INTERACT FORUM

Please login or register.

Login with username, password and session length
Advanced search  
Pages: [1]   Go Down

Author Topic: MC for speaker processor  (Read 5524 times)

Juniorjbl

  • Recent member
  • *
  • Posts: 43
MC for speaker processor
« on: November 29, 2015, 01:56:27 pm »

Would JR be able to use all of it's audio processing (DSP) functions if you were to use a MOTU 24Ai (into the computer) and a 24Ao (output) on the downstream side of an Atmos processor.

I ask this just in case we are no longer able to play/decode Bluray and HD Bluray disks in the future (DTS X and Atmos and future codecs) without buying a prepro.

I completely despise the thought of using a receiver and do not want to spend the $$$$$$ on some 10K prepro.

My goal here is to use JR and a MOTU 24Ao and do all active processing including XO's for my mains and surrounds. I can play BD's now so not an issue. But what about the future?

Also how much CPU do you need to do say 10 filters per channel and 18 channels?

Thanks for any info you might have. :)
Logged

Juniorjbl

  • Recent member
  • *
  • Posts: 43
Re: MC for speaker processor
« Reply #1 on: December 01, 2015, 08:37:20 pm »

Anyone?
Logged

mwillems

  • MC Beta Team
  • Citizen of the Universe
  • *****
  • Posts: 5234
  • "Linux Merit Badge" Recipient
Re: MC for speaker processor
« Reply #2 on: December 02, 2015, 08:52:39 am »

Would JR be able to use all of it's audio processing (DSP) functions if you were to use a MOTU 24Ai (into the computer) and a 24Ao (output) on the downstream side of an Atmos processor.

JRiver's ability to take a multichannel input is present at the moment, but not perfect.  On Windows you have a choice of an ASIO input driver (which I haven't used much) and a WDM input driver (which only supports up to 8 channels AFAIK).  So you'd need to go the ASIO route.  Mojave has some fairly serious Motu kit and uses the ASIO input driver pretty aggressively, so he may be able to comment, but I seem to recall an issue with multi-client ASIO drivers being a potential stumbling block. 

Quote
Also how much CPU do you need to do say 10 filters per channel and 18 channels?

That depends entirely on how you're doing your filtering.  If you mean conventional PEQ filtering, you don't need much CPU.  I run about 30 filters per channel on a 4 channel setup (120 total filters) and my PC typically doesn't even register increased cpu usage (i.e. it hovers at 2-3% like it does at idle). Admittedly it's an i7, but it's a five year old i7.  I would think any modern i5 would be plenty for 180 filters assuming you could get the i/o sorted out.

If you're talking about convolution, that's a different story entirely.
Logged

mattkhan

  • MC Beta Team
  • Citizen of the Universe
  • *****
  • Posts: 4232
Re: MC for speaker processor
« Reply #3 on: December 02, 2015, 09:13:50 am »

I don't have a motu but I have tried to use ASIO line in with more than 2 channels and jriver was unable to open such a device. My hardware & driver supports it so I don't know whether the issue is jriver, my asio driver, the OS or some combination thereof. If you get it to work then I don't know of a reason why you couldn't just route it through jriver's audio engine.

However I don't know how you'd maintain sync with the video stream when that is being done upstream in some prepro and video is routed separately to the display. Perhaps I'm missing something about your use case though?
Logged

mojave

  • MC Beta Team
  • Citizen of the Universe
  • *****
  • Posts: 3732
  • Requires "iTunes or better" so I installed JRiver
Re: MC for speaker processor
« Reply #4 on: December 02, 2015, 09:17:34 am »

I think a MOTU 16a is a more economical choice than a 24ao and 24ai unless you need the extra channels.

Yes, you can route all the audio into JRiver for DSP and then back out with the MOTU drivers. However, you have to use the ASIO Line-in which isn't easily accessible with a remote or app. You can leave the HTPC always on, though, and have a Zone on ASIO Line-in mode. Use the other Zones for playback from the HTPC.

You will have some latency on the audio, but should be able to delay video in the Atmos processor. You can also use a display with high latency, like the JVC projectors that have ~140 ms of latency.

This doesn't apply to you since you want to use active crossovers, but others might find it helpful. Another option to consider that I recently helped someone implement is to use a Yamaha CX-A5100 processor as the processor and DAC. This is cheaper than a 24ao/24ai yet still uses the 32-bit ESS Sabre chips. It also is rated for the same voltage output. It can do Atmos 5.1.4, but not 7.1.4.

The Yamaha is a lot like JRiver in that it can have multiple DSP Presets and you can even combine manual EQ with the YPAO -RSC room correction. It also has the most extensive set of measurements I've seen for integrating the Atmos speakers. You can access all the setting on the Yamaha with a computer and web application which makes setup easier than with other processors.

In the system I helped with, Dolby TrueHD content is bitstreamed to the receiver. It is then decoded by the receiver and has the appropriate DSP applied. All other content on the HTPC (music, DVD, Blu-ray with DTS-HD) is processed by JRiver and uses JRiver's DSP and is then sent on to the Yamaha as PCM with no further DSP applied.
Logged

mwillems

  • MC Beta Team
  • Citizen of the Universe
  • *****
  • Posts: 5234
  • "Linux Merit Badge" Recipient
Re: MC for speaker processor
« Reply #5 on: December 02, 2015, 09:19:21 am »

I don't have a motu but I have tried to use ASIO line in with more than 2 channels and jriver was unable to open such a device. My hardware & driver supports it so I don't know whether the issue is jriver, my asio driver, the OS or some combination thereof. If you get it to work then I don't know of a reason why you couldn't just route it through jriver's audio engine.

I think you may be running into the multi-client issue that mojave has previously mentioned (unless it works just fine with 2 channels).  

I guess the issue was that JRiver as an input and as an output presents to the audio device as two different programs (clients) trying to open it in ASIO mode at the same time and most devices lack the ability to accommodate multiple clients at once.  Steinberg has released a generally usable multi-client driver that works with most ASIO hardware and that's the usual recommended workaround.
Logged

mattkhan

  • MC Beta Team
  • Citizen of the Universe
  • *****
  • Posts: 4232
Re: MC for speaker processor
« Reply #6 on: December 02, 2015, 09:52:24 am »

I think you may be running into the multi-client issue that mojave has previously mentioned (unless it works just fine with 2 channels).  
no, it works fine in 2 channels (and I already use the steinberg multiclient driver as well)

Logged

mwillems

  • MC Beta Team
  • Citizen of the Universe
  • *****
  • Posts: 5234
  • "Linux Merit Badge" Recipient
Re: MC for speaker processor
« Reply #7 on: December 02, 2015, 10:13:15 am »

no, it works fine in 2 channels (and I already use the steinberg multiclient driver as well)



That is a puzzlement then as it sounds like Mojave can confirm the ASIO line in does support multichannel.  I assume you've tried the line in apart from the ASIO driver (the line in support pre-dates the ASIO driver and may have different functionality).
Logged

mojave

  • MC Beta Team
  • Citizen of the Universe
  • *****
  • Posts: 3732
  • Requires "iTunes or better" so I installed JRiver
Re: MC for speaker processor
« Reply #8 on: December 02, 2015, 10:19:20 am »

I just tested and the ASIO Line-in is crashing JRiver with more than 2 channels selected. I even set audio output to Null to isolate the input on the ASIO driver. I'll post it as a bug. I'm pretty sure I've used it multi-channel before.
Logged

mwillems

  • MC Beta Team
  • Citizen of the Universe
  • *****
  • Posts: 5234
  • "Linux Merit Badge" Recipient
Re: MC for speaker processor
« Reply #9 on: December 02, 2015, 10:26:46 am »

Puzzle solved!
Logged

mojave

  • MC Beta Team
  • Citizen of the Universe
  • *****
  • Posts: 3732
  • Requires "iTunes or better" so I installed JRiver
Re: MC for speaker processor
« Reply #10 on: December 02, 2015, 10:30:08 am »

I don't alway check my MOTU driver, but did during testing. I was pleased to find a driver/firmware update which was released yesterday.

Logged

Juniorjbl

  • Recent member
  • *
  • Posts: 43
Re: MC for speaker processor
« Reply #11 on: December 02, 2015, 06:07:26 pm »

Thanks for all of the responses!! :)

Please forgive me if I should know this but, If I were to play an Atmos disk in a normal JR HTPC with a 4 layer BD internal drive (I have an LG) what would I get out if anything?

If this is a no go then I guess getting a BSS BLU160 would be my other option and no more JR for the theater. I do have a separate 2 ch room so music in the theater is most likely never going to happen.

Is Atmos/DTS:X really all that? I'm not sure I could even get the extra speakers mounted anywhere in my theater. My ceiling is only 6' 11" and there is a floor above. :( 

Thanks
Logged

mojave

  • MC Beta Team
  • Citizen of the Universe
  • *****
  • Posts: 3732
  • Requires "iTunes or better" so I installed JRiver
Re: MC for speaker processor
« Reply #12 on: December 02, 2015, 09:18:13 pm »

Thanks for all of the responses!! :)

Please forgive me if I should know this but, If I were to play an Atmos disk in a normal JR HTPC with a 4 layer BD internal drive (I have an LG) what would I get out if anything?

If this is a no go then I guess getting a BSS BLU160 would be my other option and no more JR for the theater. I do have a separate 2 ch room so music in the theater is most likely never going to happen.

Is Atmos/DTS:X really all that? I'm not sure I could even get the extra speakers mounted anywhere in my theater. My ceiling is only 6' 11" and there is a floor above. :( 
Thanks
Atmos/DTS:X rely on speaker separation to provide greater immersion. If you have limited height, then I don't think you will get as much benefit. Some are using flush mounted ceiling speakers on height limited rooms. I'd prefer to see someone have a higher quality 5.1 or 7.1 channel system than try to support Atmos.

JRiver can only decode the first three substreams of an Atmos encoded TrueHD track. You can bitstream from JRiver to a processor or receiver that has Atmos decoding and can decode additional substreams. Here is a good description of what is happening:

Quote
The TrueHD stream is structured as a series of substreams. The first substream contains the entire soundtrack as a 2-channel mix. So if you're a TV manufacturer, you don't have to use a fancy decoder that unpacks all 7.1 channels only to downmix them to 2 channels for playback over the two TV speakers. Instead, you use the most basic TrueHD decoder that simply unpacks the first substream.

The second substream contains the audio (data) and instructions (metadata) that, when combined with the first substream, results in a discrete 5.1-channel track. The third substream contains the audio and instructions, when combined with the first two substreams, to recreate a 7.1-channel track. The fourth substream for 9.1-channel soundtracks, and so on, with additional substreams allowing for reconstituting mixes with more discrete channels. However, in an Atmos soundtrack, Dolby chose to use the fourth substream for the audio objects (data) and the information about them (metadata).

If your AVR has an older TrueHD decoder (or you use JRiver), then it won't recognize the information in the fourth substream of an Atmos track. It will simply decode the first three substreams, recovering the original 7.1 mix of the soundtrack. Newer TrueHD decoders will recognize the fourth substream and unpack the contents. Since the backwards compatible 7.1 portion of the soundtrack contains all the audio in the mix, you've got copies of the objects sitting in those channels. This is a problem, since you don't want objects in two places at once. So the data in the fourth substream is inverted and used as a cancellation signal to delete that same information in the 7.1 channels. The end result is the original bed channels and audio objects of the encoding master
Logged

Juniorjbl

  • Recent member
  • *
  • Posts: 43
Re: MC for speaker processor
« Reply #13 on: December 02, 2015, 09:30:24 pm »

Ok cool!

So if I am understanding this correctly, I can continue on the path I have been on (5/7.1) and continue to enjoy all that I have had in the past.

My entire system is based on active filtering with 9 amplifiers so I prefer to keep it that way.

Thanks Mojave. I really appreciate you took time to answer my questions! :)
Logged
Pages: [1]   Go Up