to the ACQUA PC.
 HAE-BGN software will instruct the user to calibrate the HATS ears. This process can be done as
instructed or skipped and the values from ACQUA Calibrations used instead if the ACQUA
system has been recently calibrated.
 It can be necessary to change the delay values for front/back channels to get the equalizer
response to fit the tolerance mask.
 The subwoofer can be repositioned if the frequency response calibration does not fit the
required tolerance mask. Generally, the changes in subwoofer position have to be at least 30 cm
to 50 cm in order to gain effects in the low frequency response.

5.
Figure 10: HAE-BGN setup for system calibration (HATS is required)

6.2.2 Connecting the HAE-BGN system to ACQUA PC

1. Once calibrated, the MFE VI.I is to be disconnected from the HAE-BGN PC and reconnected to
the ACQUA PC.
2. In Skype setup, it is recommended using one LEMO cable for the HATS right ear and another for
the ¼” measurement microphone used for 3QUEST reference signal. Only one should be
connected to the MFE VI.I Mic Channel 2 at any given time.
If one LEMO cable is used, then the cable can be left connected to MFE VI.I for the tests in
reverberant room, but the tester needs to switch the other end of the cable between the ¼” for
3QUEST or the HATS right ear for echo cancellation tests.
3. Connect the Pulse Splitter Box (PSB III) to MFE VI.I (Coax <-> MFE Pulse cable).

Page 20 of 53
 Skype Audio Lab Setup Guide

4. Connect the PSBIII AES/EBU output to HAE BGN PC sound card input.
5. Connect RME AES/EBU output to the LabBGN input (or loop the cable through all PEQ V
equalizers in case of older setup). The AES/EBU playback loop should end in PSB III forming a
”closed loop” of AES/EBU output and input cable path.
Skipping this step allows the play back of noises if initiated manually, but automated tests are
not going to work and give an error message “No synchronization Can’t start a recorder…”
6. Connect HAE-BGN PC and ACQUA PC using HAE Remote Control-Cable (USB-USB cable).

Figure 11: Settings for RME sound card in HAE-BGN setup

Page 21 of 53
 Skype Audio Lab Setup Guide

7.0 Setting up MFE VI.I for testing

This section describes the MFE VI.I setups for all device types that are to be tested against the Skype &
Lync Audio Test Specification. If a setup is customized to cover only a subset, then some subsections
might not apply.

7.1.1 Naming conventions for test setups

It would simplify the testing if the naming convention of test setups is intuitive. That is very important in
case multiple people use ACQUA.

The optimal naming convention depends on the solution categories tested and the equipment available.
In Skype labs, the following naming conventions are used in ACQUA:
 Measurement/report: Lab descriptor; Audio UI descriptor if relevant; Test set description if
relevant; Measurement equipment descriptor.
Examples: Anechoic - Speakerphone (mouth + mic), ETSI (B&K HATS).
 Calibration assignment: Lab descriptor; Test set description if relevant; Measurement
equipment descriptor.
Examples: Anechoic – Mouth + ½” mic, ETSI – 3QUEST(B&K + ¼”mic)
 MFE VI.I Settings: Lab descriptor; Audio UI descriptor if relevant; Test set description if
relevant; Measurement equipment descriptor.
Examples: Anechoic - Speakerphone (mouth + mic), ETSI – 3QUEST.

7.1.2 Static settings in ACQUA (for Skype measurements)

On the ACQUA Settings drop-down menu, choose Settings.

Figure 12: Static settings in Measurement Settings

All settings shown on Figure 12 are the same, regardless of the device under testing or the lab.

Page 22 of 53
 Skype Audio Lab Setup Guide

7.1.3 Measurement Settings in ACQUA

In anechoic test environment, devices with headset or handset audio UI must be tested with a HATS.
Devices with speakerphone audio UI must be measured with artificial mouth and ½” free field
measurement microphone. Screen shots from ACQUA are shown in Figures 13, 14, 15.

Figure 13: Measurement Settings for devices with headset or

handset audio UI tested in an anechoic test environment

Figure 14: Measurement Settings for devices with speakerphone

audio UI tested in an anechoic test environment

Page 23 of 53
 Skype Audio Lab Setup Guide

In a reverberant test environment, HATS is always used and the same measurement settings apply for all
audio UIs.

Figure 15: Measurement Settings for testing in a reverberant test environment

7.1.4 Calibration Assignments in ACQUA

The calibration assignments are dependent on the input/output equipment used. The generic options
are the following:

 HATS ears and mouth.

 Artificial mouth and ½” measurement microphone used for testing devices with speakerphone
audio UI in an anechoic room.
 HATS mouth and ¼” measurement microphone used for 3QUEST tests in a reverberant test
environment.

If there are multiple articles of any of the above are used, then a separate calibration assignment needs
to be done for every pair of input/output used.
For example, if one HATS is used in an anechoic test environment and another in a reverberant test
environment, it means that different calibration assignments must be used.

Page 24 of 53
 Skype Audio Lab Setup Guide

Figure 16: Calibration Assignment for testing with HATS

Figure 17: Calibration Assignment for testing devices with

speakerphone audio UI in an anechoic test environment

Page 25 of 53
 Skype Audio Lab Setup Guide

Figure 18: Calibration Assignment for 3QUEST tests

7.1.5 MFE VI.I “Skype” settings in an anechoic test environment

In an anechoic test environment, devices with headset or handset audio UI must be tested with a HATS.
Devices with speakerphone audio UI must be measured with an artificial mouth and a ½” free field
measurement microphone. Screen shots from ACQUA are shown in Figures 19 and 20.

Figure 19: MFE VI.I Settings for testing with HATS

Page 26 of 53
 Skype Audio Lab Setup Guide

Figure 20: MFE VI.I Settings for testing with artificial mouth and
½” measurement microphone

The commented snapshots above indicate also which physical connectors should be used on MFE VI.I.
As the same ones are used for both then it means that the equipment needs to be physically
reconnected when needed.

Using incorrect calibration assignments might ruin output equipment.

7.1.6 MFE VI.I “Skype” settings in a reverberant test environment

In reverberant test environment, HATS is used. However, there are two hardware configurations used:
 AEC tests – HATS is used
 3QUEST tests – instead of HATS ear a ¼” measurement microphone is used

This means that while executing the sequence the wiring, and also ACQUA settings, have to be
reconfigured.

Figure 21: MFE VI.I Settings for AEC tests

Page 27 of 53
 Skype Audio Lab Setup Guide

Figure 22: MFE VI.I Settings for 3QUEST tests

Page 28 of 53
 Skype Audio Lab Setup Guide

8.0 Setting up the DUT

The DUT has to be positioned according to the latest available Skype & Lync Audio Test Specification.
The positioning according to specification version 1.0 is summarized in the following sections:
 Annex-D: How to position DUT in an anechoic test environment
 Annex-E: How to position DUT in a reverberant test environment

This section is mainly instructing about how to configure the DUT to use the correct processing, and
provides some tips about using the Skype audio tests in ACQUA 3.2.

8.1 Controlling the audio processing

By default, Skype uses the raw audio pin on any device running Windows 8.1 (or newer). This might
differ on some other platforms — on some, only custom processing is used and on others the custom
and Skype audio processing may be cascaded.

For testing purposes, it is critical to understand what is tested. The Skype & Lync Audio Test
Specifications have different requirements based on if the raw pin (and Skype processing is used) or
custom processing is used.

8.1.1 Using DUT Editor to test raw mode

The DUT Editor is provided by HEAD acoustics bundled with the Skype package. It is supported only on
Windows platforms. On other platforms, the XML commands must be used as described in the section,
Using XML keys to control the audio processing.

The DUT Editor can be run in master/subordinate mode on REF PC and peered over TCP with DUTagent
runtime. This enables remote controlling of the audio settings. This means the anechoic chamber does
not have to be opened or monitor/mouse extensions used every time there is a need to reconfigure the
settings.

The Skype test sequence in ACQUA is invoking instructions via Info SMD-s (pop ups during test) if there
is a need to reconfigure Skype settings.

8.1.2 Using XML keys to control the audio processing

The exact commands and instructions are provided in a separate white paper, Skype Audio
Configuration Options, available upon request from the Skype Media Hardware Certification team.

8.2 Executing the test sequence in ACQUA

8.2.1 Choosing correct specification and ACQUA settings
The ACQUA Skype package contains four databases:
 Unified Specification – Headset/Handset – Custom Processing
 Unified Specification – Headset/Handset – Skype Processing
 Unified Specification – Speakerphone – Custom Processing
 Unified Specification – Speakerphone – Skype Processing

The database name reflects audio UI and Skype & Lync Audio Test Specification.

Page 29 of 53
 Skype Audio Lab Setup Guide

After choosing the database, the next step is to choose the correct Measurement Settings, Calibration
Assignments, and MFE VI.I “Skype” settings. More guidance about how to prepare or choose these
settings is available in Section 7.0, Setting up MFE VI.I for testing.

8.2.2 Preparation sequence

Set up the solution in anechoic test environment. Make a Skype call from DUT to REF PC.

When running the preparation sequence, it is advised to keep the anechoic room door open. This way
the tester can hear if DUT and measurement equipment work as intended. In addition, it enables easier
access to the device in case some settings have to be changed before running the full sequence.

Figure 23: Preparation sequence

The Receive level might need adjustment from DUT physical volume buttons or the operating system’s
master volume. Select the category specific Receive (speaker) loudness test and verify that the correct
level is achieved. Iterate by adjusting the level and measuring this SMD until it fits in the range. In case
of multiple settings, fit and then use the highest value fitting the limits.
If the device does not provide enough playback level even when all settings are in maximum, run the
tests with maximum available playback volume.

Once preparation is done, it is safe to close the door to the anechoic room and start the tests.

8.2.3 Choosing the correct tests

In ACQUA 3.2, new filtering tools are available. The Unified Specification in ACQUA is built so that it has
two levels of tags:
 Differentiating audio UIs – this is used when selecting the tests to be executed
 Differentiating standard and premium level requirements – this typically is used for reporting

Figure 24: Filtering in ACQUA 3.2

Page 30 of 53
 Skype Audio Lab Setup Guide

Headset/Handset database tags Speakerphone database tags

headset_mono handheld
headset_stereo laptop
handset personal
group
long_range
room_system

Execution of the tests in an anechoic room is sectioned as follows:

1. Prepare the solution for testing.

2. Filter out category-specific tests.
3. Start the testing from Send path tests. Follow the instructions invoked by ACQUA. Continue with
Receive path and Echo path.

Figure 25: Send path, Receive path, and Echo path

folder in Unified Specification

Page 31 of 53
 Skype Audio Lab Setup Guide

4. Run the frequency response tests as instructed by the notifications in ACQUA. The frequency
response tests assume that the measurement equipment or DUT must be repositioned. The final
result of the test is an average of the three measurements.

Figure 26: Frequency response tests in Unified Specification

5. Set the playback level to maximum, and run the EQUEST test.

Figure 27: EQUEST tests in Unified Specification

Page 32 of 53
 Skype Audio Lab Setup Guide

If reverberant room tests need to be done, then the steps would be the following:

Run AEC testing:

1. Set up the device in reverberant test environment. Use the same nominal playback level as used
in anechoic room.
2. Connect HATS to MFE VI.I, if necessary, and then change the Measurement Settings, Calibration
Assignment, and MFE VI.I Settings “Skype” in ACQUA according to the setup.
3. Select and run the tests. Note that if the playback level is not set to maximum then EQUEST
tests with maximum playback level must be done separately as in anechoic room.

Figure 28: Reverberant room AEC tests in Unified Specification

Page 33 of 53
 Skype Audio Lab Setup Guide

Run 3QUEST tests:

1. Connect ¼” measurement microphone to MFE VI.I.

2. Change the Measurement Settings, Calibration Assignment, and MFE VI.I Settings “Skype” in
ACQUA according to the setup.
3. Make sure that HEADAutoEQ software has been started on HAE-BGN PC.
4. Select and run the applicable 3QUEST tests.

Figure 29: Reverberant room 3QUEST tests in Unified Specification

Page 34 of 53
 Skype Audio Lab Setup Guide

9.0 Annex-A: How to import a database

Sometimes there is a need to import a custom made database. This is typically done using backup files
(with the .bak file extension).

In ACQUA it is can be done the following way:

1. Close ACQUA
2. Start the ACQUA Database Administrator.
3. Click Restore Database.
4. Choose the location of backup file.
5. Choose where the local database (with file extension .mdf) will be created.
6. Click Restore.
7. Close ACQUA Database Administrator.
8. Start ACQUA.
The imported database (and the projects it contains) should appear in the available databases.

Page 35 of 53
 Skype Audio Lab Setup Guide

10.0 Annex-B: Step by step guide how to recalibrate a G.R.A.S. ¼”

measurement microphone using B&K pistonphone
In ACQUA it can be done the following way:
1. On the ACQUA Preparations drop-down menu, choose Input calibration.

Figure 30: Input calibration - ACQUA Preparations drop-down menu

2. Choose Acoustical, and then click Next.

Figure 31: Input calibration - Choose calibration type

3. Choose the device from the list.

If adding a new device, click New.

Figure 32: Input calibration - Choose or create a new calibration

Page 36 of 53
 Skype Audio Lab Setup Guide

4. Enter calibrator’s nominal output level.

Figure 33: Input calibration - Insert the microphone air tightly to pistonphone
and enter the nominal value to ACQUA

5. Choose where the microphone is connected. The recommended connection is to MFE VI.I
channel number 2.

Figure 34: Input calibration - Choose how the microphone is connected

Page 37 of 53
 Skype Audio Lab Setup Guide

6. Enter calibration duration. The recommended value is 1000 ms.

Figure 35: Input calibration - Enter calibration duration

7. Verify the parameters set are correct, and then start the calibration.

Figure 36: Input calibration - Start

Page 38 of 53
 Skype Audio Lab Setup Guide

11.0 Annex-C: Step-by-step guide for how to recalibrate G.R.A.S. mouth

In ACQUA it can be done the following way:
1. On the ACQUA drop-down menu, select Preparations → Input calibration…

Figure 37: Output calibration - ACQUA Preparations drop-down menu

2. Choose the device from list.

If adding a new device, click New.

Figure 38: Output calibration – Choose device

3. Choose measurement device.

If the correct device is not in the list, then that must be calibrated first.

Figure 39: Output calibration - Choose calibrator

Page 39 of 53
 Skype Audio Lab Setup Guide

4. Choose where the output device is connected. The recommended connection is to MFE VI.

Figure 40: Output calibration – Choosing frontend

5. Choose the channel where the input device (measurement microphone) is connected.

Figure 41: Output calibration – Choosing MFE VI.I channel

where measurement microphone is connected

Page 40 of 53
 Skype Audio Lab Setup Guide

3. Choose the channel where the output device (mouth) is connected.

Figure 42: Output calibration - Choosing MFE VI.I channel

where mouth is connected

4. Choose the band where the artificial mouth is to be calibrated.

Figure 43: Output calibration – Choosing band for calibration

Page 41 of 53
 Skype Audio Lab Setup Guide

5. Mount the measurement microphone in front of the artificial mouth following the vendor’s
instructions.

Figure 44: Output calibration – Mounting microphone

6. Verify the parameters set are correct, and then start the calibration.

Figure 45: Output calibration - Start

Page 42 of 53
 Skype Audio Lab Setup Guide

12.0 Annex-D: How to position DUT in an anechoic test environment

12.1.1 Device with headset audio UI in an anechoic test environment
The headset will be placed on HATS and adjusted to resemble the positioning on a real human as closely
as possible.

Figure 46: Headset positioning on HATS

12.1.2 Device with handset audio UI in an anechoic test environment

If the manufacturer/vendor provides guidelines as to how the handset should be tested by providing
standard handset positioning coordinates as defined in ITU-T P.64, Annex E, the audio test engineer will
take such recommendations into consideration.
If not instructed otherwise, the measurements shall be conducted with an application force of 8N ±2N
between the phone and the right ear.

Figure 47: Putting handset in holder

Figure 48: Positioning the handset on HATS

Page 43 of 53
 Skype Audio Lab Setup Guide

12.1.3 Device with handheld speakerphone audio UI in an anechoic test environment

Figure 49: Positioning handheld device

Figure 50: Positioning tablet with a leg using an artificial mouth

Page 44 of 53
 Skype Audio Lab Setup Guide

12.1.4 Device with personal speakerphone audio UI in an anechoic test environment

Figure 51: Positioning a USB accessory using an artificial mouth

Figure 52: Positioning a standalone webcam using an artificial mouth

Page 45 of 53
 Skype Audio Lab Setup Guide

Figure 53: Positioning a laptop using an artificial mouth

12.1.5 Device with group speakerphone audio UI in an anechoic test environment

Figure 54: Positioning a group speakerphone with using artificial mouth

Page 46 of 53
 Skype Audio Lab Setup Guide

Figure 55: Positioning a group speakerphone with satellite microphones

using artificial mouth

12.1.6 Device with long-range speakerphone audio UI in an anechoic test environment

Figure 56: Positioning a long-range speakerphone with using artificial mouth

Page 47 of 53
 Skype Audio Lab Setup Guide

13.0 Annex-E: How to position DUT in a reverberant test environment

The principles of setting up the DUT are the following:

 HATS is used for measurements.

 DUT has to be set up in a reverberant room where HATS was during calibration.
 Unless stated otherwise, the same nominal playback level should be used as in an anechoic test
environment.
 If the headset or handset audio UI is tested, then the DUT has to be set up on HATS as described
in Annex-D: How to position DUT in an anechoic test environment
 If the speakerphone audio UI is tested, then the DUT has to be set up in a manner that simulates
the position of the artificial mouth as described in Annex-D: How to position DUT in an anechoic
test environment.
Echo path tests on long-range speakerphones is an exception. In this case, the DUT is set up in a
realistic condition instead of simulating the levels.

Figures 57 through 62 describe the positioning of DUT in a reverberant test environment.

13.1.1 Device with headset or handset audio UI in a reverberant test environment

Figure 57: Background noise test setup for headsets and handsets

Page 48 of 53
 Skype Audio Lab Setup Guide

13.1.2 Device with handheld speakerphone audio UI in a reverberant test environment

Figure 58: Background noise test setup for handheld speakerphones

Page 49 of 53
 Skype Audio Lab Setup Guide

13.1.3 Device with personal speakerphone audio UI in a reverberant test environment

Figure 59: Background noise test setup for personal speakerphones

Page 50 of 53
 Skype Audio Lab Setup Guide

13.1.4 Device with group speakerphone audio UI in a reverberant test environment

Figure 60: Background noise test setup for group speakerphones

Page 51 of 53
 Skype Audio Lab Setup Guide

13.1.5 Device with long-range speakerphone audio UI in a reverberant test environment

Figure 61: Background noise test setup for long-range speakerphones speech
quality in presence of noise testing

Page 52 of 53
 Skype Audio Lab Setup Guide

Figure 62: Test setup for long-range speakerphones, echo path testing

Page 53 of 53