Installing new firmware to your Realiser A16

Rev 1.98 Jul 24 2020

Updating your A16 with rev 1.98 JUL 24 2020 is only necessary if your A16’s firmware is older. The
current revision of your firmware is found in SETTINGS>UPDATES/ABOUT as described below in step 5. If
an update is required please begin with STEP 1.

Issues addressed in this update

(all hardware variants)

1) In all modes of operation, the USB stereo line output carries the user A headphone PCM signal.
Previously this PCM signal exhibited a 6 dB drop in level as well as an inter-channel misalignment of
2 samples compared to the user A headphone SPDIF digital output. This discrepancy has been
rectified. Please note that this USB stereo line output is only designed to work at a fixed USB
sampling rate of 48kHz.

2) The APM89L Atmos decoder included in revisions 1.91 and 1.92 was an experimental version (v2.3.2
8901 May) that did not allow wide speaker operation. This erroneous version has been replaced
with the correct version (v2.2.5 8901 July) in this update.

A16 Firmware update procedure

(all hardware variants)

STEP 1. The new firmware for the Realiser A16 is uploaded through the micro-SD card slot on the front
panel. First, obtain a micro-SD card (commonly 16 GB) and ensure it is formatted as FAT32. Second,
create a ‘realiser’ folder in the root directory and copy the firmware file FIRMA001.SVS into the realiser
folder. Insert this micro-SD card into the slot on the front of your A16.

STEP 2. Power up the A16 ensuring the power indicator LED is steady green. You can power it up using
the remote control or by momentarily depressing either User A or User B volume knobs. Now turn off
the A16 by pushing in and holding in the User A volume knob for at least 3 seconds. The LCD screen will
switch off and the power indicator LED will turn red. Release the User A volume knob.

STEP 3. Push in and hold in the User B volume knob and, simultaneously, push in and release the User A
volume knob. Then release the User B volume knob. The action of holding in B and depressing A
activates the firmware update manager as shown below. The power indicator LED will also be blinking
green.

STEP 4. Using the remote control, press the ENTER key twice to begin the firmware update. The A16 will
enter a long period (20-30 minutes) of authenticating the software, loading and rebooting. When the
unit first reboots it will begin updating the firmware for the individual hardware modules. Once the
firmware modules have been reprogrammed the unit will reboot using the normal power-up sequence
to the Speaker Map display for User A.

STEP 5. To confirm the firmware update was a success please check the revision numbers displayed in
UPDATES/ABOUT accessed via the SETTINGS page.

Confirm the A16 firmware revision is 1.98 JUL 24 2020, the FPGA firmware revision is 0.50 JUL 24 2020
and the head tracker firmware 1.20 Nov 08 2019. For APM89L hardware variants the APM firmware will
V2.2.5 8901 JUL. If your HT firmware shows an earlier date, then you should update the HT firmware
following completion of this update. See previous updates for instructions on how to undertake this.
 Updates/About page (APM89L variant)

STEP 6. The firmware update is now complete. A full restore must be run in order to properly initialize
new menu and Preset features First press BACK and then navigate to ‘Settings’ and press ENTER.

Then navigate to ‘Restore factory setup’ and press ENTER again.

Select ‘full restore’ and press Enter again.

The full restore will take approximately 10 minutes to complete, thereafter the A16 will automatically
return to the User A Speaker Map display.

Changes to existing features

(all hardware variants)

APM89L/APM110 decoder SVS HP audio level

Previously audio output from the AMP89L/APM110 decoder was first lowered by 9dB to ensure
sufficient headroom is available for all bass management modes of operation. In previous
firmware the decoder audio was input to the SVS headphone rendering algorithm without
modification leading to a 9dB level drop when comparing the same audio input via USB or Line.
As of rev 1.98, audio sourced from the decoder is now boosted by 9dB prior to SVS rendering
processing making the audio levels consistent across all inputs, as heard over the headphones.
Note that as a result of this change, when listening to audio tracks that are input via HDMI or
SPDIF, your headphone volume will need to be reduced by 9 dB (or 9 volume units) to bring the
loudness back to your regular listening level.

APM89L/APM110 decoder AV audio level

Unlike the SVS headphone mode, in AV line output mode the 9dB level drop implemented in the
APM decoder cannot be easily reversed. Hence the line out levels of Atmos/DTS/PCM decoded
audio remain 9dB lower compared to audio sourced from either the USB or line inputs. As of rev
1.98 a unity gain switch ‘APM 0dB’ for the APM89L/APM110 decoder has been added to the
MISC settings page. When this switch is enabled the APM decoder will output audio without any
attenuation. Use of this switch should be restricted to those applications where it is necessary to
switch between different inputs while in AV mode and have a consistent volume level between
them. Note that if APM 0dB’ is enabled the signal level at the line outputs in AV mode will be
9dB higher. Please ensure you lower your amplifier volume by 9dB to bring the loudness back to
your regular listening level prior to changing enabling APM 0dB.
Note that the APM 0dB switch requires the user A or user B Preset be reloaded before taking
effect.

APM89L/110 -9dB attenuation override

Tri-level volume now uses offset instead of absolute values

Previously the tri-level volume feature involved logging three different volume levels and then
jumping between these three levels using the remote volume buttons. As of rev 1.98 this tri-
level scheme has been replaced by one that jumps between the current volume (or reference
level) and two other volumes (or reference levels) that are offset from the current volume
(reference) and has been renamed to Tri-ref. The new scheme is described in the new features
section below.

Adjusting the Volume

Previously when changing the headphone volume using either the remote control or the volume
knobs, the first keystroke or rotary step only displayed the current volume level and required
subsequent keystrokes/steps to effect a change in volume. As of rev 1.98 the volume is now
changed on the first keystroke/step. Note that each +/- volume keystroke on the remote control
corresponds to +/- 1dB change in volume (or +/-1 volume units).

New Features
(all hardware variants)

Headphone Clip Attenuation Override

The A16 headphone volume control (user A and user B) is implemented inside the SVS rendering
DSPs using a digital gain control. A consequence of this approach is that it is relatively easy to
clip the PCM signal output to the headphone DACs (as well as the SPDIF transmitters and USB
stereo line out) when a high-volume setting coincides with high input signal levels. To mitigate
this potential problem the A16 steadily reduces the digital gain (volume) each time a clip event
is detected, thereby ensuring the listener is not subjected to sustained clipping distortion. In
practice this means that the A16 will automatically reduce the dialled in volume during loud
portions of the sound track until clipping ceases. The problem can normally be avoided either by
moving the A16 gain slider (L-M-H) on the front panel to a higher setting when the headphone is
directly by the A16, or by increasing the headphone amplifier gain when using an external
headphone amplifier. Previously any automatic reduction in volume setting was not reflected on
the A16 display until an attempt to adjust the volume was made by the user. As of rev 1.98, the
volume shown in the top right-hand corner of the speaker rendering page will now match the
actual gain in real time.

Real time volume display

Nonetheless there are times when it is useful to not have the system automatically attenuate
the gain on clipping. As of rev 1.98 a ‘Limiter off’ switch has been added to the AUDIO Settings.
When enabled, the digital gain will always remain at that set by the volume, for both user A and
user B and for both 16-ch and 24-ch modes of operation.

Note that the ‘Limiter off’ switch is implemented in real time and does not require the reloading
of Presets.

Attenuation override switch in AUDIO settings

Balanced Headphone Output Mode

As of rev 1.98 the user A headphone and user B headphone jacks can be configured to work as a
single balanced (differential) drive for user A using the Bal-HP switch in the AUDIO settings page.
In this case the user A headphone jack (or the user A rear RCA jacks) carries the balanced signals
for the left ear and the user B headphone jack (or the user B rear RCA jacks) carries the balanced
signals for the right ear (see schematic below). Only headphones that are wired with a separate
return wire to each driver can make use if this mode. Typically a custom balanced headphone
cable will be required.

Balanced headphone switch in AUDIO settings

The balanced drive operates in both 24ch and 16ch SVS modes and all HPEQ routines, and user
A headphone signal continues to be available at the SPDIF output. However, in 16ch mode the
user B headphone audio can only be heard through the user B SPDIF output. In balanced output
mode the voltage across the headphone drivers is twice that of the unbalanced mode and, as a
result, will sound twice as loud for the same volume setting. Please ensure your headphone
volume is reduced appropriately before engaging this mode. Please also ensure that the user A
and user B gain sliders on the front panel are set to the same position. Moreover, we also
recommend using the lowest slider gain settings for both until you become more familiar with
this mode of operation.

Note that the ‘Bal-HP’ switch is implemented in real time and does not require the reloading of
Presets.
Tone Generation

To aid decoder speaker identification rev 1.98 implements the tone generator feature of the
APM89L/APM110 decoder subassembly. The tone generator is controlled by the ‘Tone Gen’
selector under MISC settings page. Having the ability to substitute decoder audio with known
tones is particularly useful for verifying high order decoder modes of operation. Presently the
tone generator will only function when an HDMI input is receiving active content and will mute
anytime this HDMI stream is interrupted. Hence it is necessary to select HDMI(1-4) before
enabling this feature and then play any content with a 48kHz soundtrack to this same HDMI
input to active the tones.

The tone generator function has two modes of operation.

1) ID Tones generates tones with a unique frequency for each listening room speaker and
outputs these tones at a level of -20dBFS to both SVS headphone rendering DSPs when
using headphones or to the line outputs when in AV mode. The following ID frequencies will
be present on active channels.
Center: 500 Hz.
Left: 1000 Hz.
Right: 1500 Hz.
Left Sur: 2000 Hz.
Right Sur: 2500 Hz.
Left Back: 3000 Hz.
Right Back: 3500 Hz.
Left Wide: 4000 Hz.
Right Wide: 4500 Hz.
Left Top Front: 5000 Hz.
Right Top Front: 5500 Hz.
Left Top Middle: 6000 Hz.
Right Top Middle: 6500 Hz.
Left Top Rear: 7000 Hz.
Right Top Rear: 7500 Hz.
Left High Front: 8000 Hz.
Right High Front: 8500 Hz.
Left High Rear: 9000 Hz.
Right High Rear: 9500 Hz.
SW: 125.00 Hz.
SW2: 62.50 Hz.
Left Rear Surround 1: 10000 Hz.
Right Rear Surround 1: 10500 Hz.
Left Rear Surround 2: 11000 Hz.
Right Rear Surround 2: 11500 Hz.
Left Screen: 12000 Hz.
Right Screen: 12500 Hz.
Left Surround 1: 13000 Hz.
Right Surround 1: 13500 Hz.
Left Cntr Surround: 14000 Hz.
Right Cntr Surround: 14500 Hz.
Center Surround: 15000 Hz.

ID tone mode
2) 1kHz Tones generates a single 1kHz sinewave tone at a level of -20dBFS and outputs this
tone to each active channel in either SVS headphone rendering DSPs when using
headphones or to the line outputs when in AV mode.

1kHz Tone mode

Note that the tone generator function runs only on the APM89L/APM110 decoder boards and
therefore cannot be used to verify listening rooms that input their audio via the USB or Line
inputs. Note also that for the Tone Gen function to work it is necessary to reload the user A
Preset after the function is selected in the MISC settings page. Likewise, to turn off the tone
generator it is necessary to return to the MISC settings page, turn Tone Gen off and then reload
user A Preset.

Pink Noise or Music playback in TEST

The TEST key is used to activate a built-in audio loop to aid verification and debugging.
Previously only a music loop was available. As of rev 1.98 it is now possible to select between
the music loop and a pink noise loop (-20dBFS) using a selector in the MISC settings page.

TEST loop selection

Note that this selector is implemented in real time and does not require Presets to be reloaded.
However, there is a 10 second pause on changing the test selection as the new test signal is
uploaded to the rendering DSPs.

Head Tracker Disable

Previously to stop head tracking in the A16 it was necessary to unplug the head tracker. If the
head tracker is unplugged while running it is possible for the heading angle fed to the SVS DSPs
to freeze at the last recorded angle and not fall back to zero. As of rev 1.98 a new Disable HT
switch has been added to the MEASUREMENT settings page to ensure the tracker heading is
properly held at zero (looking straight ahead and level). Disable HT should be enabled if the user
does not intend to use head tracking whilst listening over headphones. Note that this switch is
implemented in real time and does not require Presets to be reloaded.

Head Tracker disable in MEASUREMENT

settings

Azimuth Angle Offset

It is possible for the virtual loudspeaker layout to appear slightly rotated as a result of errors in
the look angles taken up during a PRIR measurement, or because the head tracker mounted on
the headphones is slightly twisted. As of rev 1.98 a new HT azi offset entry has been added to
the MEASUREMENT settings page with a range of +/- 10 degrees. If, for example, the
soundstage exhibits a +5 degree rotation, then the HT azi offset would be set to -5 degrees in
order to cancel out the apparent rotation. Presently the HT azi offset value will offset both user
A and user B sound rooms at the same time. Note that the offset value is fed to the rendering
DSPs in real time. Hence it is possible to listen over the headphones while altering the offset
angle. Note also that the offset angle is independent of the HT function and is effective even if
the HT has been disabled or disconnected.
 Azimuth angle offset in MEASUREMENT
settings

Reference Level Calibration (16-ch SVS mode only)

Movies soundtracks are commonly created in dubbing stages and mixing studios that are
calibrated to a particular loudness or reference level. By playing back a movie soundtrack at the
same reference level, the listener can replicate the sonic experience that would have been
intended by the director during the production. Soundtracks that are destined for movie theatre
playback are typically mixed at a reference level of 85dB SPL (measured at the sweet-spot while
playing a -20dB pink noise signal through a single main speaker). Lower reference levels may be
used for soundtracks targeted for home playback. For example, 79dB SPL and 76dB SPL
reference levels are common for home movie and music playback. As of rev 1.98 it is now
possible to calibrate the binaural sound pressure levels (SPL) as delivered by your headphones
for any particular virtual listening room and to then set the listening levels referenced to this
calibrated level. In this case the volume units are replaced by the reference level in dB SPL and
the desired loudness experience is set by adjusting the playback reference level. For example,
when you visit a cinema the playback reference level you will hear is typically between 79dB and
85dB SPL. By calibrating the A16 virtual listening room and headphone combination, this same
playback level can be dialled in with ease allowing the listener to acoustically replicate what
they experience at the movies with a high degree of confidence.
Theory of Operation

A reference level is the sound pressure level (SPL) of a full-band pink noise signal played out the
left front or right front loudspeaker 20dB down from peak level, measured at the listening
position. Typically, the SPL meter is set to filter the incoming microphone signal using a C filter
and the calculated SPL value averaged using the slow setting. Hence to calibrate the playback
volume to a reference level of 85dB, for example, one simply needs to play the pink noise signal
from a DVD or Blu-ray player and adjust the speaker volume until the SPL meter reads 85dB
SPL(C) at your listening position. From that point on, when listening to movie soundtracks, one
just brings the volume back to the same position in order to listen at the 85dB reference level.

Setting up the Reference Level using a simple comparison

Setting the reference level in the A16 is essentially the same process. The only difference is that
the listening room is a virtual experience in a headphone. Regular SPL meters cannot easily
measure the volume levels in your ears. However, one way of overcoming this problem is to use
a simple comparative procedure. In this case one sets up a real loudspeaker in a real room and
sets its volume to the desired reference level using a pink noise test signal and the SPL meter
positioned in the sweet-spot. By sending the same pink noise signal into the A16, and selecting
the left front loudspeaker in the headphone, a reference volume setting in the A16 is obtained
when the pink noise intensity heard over the headphones is the same as what is heard directly
from the speaker. Once you have this A16 volume value it can be entered into the Reference
Level Management page of the Preset that holds the listening room used in the comparison.

Reference level management for Presets

For user A Presets the Ref Level Management entry point is found on the second page of the
Preset. For user B Presets it is found in the main Preset page. On leaving the factory (or after a
factory restore) the reference levels are set as shown below. The first three lines define the
reference levels for the Atmos, DTS and PCM rooms respectively. The fourth line defines the tri-
ref function to be explained later.

Default Preset Reference level setup

Each room reference level can be enabled or disabled. When enabled the A16 will use the Ref
SPL volume value and the real time volume indicator will change from ‘V’ to ‘R’ and the volume
screen will use REFERENCE as opposed to VOLUME as shown below. Note that these settings do
not take effect until the Preset is reloaded.
 Example vol screen when Ref SPL active

Assuming in the example of the sound room comparison, the reference level was measured at
85dB on the SPL meter and the volume of the left speaker of a virtual PCM room, loaded to the
SVS renderer, matched the intensity of the real speaker when set to 78. Then the Ref SPL and
Vol entries for the PCM Ref would be as follows.

Using reference level SPL for PCM room

By enabling the PCM Ref switch, the A16 will display the volume in reference SPL as opposed to
volume units. However, the Preset must be reloaded for this to occur. Using a comparison
method is easy to understand and relatively easy to undertake. However, unless the virtual PCM
room is a measured copy (PRIR) of the real listening room you are using to make the
comparison, the final reference level may only be accurate to within 2 to 3 dB since the spatial
and tonal difference between them will make it difficult to achieve anything closer.

Setting up the Reference Level using the Cal SPL Method

Rather than subjectively comparing the loudness of a virtual speaker to a real speaker, the Cal
SPL method instead makes use of the SVS microphones to measure the loudness level in the
listeners ears while wearing their headphones. Since all SVS microphones exhibit the same
sensitivity to within +/-1dB, their voltage levels can be fairly accurately mapped to actual SPL. A
binaural SPL metering algorithm is run in real time within the A16 itself, with the microphones
placed in the listeners ears. For Cal SPL the user loads the virtual listening room to be calibrated
(running on DSP A), feeds a pink noise signal into the virtual listening room, and then measures
the pink noise level at the headphones by analysing the SVS microphone signals using the
binaural SPL metering algorithm (running on DSP B). In the example below a 9.1.6ch PCM
listening room will be subject to the reference level calibration procedure.

Example 9.1.6ch PCM listening room

The user must then decide the pink noise source. Either an external signal is to be input via the
LINE inputs (in this case) or the pink noise can be generated internally by enabling the pink noise
SPL Gen in the MEASUREMENT settings page. In this example we will use the internal generator.

Enabling SPL pink noise

Finally, the user should press the PA key to display the speaker rendering page and then lower
the A16 user A headphone volume to a safe level. Once the volume has been lowered the SPL
calibration is started by pressing the CAL key.

Cal SPL page

The SPL calibration routine always starts up with the front left virtual speaker in solo. Since in
this example the pink noise is internally generated, the L-R headphone level meters immediately
indicate the peak PCM signal level (green bars) heard through the headphones, and this level
will rise and fall as the user A volume is adjusted in the normal way. Other speakers can be
selected using the solo buttons on the remote control, but never more than one. Next the user
inserts the SVS microphones into their ear canals (plugged into the A16 Mics jacks on the front
panel) and then dons their headphones (plugged into user A HP jack), with the microphones still
inserted. Since the headphone sensitivity is unknown it is recommended that the user A volume
be first reduced by 30 units and then the user A switched gain (L-M-H) on the front panel set to
H. By increasing the analogue gain in this way, the PCM headroom within the SVS rendering
DSPs is maximised. If appropriate the HP balanced mode could also be deployed to further boost
the analogue gain.

In the example below the user wishes to calibrate the reference level by finding the user A
volume setting (using the remote control or using the volume knob) that brings the SVS
microphone SPL levels to 79dB. In this example this is achieved with the volume set to 48 (V48 in
the top right corner). The SPL measurement running in DSP B averages over a 10 second window
so it is necessary to allow volume changes to take full effect before making further adjustments.

Reference level of 79dB SPL @V48

As with the comparative procedure described earlier, the SPL value of 79 and the volume value
of 48 may be entered manually into the Ref Level Management entry page of the Preset used to
load the listening room under test, or indeed any Preset that uses the same room. Note that the
Preset must be reloaded for these to take effect. Alternatively, the values can be entered
automatically by depressing the user B volume knob. In this case the PCM Ref switch is also
enabled causing the A16 volume to switch to reference dB SPL. Whilst the PCM Ref switch
remains active, each time the A16 loads that room from that particular Preset, the A16 will
switch to reference volume units. Disabling the switch will cause the system to revert back to
regular volume units.

To exit the CAL SPL routine, press the CAL key again and the A16 will exit and then reload both
user A and user B Presets.
 Depressing vol B knob switches to Ref SPL

Note that Cal SPL does not reference directly listening rooms in user B Presets. To calibrate a
listening room intended for use on the B side, first create a user A Preset that references the
listening room. Then run Cal SPL on that room and take a note the final Ref SPL and Vol
numbers. Then enter these manually in the Ref Level Management entry page of the user B
Preset that contains the same listening room.

Calibrating an Atmos Room

Since the SPL calibration routine operates on the listening room currently loaded in the user A
Preset, then to calibrate an Atmos listening room, one must ensure it is loaded prior to
activating Cal SPL. The easiest way to force an Atmos room load before activating Cal SPL is to
select an HDMI input as the audio source, set the HDMI PCM upmixer to Dolby Surr in PCM
Management and then reload the user A Preset.

Estimation Reference Level PCM Headroom

For any reference level it is critical to know whether the combined signal level of the virtual
listening room sound stage at that volume level will have sufficient digital headroom as the
input signals approach peak. To help with this calculation a headroom estimation algorithm is
used to display yellow headroom bars that ride atop the green headphone peak level bars.

Headroom estimation at 85dB SPL

In the example above the Cal SPL routine is presently estimating the headphone PCM headroom
necessary to cope with peak incoherent signals on all inputs of a 9.1.6ch room set to a reference
listening level of 85dB SPL. The peak headphone levels are shown in green. The L-ch is peaking
around -36dBFS while the R-ch is peaking around -40dBFS. The yellow bars indicate the increase
in level that would occur if all 16 channels were fed independent peak pink noise with all
speakers set to 0dB gain, except the SW which is set at +10dB.

The estimation calculation uses the number of speakers, the SPL Headroom and the SPL SW
Loss. Both SPL Headroom and SPL SW Loss can be adjusted from the Measurement Setting page
as shown below. SPL Headroom is simply the maximum peak input signal above reference level.
Typically, this is +20dB. SPL SW Loss is the drop in measured SPL when switching from the main
speakers to a subwoofer using a C filter (without the +10dB boost). Typically, a 80Hz subwoofer
exhibits a drop of 5 to 6dB. In Cal SPL a SW reference level of 85-5+10=90dB SPL is used in the
estimator. Note that the headroom estimator does not take into account Bass Management or
Direct Bass processing but in most circumstances neither of these are likely to exceed a direct
LFE-SW signal level.

In this example the routine estimates that the peak level could increase by a further 32dB giving
at maximum binaural SPL of 117dB and leaving a margin of 4dB before the onset of clipping. In
other words, we expect that typical dynamic excursions away from reference level, as might
occur in action movies, should not clip the headphone PCM signal. When the estimation routine
determines that insufficient digital headroom is available to guarantee clip free playback, the
yellow bars are replaced by red bars to act as a warning. In such cases it would be necessary
either to use more sensitive headphones or increase the analogue gain driving the headphones
(L-M-H gain switch or external HP amplifier) or use a room with fewer speakers or reduce the
reference listening level. Another option would be to accept the possibility of clipping and to
consider enabling the automatic clip attenuation. In this way the A16 could take action to
reduce the listening levels if clipping does eventually occur.

Digital headroom parameters

Note also that the PCM headroom calculation applies not only to the A16s internal headphone
DACs but also to both SPDIF HP outputs and the USB stereo line out (in the case of user A). Note
also that you must recalibrate the reference level if either the headphone type or the gain
position on the A16 front panel slider switch are different to those used during the original
calibration.

Reference Levels and External Headphone Amplifiers

For the reference level and headroom estimation to make sense, all volume adjustments must
use the A16 volume controls (volume knob or remote control), both for the measurement itself
and all subsequent use. An external headphone amplifier is simply external gain and its volume
setting forms part of the calibrated chain and therefore must remain fixed. Ideally the volume of
the external amplifier would be left at maximum. This ensures the amplifier itself does not run
out of steam due to lack of gain, but this may be excessive especially for low impedance
headphones or those with a high sensitivity. It all depends on the parameters of the amp-
headphone combination and would require some experimentation to find the best setting.

Steps to calibrate a PCM Listening Room reference level using internal pink noise

1) Set the audio source to LINE, USB or HDMI (playing a PCM track)
2) Plug in headphones and SVS microphones
3) Create a Preset for user A that assigns the PCM listening room required to be calibrated
4) In that Preset assign a HPEQ file for the headphone you intend to use
5) Ensure AV mode is disabled
6) Load the user A Preset just created
7) Ensure TEST mode is not active
8) Ensure the A16 is running in 16ch SVS mode
9) Navigate to SETTINGS->SYSTEM->MEASUREMENT SETTINGS and enable SPL Gen
10) In the same page set SPL headroom to 20 and SPL SW loss to 5
11) Return to the home page
12) Press PA key to display the speaker render page for user A
13) Press the CAL key and the SPL calibration screen will load after approximately 5 seconds
14) Initially reduce the headphone volume to a safe level
15) Insert SVS microphones and don headphones without disturbing the mics
16) Adjust the user A volume slowly until the desired binaural SPL is reached
17) Wait at least 10 seconds to let the meter settle between each adjustment
18) Observe the headphone headroom for any final SPL setting
19) If happy depress user B volume knob to log the reference SPL and volume numbers.
20) Press CAL key to exit

Headphone SPL Headroom

Although we now have some idea of the likely headphone voltage excursions the remaining
issue is whether the headphone (or headphone plus external amplifier) in use has sufficient
headroom to cope with such excursions and what levels of distortion are likely to accompany
them if they do. This is subject to further study.

Using only the Reference level SPL measurement

The binaural SPL measurement can, on its own, be useful where there is a need to estimate the
SPL of a headphone not being driven by the A16 but from some other audio chain. For example,
by playing -20dB pink noise through a headphone connected to a PC or a phone, the SPL levels
are easily determined by activating the Cal SPL routine in conjunction with the SVS microphones.
In this case the digital headphone headroom estimator should be ignored.

Using only the Headroom Estimator

Conversely, the headphone PCM signal headroom estimator can, on its own, be useful just to
predict the available headroom for any particular A16 volume level. In this case neither SVS
microphones nor the headphones are required and the binaural SPL number can be ignored. The
user simply loads the listening room using a user A Preset, enables the SPL pink noise function
and activates Cal SPL. For any desired volume, if yellow bars are displayed above the green
headphone peak level meters then sufficient headroom exists. If red bars are present then
insufficient headroom exists and clipping is possible as the input levels approach peak
amplitude. If the headphone clip attenuation is enabled then the volume will automatically be
lowered if clipping does in fact occur. Headroom estimation is also possible using the metering
function described elsewhere in this document.

Vol=54, headroom sufficient

 Vol=64, headroom insufficient

Measuring binaural SPLs for Subwoofers

While running the Cal SPL routine the binaural SPL for the subwoofer can be measured by
pressing the remote LFE key. In theory, for a subwoofer (80Hz low-pass) set to the same gain as
the main speakers, the binaural SPL would be approximately 5dB below that for the main
speakers. For example, if the desired SPL was 85dB SPL then the SW should measure around
80dB SPL (for a subwoofer with a 10dB gain, then the SPL would measure at 90dB). However,
measuring SW SPL is somewhat inaccurate in practice, particularly using the apparatus the Cal
SPL routine must use. Pink noise is louder in the low frequencies than in the high frequencies.
This means that the SPL calculation is influenced more by the lows than the highs. As a result,
any low frequency roll-off in the measurement chain will progressively degrade the SPL
measurement accuracy as the bandwidth of the pink noise decreases. For example, a 20Hz roll-
off will reduce the accuracy of a full-band pink noise (20Hz-20kHz) measurement by less than
1dB whereas the same roll-off when measuring a Subwoofer (20Hz-80Hz) will introduce an
underestimation of almost 3dB. In our case the SVS microphones and the headphones are the
source of the roll-off and hence there is a risk that the SW gain will be set too high because of
this inherent underestimation. Our recommendation would be to calibrate the subwoofer using
an SPL meter before capturing a PRIR. In that way the SW levels will already be at the correct
level relative to the main speakers. If you must use the Cal SPL routine to set the SW speaker
gain, we suggest the following procedure.

1) Select the left front speaker and adjust the A16 volume to attain the desired reference level.
2) Select the SW speaker and take note of the binaural SPL.
3) Exit Cal SPL and navigate to Listening Rooms and locate the SW for the room you are
calibrating. Adjust the SW speaker gain (see below) such that the combined gain + binaural
SPL will be approximately 8dB below reference level, for a 0dB SW chain, or +2dB above the
reference level, for a +10dB SW chain. Let’s assume the +10dB SW boost will be applied to
the SW signal. Hence the SW speaker gain should be adjusted to produce a binaural SPL 8dB
below reference. For example, if the reference level is 85dB and the measured SW SPL was
78dB then the SW gain should be set to -1dB so that the final SW SPL will be 77dB.
4) Exit the Listening Room menu (the room you were editing should be reloaded automatically)
5) Re-enter the Cal SPL routine. Confirm the SW SPL is correct relative to a main speaker
reference SPL.

Adjusting the SW speaker gain

Please note that the SW speakers supplied in the factory PRIR files are already level matched to
the main factory speakers. If the LFE signal is to be boosted by +10dB then no gain changes are
required. Apply a 10dB gain to the SW speaker if the LFE will be fed directly to the SW without
the boost.

Reference Levels for 24-ch SVS mode

Because the SPL metering algorithm runs on the B-side DSP it is presently only possible to
conduct 16-ch reference level measurements. However, if the speakers that occupy channels 17
through to 24 are from the same listening room as those in channels 1-16, it is possible to simply
calibrate the first 16 channels and copy the calibration data to the 24ch room. The only error
that arises is that the estimated headroom will be underestimated by approximately 1.5dB for a
24ch room compared to the 16ch room.

Tri-Ref volume control

Previously a tri-volume function allowed three Preset volume settings to be stored for each
Preset. As of rev 1.98 this has been replaced by a tri-reference offset function linked to the room
reference level data described earlier. The room reference level data includes a Ref SPL and
corresponding VOL for each room type (Atmos/DTS/PCM). When Tri-Ref is enabled any attempt
to change the A16 volume (by knob or remote control) causes the volume to step either through
Ref SPL –> Ref SPL+Ref2 –> Ref SPL+Ref3 when operating in reference level mode, or through
Vol –> Vol+Ref2 –> Vol+Ref3 when not operating in reference level mode.
Tri-ref offsets of +3dB and -3dB

VOL

VOL + REF2

VOL + REF3
Additional Level Meter Features

The level meters have been modified to include the following features when operating in 16-ch
SVS mode;

1) The headphone volume (or reference level) can now be adjusted in real time.
2) The volume (or reference level) is now displayed in real time.
3) The headphone RMS (loudness) levels are superimposed onto the peak headphone meters
(black bars).
4) Additional headphone headroom and headphone SPL metering modes are described below.
5) Irrespective of the headphone metering mode, the tactile meters always use the same scale
as the input signals.

General Headphone headroom estimation metering in 16ch SVS mode

When reference levels are disabled for the loaded listening room, an estimated digital
headphone PCM headroom indicator (yellow bars) for the current headphone volume is
displayed when the pink noise TEST is active and the first channel of the listening room
is solo. When insufficient headroom exits, these bars turn red and remain just below the
clip segments. When the headroom exceeds 20dB the yellow bars are deactivated. The
estimation calculation uses the number of speakers, the SPL Headroom and the SPL SW
Loss and assumes 0dB gain for the main speakers and +10dB gain on the SW, as in the
Cal SPL routine.

Reference Level Headphone headroom estimation metering in 16ch SVS mode

When reference levels are active for the loaded listening room, an estimated peak
headphone headroom indicator (yellow bars) for the current headphone volume is
displayed. When insufficient headroom exits, these bars turn red and remain just below
the clip segments. When the headroom exceeds 20dB the yellow bars are deactivated.
By pressing the LEFT and RIGHT ARROW keys the main meters toggle between peak
headphone level and headphone SPL. However, the RMS levels (black bars) are not
displayed in the SPL mode. Note that for both modes the headroom (yellow bars) is
calculated directly from the calibrate reference level data for that room.
 Peak, RMS and headroom meters (pink noise)

Ref Level Peak, RMS and headroom meters

Ref Level SPL and headroom meters

A16 Internal Flash Memory Health Status

A small number of A16 users have reported sluggish behaviour from time to time. Our own
investigations suggest this is related to a slowing down of the internal flash memory used to
store user PRIRs, HPEQs, Listening Rooms and Presets. On some occasions this memory will stop
altogether when trying to access certain areas of the memory. In both cases, the A16 only
recovers when the firmware is reinstalled suggesting that the memory problem is cleared on
erasing and reprogramming. The A16 uses a SanDisk 16GB micro SD card internally for flash
memory storage and as a precaution we have recently switched to using a Micron 32GB
industrial micro SD card in all new builds. Industrial SD cards differ from consumer cards in that
they incorporate a separate memory protection and wear-levelling processor that runs in the
background and whose job is to maximise the lifetime of flash memory and to protect from read
and write corruption during power outages. The key advantage of industrial designs is that we
are able to read out flash health metrics from the card at any time. This therefore allows us to
better determine if the sluggish behaviour is caused by a reduction in flash health or by an, as
yet, undetected problem with our flash card driver.

A16 internal industrial micro SD card

As of rev 1.98 a flash memory health log is maintained for A16s which incorporate the industrial
SD card. The log is accessed from the Updates/About page. The log is updated each time the
Realiser is powered up, and each time the enter key is pressed when on the A16 internal flash
health page. The log records the date (day-month-year) for every 1% drop in health. As each
new record is logged the older data drops down the table and eventually falls off the end.

New internal micro SD card health log

We do not know for certain if an industrial SD card will solve the sluggish behaviour experienced
by some A16 users. Nonetheless, using an industrial card moving forward brings some certainty
as to the longevity of the A16 flash memory. Initially we will contact those users that have
previously reported the problem to see if they are willing to install the new card and hopefully
get some answers in the following months. If this proves positive then we will make the new
card available to anyone who wants to install it themselves. Note that replacing the internal
micro SD card is not as simple as changing it out. The replacement card must be pre-
programmed by the factory with certain data in order for your A16 to function correctly.

Known issues that are being worked on

1) The APM110 Atmos decoder is not decoding Dolby MAT streams correctly. MAT streaming is
commonly used in Apple TV and Sony PlayStation products when decoding Atmos content.
2) With some players the APM110 Atmos decoding does not always start up after changing
modes. It can be started manually by momentarily switching away from the current source
and then returning. For example, if using HDMI(1) then select HDMI(2) and then reselect
HDMI(1).
3) The reference SPL metering mode (as opposed to regular peak level metering mode) only
functions for user A at this time.
4) Some of the reference level features have not yet been incorporated into the IP command
server.

Imminent A16 Features (APM110/HSR41T hardware)

1) Add eARC audio source

2) HDMI AV bypass standby mode
3) Extend ALL/ASYNC PRIR measurements to 24 channels
4) Extend OH key group to 10 overhead/height speakers

New A16 Features in Development (all hardware variants)

1) DTS-X certification
2) Auro3D certification
3) Vertical Headtracking mode
4) Low delay gaming mode
5) Separate Denoise function for existing PRIR files
6) Allow Sound Rooms to be written-to/read-from the SD card (like PRIR files)
7) Extend optical stabilisation mode to User B head tracking
8) Headphone only mode (with HPEQ)
--------------------------------------------------------------- END --------------------------------------------------------------------