HTP-1 Volume-Control Findings¶

This document records the current understanding of the HTP-1 volume-control implementation. It is based on inspection of the avController source, related UI code, the Peak Monitor implementation, CS3318 data-sheet limits, hardware output-stage information, and observed behavior.

Treat this as an engineering reference for changes to volume-control code, UI labels, on-device help, and online documentation. Where a point remains unverified, it is marked as an open question rather than presented as fact.

Terminology¶

Master Volume¶

The main volume control.

The underlying Master Volume (/volume) range is:

-100 dB to +22 dB

Zero Point can shift the value shown to the user without changing this underlying range or the actual playback level.

Maximum Volume¶

Maximum Volume only clamps the user-accessible Master Volume range.

It does not:

move the internal 0 dB reference point;
change the gain calculation;
redefine Reference Output Voltage;
change actual gain below the cap.

Maximum Volume is stored on the underlying Master Volume scale. When Zero Point is non-zero, the value shown to the user is shifted by the same amount as the displayed Master Volume.

Power On Volume¶

Power On Volume defines the underlying Master Volume applied when the HTP-1 starts.

Its underlying range is:

-100 dB to 0 dB

The 0 dB upper limit is a safety ceiling and is independent of Maximum Volume.

Power On Volume is stored on the underlying Master Volume scale. When Zero Point is non-zero, the value shown to the user is shifted by the same amount as the displayed Master Volume.

For example:

Zero Point = -15 dB
stored Power On Volume = -35 dB
displayed Power On Volume = -20 dB

After restart, the HTP-1 restores the stored -35 dB underlying Master Volume, which is again displayed as -20 dB.

Mix Out Volume¶

The Mix Out uses a separate digital volume control and does not use the main output's Zero Point.

The confirmed ranges are:

/secondaryVolume         -100 … 0 dBFS
/secondaryPowerOnVolume  -100 … 0 dBFS

Unlike the main outputs, Mix Out does not provide the main analog volume control's gain range above 0 dB.

Reference Output Voltage¶

Previously called Maximum Output Level.

The setting ranges from 0.1 to 4.0 Vrms and defines the nominal balanced output produced by a 0 dBFS sine wave at internal 0 dB Master Volume, subject to the normal board correction and calibration tolerances.

It does not mean:

every signal produces that RMS voltage;
the instantaneous output voltage is always fixed at that value;
processing cannot raise or lower peak levels before the master-volume stage;
a momentary full-scale sample has an independently defined Vrms value;
the setting can override the HTP-1's approximately 4 Vrms hardware limit.

The name Reference Output Voltage distinguishes the calibration setting from the hardware's absolute clean-output capability.

Maximum Digital Headroom¶

Maximum Digital Headroom specifies the maximum amount of digital attenuation the HTP-1 can retain to accommodate gain introduced by DSP processing.

It is not a fixed reserve guaranteed at every Master Volume setting.

As Master Volume rises beyond the point reported as Highest volume with full digital headroom, the HTP-1 gradually consumes the configured reserve to allow higher playback levels.

Configured and available digital headroom¶

These are different quantities:

Configured digital headroom is the value selected with Maximum Digital Headroom.
Currently available digital headroom is the reserve remaining at the current underlying Master Volume.

Documentation must not use these terms interchangeably.

Peak Monitor¶

Peak Monitor records the largest instantaneous post-processing sample at its measurement point.

A linear value of 1.0 represents digital full scale. It does not calculate:

RMS level;
average level;
crest factor;
sustained power;
an analog Vrms value.

If Peak Monitor indicates that H dB of headroom is required, the relevant interpretation is that the largest observed post-processing sample was about H dB above digital full scale before the protective digital attenuation was applied.

Zero Point¶

Zero Point changes the displayed scale used for main-volume positions. It does not change any underlying volume setting or gain.

The relationship is:

displayed Main Volume     = /volume     - /cal/zeroPoint
displayed Power On Volume = /powerOnVol - /cal/zeroPoint
displayed Minimum Volume  = /cal/vpl    - /cal/zeroPoint
displayed Maximum Volume  = /cal/vph    - /cal/zeroPoint

The displayed Highest volume with full digital headroom is shifted by the same Zero Point offset.

Zero Point does not change:

the underlying Master Volume;
Power On Volume;
Minimum Volume or Maximum Volume;
analog or digital gain;
output voltage;
the internal 0 dB reference point;
Maximum Digital Headroom;
currently available digital headroom;
Mix Out Volume or Mix Out Power On Volume.

The confirmed underlying ranges are:

/volume                              -100 … +22 dB
/powerOnVol                          -100 …   0 dB

/secondaryVolume                     -100 …   0 dBFS
/secondaryPowerOnVolume              -100 …   0 dBFS
/shaker/savedSecondaryPowerOnVolume  -100 …   0 dBFS

Zero Point is never added to or subtracted from these stored values. Changing Zero Point therefore changes their presentation where applicable, but not the actual playback level, power-on level, or volume limits.

For example, with Zero Point set to -15 dB, an underlying Master Volume or Power On Volume of -35 dB is displayed as -20 dB.

When Mix Out tracks Main Volume, the tracked Mix Out level follows the underlying Main Volume but is clamped to 0 dBFS if Main Volume rises above internal 0 dB.

Confirmed implementation details¶

Volume ranges¶

The confirmed volume-related ranges are:

/volume                              -100 … +22 dB
/powerOnVol                          -100 …   0 dB

/secondaryVolume                     -100 …   0 dBFS
/secondaryPowerOnVolume              -100 …   0 dBFS
/shaker/savedSecondaryPowerOnVolume  -100 …   0 dBFS

The main output and Mix Out therefore do not share the same upper volume limit.

The main output's +22 dB range is associated with the analog volume-control stage described below. Mix Out is a digital downmix level and is limited to 0 dBFS.

Power-on restore behavior¶

Power On Volume values are stored on their underlying scales.

At startup:

/volume = /powerOnVol

and:

/secondaryVolume = /secondaryPowerOnVolume

subject to Seat Shaker routing and volume-tracking behavior.

Zero Point is not part of this restore calculation.

This distinction is important. Applying Zero Point while restoring Power On Volume would change the actual playback level after restart rather than merely changing its displayed value.

Voltage conversion¶

The user-facing voltage setting is converted approximately as follows:

Vrms
→ 20 × log10(Vrms)
→ user clip/reference level in dBV
→ board correction
→ _correctedClipVolume_dB

The production-board correction found in the source is approximately:

2.556 dB

derived from:

((12.077 + 12.035) / 2) - 9.5

The full engineering rationale for this correction has not yet been verified. Do not expose it in normal user documentation unless that rationale is known.

Internal 1 dB reserve¶

The implementation retains an additional approximately 1 dB digital reserve.

With configured headroom H, the highest underlying Master Volume that retains the full configured headroom is approximately:

Highest MV with full configured headroom ≈ 1 dB - H

Examples:

12 dB configured headroom → approximately -11 dB MV
18 dB configured headroom → approximately -17 dB MV

The UI displays this threshold using the current Zero Point scale.

Use the live UI readout as the authoritative value where available.

Analog gain limit¶

volAna is capped at +22 dB, matching the maximum gain setting of the CS3318 analog volume-control stage.

Conceptually, the implementation does this:

if (volAna > 22) {
    volDig += (volAna - 22);
    volAna = 22;
}

Any requested gain beyond the analog stage's +22 dB control limit is transferred to the digital stage.

This analog gain range applies to the main output path. It must not be used as the range for Mix Out.

CS3318 signal-swing limit¶

The CS3318 is operated from approximately ±9 V analog rails.

Its data sheet specifies the full-scale input/output range for THD+N < 1% as:

(VA-) + 1.35 V  to  (VA+) - 1.35 V

With ±9 V rails, this corresponds to approximately:

±7.65 Vpeak
15.3 Vpp
5.41 Vrms for a sine wave

This is a signal-swing limit, not the same thing as the CS3318's +22 dB gain setting.

Output stage and maximum external balanced output¶

The analog output stage after the CS3318 has approximately -2.7 dB gain.

Applying that fixed attenuation to the CS3318's approximately 5.41 Vrms maximum sine-wave output gives:

5.41 × 10^(-2.7/20) ≈ 3.96 Vrms

This explains the HTP-1's approximately 4.0 Vrms clean balanced-output limit.

The resulting analog structure is approximately:

CS3318 maximum clean sine-wave output: ≈5.4 Vrms
                           ↓ -2.7 dB
Balanced XLR output:                    ≈4.0 Vrms

Therefore:

the CS3318's unused gain-setting range is not necessarily unused clean output-voltage capability;
increasing analog gain also raises any full-scale peak;
once the required output reaches the hardware limit, further gain only moves clipping into the analog path;
exceeding approximately 4 Vrms cleanly requires hardware changes.

Gain algorithm¶

The system does not follow one universal, simple sequence such as:

analog only → digital only → analog only

Its behavior depends on both Reference Output Voltage and Maximum Digital Headroom.

A useful approximation is:

Region 1: Full configured headroom¶

At lower underlying Master Volume settings, analog gain increases while the full configured digital headroom remains available.

Region 2: Configured headroom is consumed¶

Above Highest volume with full digital headroom, the HTP-1 begins consuming the configured digital reserve to allow higher playback levels.

At approximately internal 0 dB Master Volume, only the normal internal reserve remains.

Region 3: Analog gain above internal 0 dB¶

Above internal 0 dB Master Volume, the processor may continue increasing analog gain, depending on the configured Reference Output Voltage and the remaining analog output capability.

A non-zero Zero Point may cause this region to appear at a different displayed volume value.

Region 4: Positive digital gain¶

Once volAna reaches its +22 dB limit, any remaining requested gain is moved to volDig.

Whether and where this region occurs depends on Reference Output Voltage.

Important consequences¶

Maximum Digital Headroom is not permanent¶

The configured value is the maximum reserve the HTP-1 attempts to preserve.

It remains fully available only up to Highest volume with full digital headroom. Above that point, available headroom decreases.

The UI displays that threshold using the Zero Point-shifted volume scale, but the underlying threshold does not change when Zero Point changes.

Reference Output Voltage is anchored to internal 0 dB MV¶

Reference Output Voltage is anchored to internal 0 dB Master Volume, not displayed 0 dB and not Maximum Volume.

Reducing Maximum Volume therefore reduces the output of a conventional 0 dBFS test signal, but it does not by itself describe the output produced by an internally boosted post-processing peak.

Changing Zero Point does not change Reference Output Voltage or the internal volume setting to which it is referenced.

Displayed 0 dB is not necessarily internal 0 dB¶

With Zero Point set to 0 dB:

displayed 0 dB = internal 0 dB

With a non-zero Zero Point, that is no longer true.

For example:

Zero Point = -15 dB
internal Master Volume = -15 dB
displayed Master Volume = 0 dB

Documentation and UI text discussing gain behavior, Reference Output Voltage, or hardware limits must therefore use internal 0 dB Master Volume where that distinction matters.

Power On Volume is stored independently of Zero Point¶

Changing Zero Point must not rewrite Power On Volume.

For example:

stored Power On Volume = -35 dB
Zero Point = 0 dB
displayed Power On Volume = -35 dB

Zero Point changed to -15 dB

stored Power On Volume = -35 dB
displayed Power On Volume = -20 dB

The actual power-on playback level remains unchanged.

Mix Out does not use Zero Point¶

Mix Out Volume and Mix Out Power On Volume remain on their own -100 … 0 dBFS scale regardless of the main output's Zero Point setting.

When Seat Shaker volume tracking routes the signal through Mix Out, the tracked level follows underlying Main Volume only up to 0 dBFS.

For example:

Main Volume = -10 dB → Mix Out tracking = -10 dBFS
Main Volume =   0 dB → Mix Out tracking =   0 dBFS
Main Volume =  +8 dB → Mix Out tracking =   0 dBFS

Peak level and Vrms must not be mixed¶

Peak Monitor reports instantaneous digital sample peaks. Reference Output Voltage is specified using the RMS value of a sine wave.

A full-scale sine with a 4 Vrms output has:

4.00 Vrms
5.66 Vpeak
11.31 Vpp

Any legal DAC signal whose samples stay within ±1.0 cannot require more than the corresponding instantaneous peak voltage at the same gain setting, although its RMS value may be very different.

Do not assign a program transient an RMS voltage solely from its sample peak. RMS requires knowledge of the waveform over time.

A 4 Vrms setting does not imply clean output above 4 Vrms¶

With Reference Output Voltage set to 4 Vrms, a full-scale sine at internal 0 dB Master Volume is already near the clean balanced-output limit.

Higher underlying Master Volume settings may request more gain, but they do not create proportionally higher clean voltage for a full-scale signal. Positive digital gain can instead drive the digital or analog path into clipping.

Gain-setting range is not output-voltage headroom¶

The earlier idea that apparently unused CS3318 gain could be used to preserve more digital headroom was based on conflating:

unused CS3318 gain-setting range

with:

unused clean analog output-voltage capability

They are not equivalent.

If a post-processing peak is attenuated to 0 dBFS at the DAC, adding analog gain raises that same full-scale peak. Once the XLR output would exceed approximately 4 Vrms for a sine-equivalent full-scale signal, clipping has merely moved from the digital domain to the analog domain.

Crest factor does not change the instantaneous clipping limit¶

Crest factor affects:

RMS level;
average power;
heating;
perceived loudness.

It does not change the instantaneous voltage required for the specific maximum sample captured by Peak Monitor.

Reassessment of “Preserve Digital Headroom” mode¶

The original proposal assumed the current algorithm left clean analog gain unused and that gain could be reallocated to preserve more digital headroom while still reaching an independently selected output voltage.

The hardware review shows that premise is incorrect.

A fixed-headroom or shifted-reference algorithm cannot create additional clean output capability. It can only:

change where gain is allocated;
change where clipping occurs;
or limit the accessible volume range.

A new gain-allocation mode therefore cannot guarantee both:

full configured digital headroom; and
an independently chosen Reference Output Voltage at the maximum listening level;

unless that combination already fits within the existing total gain and approximately 4 Vrms hardware limit.

The preferred product improvement is a calibration workflow that configures and explains the existing controls rather than introducing a second gain algorithm.

UI guidance¶

Displayed and underlying volume values¶

Any UI that presents main-volume positions must distinguish between the underlying value and the Zero Point-shifted displayed value.

The intended model is:

displayed value = underlying value - Zero Point

for:

Main Volume;
Power On Volume;
Minimum Volume;
Maximum Volume;
Highest volume with full digital headroom.

Zero Point must not be applied to:

Mix Out Volume;
Mix Out Power On Volume;
Maximum Digital Headroom;
currently available digital headroom;
stored underlying MSO values.

User input entered on a displayed Zero Point-shifted scale must be converted back to the underlying scale before being stored.

For example:

Zero Point = -15 dB
user enters displayed Power On Volume = -20 dB
stored /powerOnVol = -35 dB

Volume limits¶

The underlying limits are:

Main Volume:          -100 … +22 dB
Main Power On Volume: -100 …   0 dB
Mix Out Volume:       -100 …   0 dBFS
Mix Out Power On:     -100 …   0 dBFS

UI controls must not reuse the main output's +22 dB ceiling for Mix Out.

“Effective voltage at that volume”¶

A value calculated as:

Reference Output Voltage × 10^(Master Volume / 20)

is mathematically valid only as:

Expected RMS output of a 0 dBFS sine wave at the selected underlying Master Volume.

It is not:

the output of the loudest program material;
a headroom-status value;
the RMS voltage corresponding to the Peak Monitor reading;
proof that the amplifier will or will not clip on a transient.

On the Levels page this value is likely to confuse users because it answers a test-signal question rather than a calibration question.

Preferred options:

remove it from the Levels page; or
move it to diagnostics and label it explicitly:

Expected output for a 0 dBFS sine at this volume

Do not label it merely as Effective voltage.

Headroom preservation status¶

A more useful Levels-page status is the relationship between Maximum Volume and configured headroom.

Examples:

Full configured headroom is preserved up to -11 dB MV.

Maximum Volume preserves all configured digital headroom.

Maximum Volume consumes 3 dB of configured digital headroom.

Where Zero Point is active, normal user-facing values should use the displayed volume scale. Engineering explanations of the gain algorithm should identify underlying or internal Master Volume explicitly.

This status should use the live implementation result rather than reproducing the gain algorithm independently in the UI where possible.

Documentation guidance¶

Preferred terms:

Reference Output Voltage
Maximum Digital Headroom
configured digital headroom
currently available digital headroom
Highest volume with full digital headroom
Maximum Volume
Power On Volume
Mix Out Volume
Zero Point
Peak Monitor
internal 0 dB Master Volume, where the distinction from displayed 0 dB matters

Recommended user-level explanation:

The HTP-1 automatically combines analog and digital volume control. At lower listening levels it preserves the full configured Maximum Digital Headroom. As Master Volume increases beyond the displayed Highest volume with full digital headroom, that reserve is gradually consumed. Above internal 0 dB Master Volume, analog gain may continue to increase until the analog stage reaches its limit. Any remaining requested gain is applied digitally and may clip sufficiently high-level signals. Zero Point changes only the displayed main-volume scale and does not change these underlying gain relationships.

Avoid wording that implies:

Maximum Digital Headroom is permanently reserved;
Reference Output Voltage is always the instantaneous output voltage;
Reference Output Voltage describes arbitrary program RMS level;
Reference Output Voltage is referenced to displayed 0 dB when Zero Point is non-zero;
a Peak Monitor sample value can be converted directly into Vrms without knowing the waveform;
unused CS3318 gain-setting range equals unused clean output voltage;
a different gain split can create clean voltage beyond the hardware limit;
Maximum Volume changes internal gain calculations;
Zero Point changes playback level;
Zero Point changes stored Power On Volume, Minimum Volume, or Maximum Volume;
Zero Point applies to Mix Out Volume or Mix Out Power On Volume;
displayed 0 dB necessarily corresponds to internal 0 dB Master Volume;
the main output and Mix Out have the same upper volume limit;
Mix Out can use the main analog stage's +22 dB gain range;
a 4 Vrms reference setting allows clean full-scale output above 4 Vrms.

Open questions¶

Full engineering rationale for the approximately 2.556 dB board correction.
Exact behavior under every processing combination and every Reference Output Voltage setting.
Exact tolerance of the practical 4 Vrms limit across production units and loads.
Whether the UI should remove the 0 dBFS sine-output calculation entirely or retain it in an advanced diagnostics section.