hyperion Axis16: Forced High-Pass Filter / Bass Cut on Top/Height Channels during Dolby Atmos Decoding?

[hushin0917]

Hi everyone, I recently acquired an AudioControl Hyperion Axis16 for my multi-channel setup (outputting via Dante to a Ferrofish Pulse16 ADC/DAC).

While performing initial setup measurements and signal verification, I encountered a few unexpected rendering/routing behaviors regarding the Height channels and LFE levels. I’d love to get the community's technical insights on this.

1. Test Setup & Source Signal

• Processor: AudioControl Hyperion Axis16 (Output via Dante)
• Audio Interface: Ferrofish Pulse16 (Dante Master Clock @ 48kHz / 24-bit)
• Source: Mac playing official Apple Music 7.1.4 spatial test tracks: Left Channel Pink Noise by Spatial Audio Tools, sent via HDMI / eARC to the Axis16.
• Measurement Method: Signal captured via Loopback routing and analyzed using REW RTA.

2. Key Findings from REW RTA (Screenshots Attached)

• Height Channel Bass Cut: When decoding native Dolby Atmos bitstream, the Top/Height channels exhibit severe low-end attenuation/roll-off below 150 Hz.
• L/R Channel Gain Offset: The Main L/R channels measure roughly 1 dB higher than the other surround channels.
• LFE Level Elevation: The LFE channel measures approximately +16 dB higher than the baseline surround channels.

3. Controlled Cross-Tests & Variables

• Mac CoreAudio Baseline: When using Mac CoreAudio software decoding (outputting multi-channel PCM bypass), NONE of these issues occur — the Height channels output full-range signals without any bass cut.
• GUI Configuration: Frequency Response for all channels is explicitly set to Full Range in the Axis16 Web UI.
• Mode Verification: Audio Mode is set to Native and DRC is set to Off.
• Dante Stability: Dante network clock is locked to Ferrofish (Master) with Latency set to 5ms–10ms, reporting 0 late packets in Dante Controller over long periods.

4. Questions for the Community

• Is the forced roll-off below 150Hz on Height channels a mandatory restriction built into the official Dolby Atmos Hardware Rendering Engine (SDK) for A/V processors, or is it a DSP bypass bug in AudioControl’s firmware?
• Is the +16 dB LFE offset expected due to standard Dolby +10 dB hardware boost combined with Dialnorm / metadata rendering differences vs. software decoding?
• Has anyone experienced similar DSP channel routing/filtering on other processors (e.g., AudioControl, StormAudio, Trinnov) outputting via Dante?

(Screenshots of REW RTA captures attached below)

 

[Littletycoon]

Hi everyone, I recently acquired an AudioControl Hyperion Axis16 for my multi-channel setup (outputting via Dante to a Ferrofish Pulse16 ADC/DAC).

While performing initial setup measurements and signal verification, I encountered a few unexpected rendering/routing behaviors regarding the Height channels and LFE levels. I’d love to get the community's technical insights on this.

1. Test Setup & Source Signal

• Processor: AudioControl Hyperion Axis16 (Output via Dante)
• Audio Interface: Ferrofish Pulse16 (Dante Master Clock @ 48kHz / 24-bit)
• Source: Mac playing official Apple Music 7.1.4 spatial test tracks: Left Channel Pink Noise by Spatial Audio Tools, sent via HDMI / eARC to the Axis16.
• Measurement Method: Signal captured via Loopback routing and analyzed using REW RTA.

2. Key Findings from REW RTA (Screenshots Attached)

• Height Channel Bass Cut: When decoding native Dolby Atmos bitstream, the Top/Height channels exhibit severe low-end attenuation/roll-off below 150 Hz.
• L/R Channel Gain Offset: The Main L/R channels measure roughly 1 dB higher than the other surround channels.
• LFE Level Elevation: The LFE channel measures approximately +16 dB higher than the baseline surround channels.

3. Controlled Cross-Tests & Variables

• Mac CoreAudio Baseline: When using Mac CoreAudio software decoding (outputting multi-channel PCM bypass), NONE of these issues occur — the Height channels output full-range signals without any bass cut.
• GUI Configuration: Frequency Response for all channels is explicitly set to Full Range in the Axis16 Web UI.
• Mode Verification: Audio Mode is set to Native and DRC is set to Off.
• Dante Stability: Dante network clock is locked to Ferrofish (Master) with Latency set to 5ms–10ms, reporting 0 late packets in Dante Controller over long periods.

4. Questions for the Community

• Is the forced roll-off below 150Hz on Height channels a mandatory restriction built into the official Dolby Atmos Hardware Rendering Engine (SDK) for A/V processors, or is it a DSP bypass bug in AudioControl’s firmware?
• Is the +16 dB LFE offset expected due to standard Dolby +10 dB hardware boost combined with Dialnorm / metadata rendering differences vs. software decoding?
• Has anyone experienced similar DSP channel routing/filtering on other processors (e.g., AudioControl, StormAudio, Trinnov) outputting via Dante?

(Screenshots of REW RTA captures attached below)

seems like audiocontrol has work on their hand.

Some of the issues you mentioned (LFE level elevation, Gain offset) were also reported by users of the Nuprime 16 AOIP processor, now I think fixed with the latest firmware. Maybe it's even the same decoding module/board supplier in the 2 products.

 

[hushin0917]

seems like audiocontrol has work on their hand.

Some of the issues you mentioned (LFE level elevation, Gain offset) were also reported by users of the Nuprime 16 AOIP processor, now I think fixed with the latest firmware. Maybe it's even the same decoding module/board supplier in the 2 products.

I hope for a quick update; this is a great Dolby decoder.

 

[hushin0917]

Normally, it should look like this screenshot. I tested different channels, and they should all align at the exact same sound pressure level (SPL) for proper decoding. This reference curve was generated using macOS's built-in Dolby Atmos decoding.

 

[darth]

H16-AES user here. Not sure if I am trying it correctly, but when I play those tracks I also notice the surrounds/tops are 1-2db-ish lower than LR. But LFE reading seems way off and read technically muted for me, even after a +10db gain I added. Maybe something wrong with that track or how this GLM software is reading, seems some rando album, no? Try

or

. I had tried some REW exported LFE signals and played via apple music which seemed fine. Findings in the last few posts of this forum https://audiophilestyle.com/forums/topic/71645-h16-firmware-issues/page/4/#findComment-1349674

 

[hushin0917]

H16-AES user here. Not sure if I am trying it correctly, but when I play those tracks I also notice the surrounds/tops are 1-2db-ish lower than LR. But LFE reading seems way off and read technically muted for me, even after a +10db gain I added. Maybe something wrong with that track or how this GLM software is reading, seems some rando album, no? Try

or

. I had tried some REW exported LFE signals and played via apple music which seemed fine. Findings in the last few posts of this forum https://audiophilestyle.com/forums/topic/71645-h16-firmware-issues/page/4/#findComment-1349674

View attachment 556153View attachment 556154View attachment 556155

I also own the H16, and it has similar bugs. In fact, that was the main reason I decided to switch and purchase the AXIS-16—I certainly didn't expect it to exhibit similar issues as well.

 

[darth]

Have you tried playing directly from the mac using dante virtual soundcard? I don't have AoIP so can't test if this approach even works. But theoretically if you use the dante virtual soundcard and setup a 7.1.4 setup in midi, macs core audio can maybe decode atmos directly on the Mac and send it over the discreet channels.

 

[hushin0917]

Have you tried playing directly from the mac using dante virtual soundcard? I don't have AoIP so can't test if this approach even works. But theoretically if you use the dante virtual soundcard and setup a 7.1.4 setup in midi, macs core audio can maybe decode atmos directly on the Mac and send it over the discreet channels.

Yes, I purchased Dante Virtual Soundcard (DVS) specifically to capture and loop back the decoded Dolby signals for verification.

 

[darth]

So did atmos decoding work without the need of the axis16?

 

[hushin0917]

So did atmos decoding work without the need of the axis16?

If you're only listening to Dolby Atmos music, you can simply use a Mac to perform the decoding for playback.

 

[grobalt]

was this behaviour shared with hyperion so they can work on it ? the AVR had some initial flaws as well with quick fix.

 

[hushin0917]

was this behaviour shared with hyperion so they can work on it ? the AVR had some initial flaws as well with quick fix.

I have simultaneously issued a formal notice and sent a letter to them, and am currently awaiting their handling of the matter.