Possible solution for DTS/Dolby/Atmos+eARC+HDCP to AES/EBU - via Dante? (for Okto DAC8PRO etc)

[Kal Rubinson]

Yup. A couple of weeks back.

 

[welwynnick]

Looks like the audiocontrol device got pushed out by another quarter. It was Q1 2025, then Q2, now Q3.

Yup, no surprise.
That launch target appears to refer to the APR-16 with analogue outputs, not the DPR-16 with AES & AoIP outputs.
I don't think I've seen a picture of the DPR-16 before - it looks like 1U / 1PCB size, like the Arvus H2-4H, which is a refreshing change from other processors.
Since it's obviously not just the AVP with the analogue outputs removed, perhaps it will be a reasonable price?

EDIT: I just noticed that while AudioControl have previously used Dirac and DLBC on some of their previous AVPs & AVRs, they now claim to offer DL ART on the APR-16:

At ISE 2025, AudioControl unveiled the Hyperion APR-16, a state-of-the-art home theater audio preprocessor designed to elevate the immersive audio experience. This 16-channel processor offers both RCA and XLR outputs, providing flexibility for various system configurations. It supports the latest immersive surround formats, including Dolby Atmos, DTS:X, Auro-3D, and IMAX Enhanced, ensuring compatibility with a wide range of content.

One of the standout features of the APR-16 is its advanced HDMI management, which includes support for eARC (Enhanced Audio Return Channel), allowing for seamless integration with modern displays and source devices. Additionally, the processor incorporates DIRAC Live room correction technology, including Bass Control and ART (Active Room Treatment), enabling precise calibration to your listening environment for optimal sound quality. The inclusion of DANTE audio-over-Ethernet protocol further enhances its versatility, facilitating high-quality audio distribution over standard networks.

 

[darth]

New product for the table
Nuprime-x H16-I (same as A but with AES3 out)

Nuprime H16-I Product

The H16-I is nearly identical to H16-A except the output and pricing (confirm by next week, expect to be lower than H16-A). refer to H16-A for details: https://nuprime-x.com/product/h16a/ Shipping: End of June / 1st week of July We follow TASCAM AES3 DB25 pinout. For H16-I, the input section of A...

audiophilestyle.com

 

[90sNintendoKid]

I've been watching this thread and I didn't notice the price of the Nuprime H16.

I am seriously considering a Nuprime H16-I and maybe two Okto dac8 PROs and Camilla DSP for EQ........

I would like to investigate this further like how do I handle syncing the DACs/clocks (not something I know a lot about).

I'm wondering whether a Windows Mini PC or Mac Mini and something like Hang Loose Host might work as well.

 

[Esprit]

New product for the table
Nuprime-x H16-I (same as A but with AES3 out)

This, on paper, is really interesting.

 

[darth]

I've been watching this thread and I didn't notice the price of the Nuprime H16.

I am seriously considering a Nuprime H16-I and maybe two Okto dac8 PROs and Camilla DSP for EQ........

I would like to investigate this further like how do I handle syncing the DACs/clocks (not something I know a lot about).

I'm wondering whether a Windows Mini PC or Mac Mini and something like Hang Loose Host might work as well.

The H16-A is 799, so the H16-I will likely be below that. I have my fingers crossed for under 500 lol though unlikely

 

[carolimagni]

I've been watching this thread and I didn't notice the price of the Nuprime H16.

I am seriously considering a Nuprime H16-I and maybe two Okto dac8 PROs and Camilla DSP for EQ........

I'm still rocking USB at the moment - but haven't abandoned the idea in the OP. I did however get an Okto since starting the thread. I've since deep-dived on the third-party DSP firmware, which is fairly powerful and worth looking into. Although when using all AES inputs you're limited to 264 MIPS across 4 cores which limits it significantly. My attached 5.1 bass management requires ~224 MIPS across 7 cores. If you want EQ with Oktos and all AES input, using the AES/USB mode and software EQ is probably the way to go.

I would like to investigate this further like how do I handle syncing the DACs/clocks (not something I know a lot about).

I'm wondering whether a Windows Mini PC or Mac Mini and something like Hang Loose Host might work as well.

I think you'd face no issues syncing the Nuprime's output amongst the Oktos, since as it's a single AES source. The Nuprime would be the clock source.
Relevant Okto manual bits:

Pure AES
Pure AES mode enables playback from the eight digital audio channels available on the four
AES/EBU input connectors. It utilizes the clock recovered from the first AES/EBU input for
processing the received digital audio data, and the dac8 PRO's internal low-phase noise
oscillators for the actual D/A conversion. The dac8 PRO's jitter eliminator is active in Pure
AES mode, ensuring the highest possible jitter rejection and performance.

Signal integrity
AES/EBU (as well as SPDIF and TOSLINK) is a synchronous connection, meaning that the
signal source device also serves as a clock source. Moreover, the clock signal is encoded
into a single bitstream with audio data using biphase mark code. This places the
responsibility of reconstructing the clock signal on the receiving device, and in the case of a
D/A converter, deviations can lead to undesirable noise or spurious signals at the output.
Although the dac8 PRO employs the highly effective Sabre jitter elimination algorithm, it is
recommended to ensure the best signal integrity possible. One important and commonly
overlooked factor is cable impedance. The source and termination impedance for AES/EBU
is 110 Ohms, and while reception will likely work with a line cable, using a dedicated cable
with 110 Ohm impedance is highly recommended.

Issue 2: Dropouts during AES/EBU playback
Recommended action: Navigate to Menu > Advanced > DPLL BW AES and increase the
value until the dropouts disappear.

Advanced submenu
DPLL BW AES: controls bandwidth of the de-jittering mechanism for AES/EBU inputs.
Increasing the value will allow for operation with more jittery signal sources at a cost
of slightly reduced jitter suppression
- range: 0 to 15 (default: 8)

However there remains a question about sampling rate. From the Okto manual (the default Linux, Mac and Win drivers are the UAC2 driver):

USB / AES
Because the UAC2 audio device is unable to change the USB host current sampling
frequency, the dac8 PRO will display a warning message, "fs mismatch," if the received
14 | dac8 PRO Owner’s Manual | okto research
AES/EBU signal's sampling frequency does not correspond to the USB host driver's current
frequency. User may be required to take action to ensure that the two sample rates match.

This begs an important question of the Nuprime unit. User welwynnick at audiophilestyle asks an excellent question:

The plot continues to thicken...

 

[carolimagni]

Apologies if you're all already aware of all this but since I mentioned the Okto's third-party firmware, feel like it's responsible to link to relevant info in-case others tunnel vision into it that and flash the generic firmwares. It's quite flexible; you technically sacrifice some DSP horsepower compared to software solutions but it becomes platform agnostic (Note: DSP compile+flash is not). It's technically third-party but with collaboration from Pavel. I would contact fabriceo before flashing any firmware and source one with your S/N on it incase you resell your unit. Read this thread, the amirm measurements thread from this post onward.
https://github.com/fabriceo/AVDSP_DAC8 Github
https://github.com/fabriceo/AVDSP_DAC8/blob/main/documents/AVDSP for DAC8_V160_1.pdf Manual

 

[edechamps]

Thankfully you will not have this problem with the LG C4. It supports everything on eARC. There is no reason to not use eARC with a C4, as far as I'm aware, aside from the number of HDMI inputs you may need.

…and it was nice while it lasted. The 2025 series from LG (C5/G5) dropped support for DTS passthrough over ARC… sigh…

This is incredibly frustrating when you realize there can be no technical reason for this: passthrough does not require any processing, it's literally just copying bits!

 

[90sNintendoKid]

I'm still rocking USB at the moment - but haven't abandoned the idea in the OP. I did however get an Okto since starting the thread. I've since deep-dived on the third-party DSP firmware, which is fairly powerful and worth looking into. Although when using all AES inputs you're limited to 264 MIPS across 4 cores which limits it significantly. My attached 5.1 bass management requires ~224 MIPS across 7 cores. If you want EQ with Oktos and all AES input, using the AES/USB mode and software EQ is probably the way to go.

I think you'd face no issues syncing the Nuprime's output amongst the Oktos, since as it's a single AES source. The Nuprime would be the clock source.
Relevant Okto manual bits:




However there remains a question about sampling rate. From the Okto manual (the default Linux, Mac and Win drivers are the UAC2 driver):

This begs an important question of the Nuprime unit. User welwynnick at audiophilestyle asks an excellent question:
View attachment 453144
The plot continues to thicken...

Thank you, much appreciated. I didn't know about the third party Okto firmware, I'll do some research.

 

[darth]

…and it was nice while it lasted. The 2025 series from LG (C5/G5) dropped support for DTS passthrough over ARC… sigh…

This is incredibly frustrating when you realize there can be no technical reason for this: passthrough does not require any processing, it's literally just copying bits!

Indeed. I hope it's just a bug and not some hidden licensing deal with Dolby considering that TV also gimped on HDR10

Also thanks to @welwynnick for asking the hard clarifying questions in the other forum

 

[welwynnick]

However there remains a question about sampling rate. From the Okto manual (the default Linux, Mac and Win drivers are the UAC2 driver):

This begs an important question of the Nuprime unit. User welwynnick at audiophilestyle asks an excellent question:
View attachment 453144
The plot continues to thicken...

First they answered by saying ambiguously:

Whatever supported from HDMI on H16 goes to AES3 output. This is the same as AoIP on H16-A.

Then I replied to get clarification:

Therefore I hope that the H16-I output sample rate mirrors the incoming rate on HDMI, and is NOT the same as AoIP on H16-A.

What a relief - Nuprime have just confirmed:

You are correct. H16-AES will show whatever sample rates from HDMI.

This is what we want - The AES3 output sample rate will be whatever the HDMI input sample rate is.
Therefore the H16-AES (as it's now called) will not have a fixed output rate like the H16-AIP (as that's now called).

 

[carolimagni]

This is what we want - The AES3 output sample rate will be whatever the HDMI input sample rate is.
Therefore the H16-AES (as it's now called) will not have a fixed output rate like the H16-AIP (as that's now called).

That's positive. For us Okto users, I suppose this means you're plain sailing if you're using AES mode. If one used the dual USB/AES mode on the Okto to do room correction in software, would you still have to adjust the USB host driver sample rate for different HDMI source material? I may be misunderstanding, perhaps this error cited in the docs only occurs if the USB host driver is set lower than the source and maybe so long as its to 192KHz you can forget about it?

USB / AES
Because the UAC2 audio device is unable to change the USB host current sampling
frequency, the dac8 PRO will display a warning message, "fs mismatch," if the received

AES/EBU signal's sampling frequency does not correspond to the USB host driver's current
frequency. User may be required to take action to ensure that the two sample rates match.

I intend to stick with correction using the onboard XMOS to keep the Oktos USB input truly platform-agnostic; so if I can achieve all I want with on-board DSP I perhaps avoid this issue.

Although I do intend to use the AES3 inputs on my Okto for a Schoeps CMD 42 Digital Microphone Amp (with PSD 42 AES3 output converter, and integrate it into my headphone DSP as USB input channel. This is surely USB/AES mode and I risk stumbling into the same issue. Although the mic interface is readily capable of 192Khz and is configurable - this means I could use this to test the Oktos behaviour and emulate what might happen with the Nuprime+Okto and different sample rate source material?

 

[welwynnick]

That's positive. For us Okto users, I suppose this means you're plain sailing if you're using AES mode. If one used the dual USB/AES mode on the Okto to do room correction in software, would you still have to adjust the USB host driver sample rate for different HDMI source material? I may be misunderstanding, perhaps this error cited in the docs only occurs if the USB host driver is set lower than the source and maybe so long as its to 192KHz you can forget about it?

I hadn't thought of that because I was focussed on what the output of the H16-AES did.
That would be a similar inconvenience to the fixed output of the H16-AIP, and when I say inconvenience, it's effectively a show-stopper.

However, doesn't the Okto have an SRC, so it can up-sample any AES3 input sample rate to fixed rate of say 192kHz for the USB loop? I'm afraid that the SRC would be downstream of the USB loop.

This is the configuration that I was pinning my hopes on.

 

[KitoDavy]

This thread is the reason why I registered ASR. I have 5 genelec 8361s coming next week. But I haven't pulled the trigger on stormaudio Evo 32aes.... I was hoping something cheaper. But not sure if the H16aes would be as good as stormaudio? At least the dirac art part can be done with glm....but then the base management... Do I still need 9301 for it...

 

[welwynnick]

This thread is the reason why I registered ASR. I have 5 genelec 8361s coming next week. But I haven't pulled the trigger on stormaudio Evo 32aes.... I was hoping something cheaper. But not sure if the H16aes would be as good as stormaudio? At least the dirac art part can be done with glm....but then the base management... Do I still need 9301 for it...

What the H16 brings to the table is Atmos decoding and multi channel digital audio outputs.
That enables you to use off-board processing (DSP) and decoding (DAC).
The processing of the H16-AES will almost certainly be like the H16-AIP, in that it only has basic parametric equalisation, nothing like the Storm.
One of the things that the Storm does is phase equalisation as well as amplitude equalisation.
That helps correct the phase (non) linearity at the cross-over of most speakers, and gives better time-domain response in-room.
You're fortunate that the 8361 already has similar capability built-in.
It's called extended phase linearity mode, and you have to switch it on (at the expense of a few milli-seconds delay).
Do you still need 9301? I don't know without further research, but you have to consider how you do volume control? It's a small price to pay to avoid using a Storm.

 

[darth]

9301B is only needed if you want to bass manage + need larger amount of digital connections. If you have the sam subwoofers, they can also bass manage using analog, but only the 7360A+ higher can do 7.1 channels. 7350A can only do 5.1, it doesn't have enough inputs for more than that.

Volume control can be done using 9310B (not to be misread as 9301B) connected to a glm module.

 

[Behrelec 37]

In addition to the Genelec speakers, I've been a little interested in the Audient Oria. It has the advantage of a USB input for example for a Mac mini. And it has AES outputs for digital Genelec speakers. Even 9.1.6 for Atmos.The disadvantage is that the input is Dante, which in its primitive state cannot change the sampling rate ? when the movie changes, and it doesn't have any kind of audio decoder, which is a big shame. The biggest question mark is also whether the unnecessary volume control can be bypassed completely? No one will ever tell you that !!! . Because the Genelecs have this SUPER GLM volume control option with Cat cables (without internet), which can be controlled for example with one wireless remote control or volume control knob or the volume control on the computer screen, so all speaker settings work immediately! At the same time, without delays, even if you have over a hundred Genelec speakers in your living room enjoying the sounds of the movie.

At least the engineer's brain from the Audient Oria perspective has already thought about remote control for the longer term (a very super super super great development for humanity!). For example, this would make controlling a tablet from the couch naturally easy. But the whole system is now unfortunately completely married to Apple, at least for now until spring 2025. You don't always get all the goodies and some connection needs are of course missing. I'm no longer particularly interested in these devices where the multichannel outputs are separated into analog connectors, because then you are completely married to the quality of the DA converters, they are outdated or often poorly implemented in terms of interference distances.

Audi's Oria has both AES digital and analog output and it's really good. I don't think the analog output can do anything with Genelec if you're looking for the best, highest quality sound from all outputs at the same time. Digitally, the Genelec option is the best way to handle multichannel sounds to the end.

 

[KitoDavy]

What the H16 brings to the table is Atmos decoding and multi channel digital audio outputs.
That enables you to use off-board processing (DSP) and decoding (DAC).
The processing of the H16-AES will almost certainly be like the H16-AIP, in that it only has basic parametric equalisation, nothing like the Storm.
One of the things that the Storm does is phase equalisation as well as amplitude equalisation.
That helps correct the phase (non) linearity at the cross-over of most speakers, and gives better time-domain response in-room.
You're fortunate that the 8361 already has similar capability built-in.
It's called extended phase linearity mode, and you have to switch it on (at the expense of a few milli-seconds delay).
Do you still need 9301? I don't know without further research, but you have to consider how you do volume control? It's a small price to pay to avoid using a Storm.

I guess I wouldn't miss too much with no going storm given the glm? For 9301, it is limited to 8 channels and lots of aes cabling.. I may prefer to go 9401 AoIP with 16 channels and db25 connection....And I didn't find other digital subwoofers either.

 

[KitoDavy]

9301B is only needed if you want to bass manage + need larger amount of digital connections. If you have the sam subwoofers, they can also bass manage using analog, but only the 7360A+ higher can do 7.1 channels. 7350A can only do 5.1, it doesn't have enough inputs for more than that.

Volume control can be done using 9310B (not to be misread as 9301B) connected to a glm module.

If I use 2 7380 with analog, then I don't think I need 9301 for 16 channels. But if I plan to go digital, I'll need 9301? One thing I don't understand is why the 9301 has glm? Doesn't every subwoofer has glm already? Given 7380 has analog function of this, I think the subs may be out of date.. at least the 9301 digital part should be embeded in the subs like the analog ones. Wouldn't you agree? But then i didn't find any other digital subs...Do you have any to recommend that goes well with Genelec?