Configuring a PC as a 8-ch pre/pro experiment

[Vasr]

The first one, I use a motu 1248 as the main device with another dac slaved to it. 5.1 system with 2 sub channels and 3 way LCR. I use the jriver convolver and peq for media played back in jriver which is fine (and would stay in use as jriver keeps video in sync with audio irrespective of filter length). wdm looped through jriver for streaming content and I find the wdm device v unstable (some memory leak in jriver makes it v crashy) so would be nice to have an alternative for this.

For this, I would skip the VoiceMeeter and use the ASIO Bridge from its developer. Set up the ASIO bridge to talk to the Motu via USB. That will expose as many channels as the USB handshake reports. Attach EAPO to its input end if you want to use any channel mapping/splitting/balancing/filtering/etc. Point your streaming output via WDM to the input of the ASIO bridge.

You can also skip the ASIO bridge and attach EAPO to the USB out directly and then point the stream via WDM to the USB and if it works use it but this might have some issues with buffering/latency/etc.

 

[Vincentponcet]

Software like Jriver make things simpler (at a cost) but it is a walled garden approach. It is like a soft-AVR on its own and they have done a good job. But everything has to be done from inside JRiver.

My experiment being described in this thread is to create a general solution that will work for any content player on the PC one may want to use and/or don't to pay for JRiver. For people who may want to do this instead of buying an AVR or going through a commercial solution like JRiver in a walled-garden approach.


The latency here has nothing to do with the power of the processor. Any software decoding of even the simplest codec like AC3 is very slow. But if you have a single player doing both audio and video rendering, then they have control over both to keep them in sync. If you have a weak CPU, it requires more time to buffer ahead and start. All media players do this - Kodi, MPC-BE, VNC, etc.

If you are using pass-through of the encoded audio to a preamp processor then there would be no decoding and so no latency issues of course. But if you have pre/pro available, the case for paying for JRiver is weaker. YMMV.

This post says using directshow GraphEdit and AC3Filter on an SPDIF input gave about 30ms latency which sounds acceptable for real time decoding https://www.avsforum.com/threads/the-inconvenient-truth-about-spdif-input.1121367/post-59130266
Did you tested it ?

 

[Vasr]

This post says using directshow GraphEdit and AC3Filter on an SPDIF input gave about 30ms latency which sounds acceptable for real time decoding https://www.avsforum.com/threads/the-inconvenient-truth-about-spdif-input.1121367/post-59130266
Did you tested it ?

No, the only reason I would want to do AC3 is to decode pass-through 5.1 coming in a bitstream which wasn't a use case for me. I had seen that forum post earlier. Seemed too unreliable to use GraphEdit and then a specific version of AC3 filter etc. Also not sure what happens if that optical input switches between AC3 and PCM after setting up that path. If it can only receive AC3, then that reduces the utility somewhat. Most things that do stereo optical can do PCM on it. So, for 2 channel optical in, that set up isn't really necessary.

 

[Vincentponcet]

No, the only reason I would want to do AC3 is to decode pass-through 5.1 coming in a bitstream which wasn't a use case for me. I had seen that forum post earlier. Seemed too unreliable to use GraphEdit and then a specific version of AC3 filter etc. Also not sure what happens if that optical input switches between AC3 and PCM after setting up that path. If it can only receive AC3, then that reduces the utility somewhat. Most things that do stereo optical can do PCM on it. So, for 2 channel optical in, that set up isn't really necessary.

The use case for Dolby digital input is when using your TV which receives DTV with 5.1 Dolby digital plus, and same for set top box provided by your Telco operator ( fiber or xDSL), they are all the time Dolby digital plus, and depending of the channel / content provides either 2.0 or 5.1

That's the source being a PCM 2 channels use case that I don't understand, what is this in your case ?

 

[Vasr]

The use case for Dolby digital input is when using your TV which receives DTV with 5.1 Dolby digital plus, and same for set top box provided by your Telco operator ( fiber or xDSL), they are all the time Dolby digital plus, and depending of the channel / content provides either 2.0 or 5.1

That's the source being a PCM 2 channels use case that I don't understand, what is this in your case ?

Most TVs allow you to set the optical out as Bitstream or PCM. If set to PCM they will decode and send PCM 2.0 not Dolby - down-mixing to Stereo if necessary. If Bitstream, they will pass through whatever they are getting as Dolby Digital (AC3)/Dolby Digital Plus (EAC3). You need to handle both. I think AC3Filter can do this but I don't know the latency for EAC3.

I use the optical for a internet radio/streamer box that sends 2.0 and only occasionally a stereo down mix from the TV just to get better sound than the TV speakers. The solution I have above is really not good for receiving encoded multi-channel in. That is a limitation of using a PC where Microsoft/Industry really do not want a PC to process incoming licensed encoder content.

 

[dualazmak]

I should also mention that this is a useful set up for use with Okto 8Pro and your choice of a multi-channel amp downstream.

The PC aggregates local content, streaming and optical sources into the USB out to the Okto and provides the common DSP features on all of them, Okto has the physical volume control. So you don't need anything else in the middle to do both music and HT (with the caveats mentioned above).

Although belated visiting this thread, I am building multichannel multi-amplifier system, just like you suggested, using OKTO DAC8PRO and software crossover EKIO;
https://www.audiosciencereview.com/...oftware-crossover-and-multichannel-dac.12489/

 

[phoenixdogfan]

After several months of experimentation (mostly trying to understand poorly written documentation for multiple tools), I was able to set up a HTPC as a sort of AVR for multi-channel content with the following configuration

View attachment 74095
Using Voicemeeter, Equalizer APO and VSTHost.

It is really not a satisfactory replacement for an AVR because you cannot have a convenient IR remote for everything, content with DRM can be a problem and the new codecs with more than 8-ch are out of the scope. But otherwise works very well from 2.0 to 7.1, integrating with your favorite media player that just works as long as it plays to a selected/default sound device sending PCM and the DSP settings/processing apply to all of them. Crossover, delay and volume can be set for each channel independently to align phase/balance. There are no latency issues for lip sync that I have encountered.

I have decided to stay with this until a decent pre/pro unit comes into market at a sensible price. Right now, they are either too expensive with features I don't need or have bare minimum features + lots of glitches and need multiple external boxes to get the full function as above. An older AVR is also an alternate solution if it has pre-outs but you would have to buy an expensive DIrac hardware box or do more than half of the work above to get it to work in the PC for multi-channel sound.

With REW based EQ I was able to get sounds very good almost as good as ARC but Dirac has some magic that makes it noticeably better. Setting up Dirac was a pain though.

Until Dirac bass management is available for PC, the bass management here works fine to integrate a sub.

Maximum 8 channels total though for the PCM output. So you can set it up for 5.2 but not 7.2.

I was going to write up some notes on how to do this but it can be quite an undertaking to describe it in a fool-proof way. I am not sure how much interest there is in replicating such a set up here to make it worthwhile.

JRiver does hav JRemote. I've gotten it to work with my phone for the app on my PC. I think switching could be accomplished by setting up multiple zones on JRiver. And if I can ever get it to work, Dirac now is multi channel, and promises to have multi sub bass management.

Don't know of a single reason why JRiver could not be used to do eq and crossovers, though it does not appear to have UI as friendly as, say, miniDSP.

With a good quiet PC, like and Intel Nuc in a fanless case, and a high quality 8 channel DAC like the you-know-who guys in Prague provide, I'm thinking it might be possible to assemble a multi channel 7.1 system as good or better than any multi kilobuck processor.

 

[phoenixdogfan]

The nuance is what I described above. It can do multi-channel in Dirac Live and in DLP and you can pick any of those configurations. But that just picks the number of channels processed independently in the DLP. There is no way to specify a crossover or a speaker as small in DLP/DL. So, it sends LFE channel to the Sub and the other speaker channels to the respective main, center and surround speakers with Dirac correction. So, if you have all "large" speakers and need no crossovers, it is fine. Otherwise you need an external crossover solution (in software or hardware) to direct the streams with a crossover. It even seems to do full range frequency sweeps to the Sub for measuring. In other words, 7.1 is just processed as 8.0 for all practical purposes, for example. it doesn't care what is connected to those channels.

The Dirac web site is a disaster with lots of old or missing or inconsistent information. It is run like a bad startup with great technology.

During the beta test phase they only had the 2.0 VST usage instructions. They have only updated the guides to the multi-channel VST usage this month for the solution I have described above using Voicemeeter that I had to discover on my own before then.

That website is a clear as mud, and like mud it's opaque but it covers the ground.

 

[AudioExplorer]

I should also mention that this is a useful set up for use with Okto 8Pro and your choice of a multi-channel amp downstream.

The PC aggregates local content, streaming and optical sources into the USB out to the Okto and provides the common DSP features on all of them, Okto has the physical volume control. So you don't need anything else in the middle to do both music and HT (with the caveats mentioned above).

I have been weighing getting the 8-channel Okto DAC since I have a multi-channel amp and a 5.1 system currently. If I wanted nothing more than HT playback for multichannel, can I connect the PC USB output to the Okto USB input, run software as you described for doing crossovers for the speakers that are small, set per channel levels in the Okto, and playback streaming sources including Netflix, ATV+, etc.? It seems like this should be possible but I just wanted to confirm since I have never done any of this before.

 

[Kal Rubinson]

I have been weighing getting the 8-channel Okto DAC since I have a multi-channel amp and a 5.1 system currently. If I wanted nothing more than HT playback for multichannel, can I connect the PC USB output to the Okto USB input, run software as you described for doing crossovers for the speakers that are small, set per channel levels in the Okto, .......

Yes.

...................and playback streaming sources including Netflix, ATV+, etc.?

Doable but not with the simplicity of an AVR.

 

[Vasr]

I have been weighing getting the 8-channel Okto DAC since I have a multi-channel amp and a 5.1 system currently. If I wanted nothing more than HT playback for multichannel, can I connect the PC USB output to the Okto USB input, run software as you described for doing crossovers for the speakers that are small, set per channel levels in the Okto, and playback streaming sources including Netflix, ATV+, etc.? It seems like this should be possible but I just wanted to confirm since I have never done any of this before.

Yes for the Okto, but I would not depend on it for streaming content with DRM (requiring a HDCP chain to play) from sources like Netflix or Apple TV.

The problem is in legally extracting digital audio from these streams without down mixing it to 2 channels and/or not downsampling. Authorized apps can only do it through a HDCP enabled chain. Without it, the streamers are required to mix down to 2.0 and/or down sample. AVRs work because they support HDCP and appear as HDCP compliant devices downstream. They can extract audio and convert to analog inside the box which is allowed. This is difficult to do within a PC.

 

[Flak]

Unfortunately, I didn't have the time to read this interesting thread in full but I imagine that this link might be useful for some:
https://live.dirac.com/config-file-settings/

 

[Vasr]

Unfortunately, I didn't have the time to read this interesting thread in full but I imagine that this link might be useful for some:
https://live.dirac.com/config-file-settings/

Thanks. Is this new?

Is there a document which explains in some detail what the settings mean more than the above and which use cases it applies to?

For example, I have no idea what Add 10db LFE Gain actually does, to use it correctly and figure out where it solves the problem I faced.

Is this something that it does during measurement only, or when playing content through it or both? Does it do it after correction or before? How does it avoid digital clipping if you play the surround sound from a movie which is recorded with LFE at -10db according to standard but can reach up to 0db (or at least -6db) in digital encoding?

The problem I faced in testing is that the measurement phase was done with the signals generated by DL where the sub signal was not cut to -10db relative to others. So if the downstream AVR was set to boost by 10db after conversion to analog for something coming on the LFE channel, then it would measure too high and Dirac would generate a large attenuation filter. When recorded content was then played through the DLP (with the attenuation in the recording), the LFE was too low. So the boost has to be disabled in the downstream device for measurement and enabled for content playing. This switch wouldn't help and wouldn't be used.

I don't see any use cases for this switch that would leave it on for any content played through it that won't run into problems.

Understanding the above and doing it correctly requires a lot from the user unless these scenarios are explained and why it is necessary to have good documentation. This is a very disappointing aspect of Dirac for the price it charges.

A problem with multi-channel with LFE is that the -10db attenuation is only used in HT content standards, not say for example multi-channel music mix from an SACD.

So, some of the better pre/pros or AVRs allow a boost setting based on the encoding coming in. So, it is boosted when it receives non-PCM (say Dolby or DTS encodings) multi-channel but not when it comes in as PCM (as most music content without LFE attenuation would). But this is irrelevant to the all PCM use required of Dirac in a PC since it needs to be decoded before use. So, not sure how to reproduce that AVR behavior when doing room correction in the PC if using it for both HT and non-HT content.

 

[Flak]

The problem I faced in testing is that the measurement phase was done with the signals generated by DL where the sub signal was not cut to -10db relative to others. So if the downstream AVR was set to boost by 10db after conversion to analog for something coming on the LFE channel, then it would measure too high and Dirac would generate a large attenuation filter. When recorded content was then played through the DLP (with the attenuation in the recording), the LFE was too low. So the boost has to be disabled in the downstream device for measurement and enabled for content playing. This switch wouldn't help and wouldn't be used.

Yes, you're correct... and it's often impossible to disable the LFE gain on the AVR.
A solution would be to attenuate the subwoofer measurement sweep only during measurement and have an additional flag for that.
It has been addded to the wish list so it may be available in the future

 

[phoenixdogfan]

Yes for the Okto, but I would not depend on it for streaming content with DRM (requiring a HDCP chain to play) from sources like Netflix or Apple TV.

The problem is in legally extracting digital audio from these streams without down mixing it to 2 channels and/or not downsampling. Authorized apps can only do it through a HDCP enabled chain. Without it, the streamers are required to mix down to 2.0 and/or down sample. AVRs work because they support HDCP and appear as HDCP compliant devices downstream. They can extract audio and convert to analog inside the box which is allowed. This is difficult to do within a PC.

JRiver can extract multi-channel from Netflix, it is HDCP enabled, is it not? And Netflix supports surround sound on the PC with their desktop app.

 

[192kbps]

Software like Jriver make things simpler (at a cost) but it is a walled garden approach. It is like a soft-AVR on its own and they have done a good job. But everything has to be done from inside JRiver.

My experiment being described in this thread is to create a general solution that will work for any content player on the PC one may want to use and/or don't to pay for JRiver. For people who may want to do this instead of buying an AVR or going through a commercial solution like JRiver in a walled-garden approach.


The latency here has nothing to do with the power of the processor. Any software decoding of even the simplest codec like AC3 is very slow. But if you have a single player doing both audio and video rendering, then they have control over both to keep them in sync. If you have a weak CPU, it requires more time to buffer ahead and start. All media players do this - Kodi, MPC-BE, VNC, etc.

If you are using pass-through of the encoded audio to a preamp processor then there would be no decoding and so no latency issues of course. But if you have pre/pro available, the case for paying for JRiver is weaker. YMMV.

Can you help me?

5.1 for CS:GO

two DBR62+Infinity RC252+ two JBL A130 Is that okay? What else do I need, an X3600H?

www.audiosciencereview.com

 

[Neddy]

I have been weighing getting the 8-channel Okto DAC since I have a multi-channel amp and a 5.1 system currently. If I wanted nothing more than HT playback for multichannel, can I connect the PC USB output to the Okto USB input, run software as you described for doing crossovers for the speakers that are small, set per channel levels in the Okto, and playback streaming sources including Netflix, ATV+, etc.? It seems like this should be possible but I just wanted to confirm since I have never done any of this before.

Here's how I've done that. Not for the feint of heart:

With the OktoDAC as my output DAC, I use an Oppo 103 for surround decoding of Mch Bluray and HDMI (AppleTV) inputs, then outputs that as 8ch analog to a Motu 8Pre ADC->USB.
That sends surround decoded 8 ch analog via USB to the PC which shows up as an input on JRiver.

Works great - tho it took seeming months to get JR/PC to recognize it reliably.
Seems (and is?) overcomplicated, but it works well enough to host Movie Nights every week.

(Oppo 103 also sends HDMI video to projector via wireless. Complicated? Nah Lotta gack, tho.)

This just how I did it - I already had the Oppo103 (wish I'd upgraded to a 205 when they were 'cheap'!) and the wireless HDMI system, so it only took the addition of a fairly inexpensive, but pretty decent Motu ADC to implement it all.

This also allows me to use JRiver as my music PC in addition, with just OktoDAC as an output device.

Still haven't tried the JRiver/Netflix multicnannel via browser thing yet...have Netflix on AppleTV anyway.

The nice thing about 5.1 would be that the Okto would leave 2 extra channels for bi-amping or multi-subs.
I'm running 7.1 so need to split out Okto main channels further for bi-amping.

Best wishes!

Edit: this also allows me to accomodate the 10db variance in LF (technically managed by Oppo decoder) via the Analog input gain adjustments on the Motu ADC, so solves that for streamed or disc content as well. Works great for movies but ripped SACDs or BRs on the PC can still be a problem, but minor.

 

[prescient]

Resurrecting an old thread: When setting up Dirac Live VST plugins is there any reason to not the use equalizer APO instead of setting up more software as indicated in the directions? My low bar of testing is that it is still outputting audio on the surrounds.

Signal path is: VM Banana -> Equalizer APO (within EAPO: Bass Management -> Dirac Live VST) -> VM Banana -> Motu mk5 ASIO

 

[XpanD]

Resurrecting an old thread: When setting up Dirac Live VST plugins is there any reason to not the use equalizer APO instead of setting up more software as indicated in the directions? My low bar of testing is that it is still outputting audio on the surrounds.

Signal path is: VM Banana -> Equalizer APO (within EAPO: Bass Management -> Dirac Live VST) -> VM Banana -> Motu mk5 ASIO

Last time I tried it EAPO's plugin support was still too limited to do it that way -- I had very distorted audio and the interface did not respond to any input. I don't think there's been any updates to that part of it since, but I'd be happy to be proven wrong there!

Running through Element and Voicemeeter Banana right now and that setup has been fine, though it is a bit cumbersome. (and Voicemeeter only works at 44.1KHz on my system for some reason, anything higher and I get garbled audio)

 

[prescient]

Last time I tried it EAPO's plugin support was still too limited to do it that way -- I had very distorted audio and the interface did not respond to any input. I don't think there's been any updates to that part of it since, but I'd be happy to be proven wrong there!

Running through Element and Voicemeeter Banana right now and that setup has been fine, though it is a bit cumbersome. (and Voicemeeter only works at 44.1KHz on my system for some reason, anything higher and I get garbled audio)

FWIW the Dirac VST plugin seems to be working OK in the EAPO chain. I can turn the VST on/off in the configuration manager and see the impact to FR for the front two channels via REW.

Frequency response at the listening position looks cleaner with the VST on, but I have yet to check surrounds due to the REW two channel limitation. I might try copying the REW signal to the surrounds to see if the VST Is working properly on all channels.

The only distorted audio that I have noticed is when identifying audio channels through the windows 5.1 config manager. I sort of assume that this is due to some weirdness with Banana Meeter.