• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

An evolving journey - DIY Surround on a Mac

Julf

Major Contributor
Forum Donor
Joined
Mar 1, 2016
Messages
3,004
Likes
3,998
Location
Amsterdam, The Netherlands
Level 4: This is the hobbyist/enthusiast level where the level 3 equipment is too limited for extracting the best audio and video. This is where my kind of setup comes in. Obviously, this needs a lot of tinkering and getting hands dirty. Not for everyone. Linux and Apple are not even in play here given the lack of necessary software but the results are several degrees above what the other levels can do for combined music and movie consumption.

What software does Linux lack for "extracting the best audio and video"?
 

LumbermanSVO

Member
Joined
Dec 31, 2018
Messages
50
Likes
71
I hear ya @LumbermanSVO, I really do.
Remember Apple's "Front Row"? (https://en.wikipedia.org/wiki/Front_Row_(software))
Oh for something that simple to sit in front of all media and be easily remote controlled?!
Maybe JRiver's Media Centre is exactly that?

I'm familiar with OBS and NDI, I've played in this space for live gig streaming and hope to have a crack at Resolume VJ'ing sometime with live video feed mixed in! If you have the expertise then of course go for it. Personally, I find it a little too heavy duty for a no-fuss home set up.

MotU are great audio interfaces, I too am tempted, especially as I could use the TRS outputs to feed my JBL LSR305's etc.
A cheaper alternative might be this from MiniDSP: https://www.minidsp.com/products/opendrc-series/opendrc-da8
It has enough processing power to accept some FIR processing on its channels, plus it's half the price of a MotU.

If Front Row was still supported on hardware that can output 4K HDR, I'd just need to solve the audio portion of the puzzle. But it's not, and people stopped porting it to newer operating systems a long time ago. That leaves me with an AppleTV, I don't use any other playback devices in my theater. It works well, software/firmware updates work flawlessly, and it has no problem pulling content from my iTunes server.

My issues with consumer AVR's is the size, heat, and super crappy UI's. I just want it to break out the audio and give me unmolested analog outs. I don't want to tell it if my speakers are "large" or "small" and have to google what the actually means on that particular model. I have a stack of amps, so I don't need internal amps generating heat and making the unit 6RU tall. I don't want a stupid ugly OSD popping up every time I adjust the volume, I don't want sleep timers. All that junk just to get outputs with questionable quality. A 1RU tall unit that breaks out the audio, has quality line level outputs, and does volume control, would be amazing.

NDI works pretty much flawlessly, but OBS is still a little buggy, it REALLY doesn't like when the input signal is lost, so you'd need to set your source to never turn off. This morning I had a different thought though, there are HDMI audio decoders, but you still have the volume control issue. Perhaps an HDMI audio extractor might be a better route, then run a digital line to your Mac, run it through AU/DAW/audio processor and bob's your uncle. I think a Harmony remote can control volume on the Mac.

I think I have a 1080p extractor laying around I can test with...

I have all pro amps, so the MOTU is VERY tempting, the MiniDSP looks neat, but the limited taps per channel is a bummer. I already have the speakers and amp channels for a fully active 5.2 setup with 3-way mains, I just need to figure out processing, the MOTU has all the output channels I need in 1RU of space.
 

audimus

Senior Member
Joined
Jul 4, 2019
Messages
458
Likes
462
What software does Linux lack for "extracting the best audio and video"?
Almost everything in my set up.

Linux is fine at the Kodi, Squeezebox level but not much beyond theory for the next level. It mostly serves ideological wars. Once you start customizing the Linux to do what you want then you land up with highly specialized distro of your own that is difficult to maintain that makes Windows maintenance look like a first world problem.

The Well-Tempered Computer
is a good site that documents audio related sites across different hardware without getting into ideological debates.
Getting low-latency bit-perfect high res audio out of a Linux box is not easy because the Pulseaudio/Alsa architecture was never designed for it. These are the equivalent of the Windows shared mode audio that everyone bypasses on Windows if remotely interested in good quality audio. Mastering Jack and getting it to work with upstream apps limits what you can do. There is no low latency software equalizer like Equalizer APO that you can just feed the REW generated filters and have it do what it is supposed to do. Most such projects have been abandoned long time ago. There is no equivalent of Voicemeeter (over Jack) if you want higher level interface to virtual sources and sinks to route your audio based on processing needs between multi-channel and stereo. Commercial DRCs require a Mac/Win box to configure.

These are just a few of the practical difficulties encountered when you actually start to get your hands dirty. Linux can be adopted as an underlying platform as Emotiva has done but then that is not practical for most people to do on their own.

Similar issues exist for video to get the best out of a Linux box which is why I switched.
X11 really limits what you can do with modern graphics cards. Apple did the right thing by not using any of that stuff and writing their own graphics processing.

There is no equivalent of madVR if you want to use a good graphics card to do the kind of video processing (upscaling or downscaling in chroma and image, removal of scaling or compression artifacts, HDR tone mapping, etc) that is better than most TVs/AVRs or Oppos.

At a more basic level, Nvidia cards still have an advantage for HTPC use because of their lower noise/heat profile but you get caught in the ideological wars between the open source community and Nvidia’s proprietary drivers. When Nvidia decided to no longer support VDPAU and switched to a new architecture for current cards, Kodi developers would no longer support it and so you lost the ability to do things like hardware decoding of 10bit HEVC which burdens the CPU too much and have to get CPUs with the latest hardware decoding but CPUs are not as easy to upgrade as video cards when new standards come out. Things like that distract from what you want to do for the kind of goals at this level.

I could go on...

Windows maintenance is the least of the problems at this level because you are not trying to use it as a general purpose desktop with zillion third party software subject to security issues, etc. So, I leave the brand zealotry and ideological debates to others and focus on getting results in practice.
 

Fitzcaraldo215

Major Contributor
Joined
Mar 4, 2016
Messages
1,440
Likes
632
I have been a JRiver fan for about 10 years. It satisfies my needs very well. I am interested in classical music, and it delivers excellent music in 7.1. Tagging and cataloging is excellent, though it requires manual attention. But, I watch TV and Blu-Rays on my TV monitor with it, as well.

The system is simple. A PC is the heart of it. It feeds the audio via USB to a ExaSound Mch DAC, the amps and speakers, and the TV via HMDI. There is a 52 TB NAS hooked to the PC via Ethernet for my extensive Mch music SACD and PCM collection and my video collection. There is also a TV cable card tuner via Ethernet delivering full Mch sound which is well integrated with JRiver. There are also optical drives in the PC with CD, DVD, Blu-Ray capability.

JRiver is excellent for control and calibration over all sources. Audio is exquisite and bit perfect, from MP3’s to CD’s to high rez in Mch. Video is similarly excellent, using the superb MadVR reproduction system.

I need no audio processor - not a prepro or an AVR. I am using Dirac Live in the PC, which JRiver interfaces well with. Consequently, the core of the apps in the PC is JRiver and Dirac, that’s it. And, I like to use an iPad via JRemote control to select my recordings from my collection.

I am using Win7. If I saw other advantages to other OS’s, I would use those. But, I cannot find any, and often there are restrictions to non-Windows apps. But, yes, Apple and Linux are both supported by JRiver.
 
Last edited:

Julf

Major Contributor
Forum Donor
Joined
Mar 1, 2016
Messages
3,004
Likes
3,998
Location
Amsterdam, The Netherlands
Once you start customizing the Linux to do what you want then you land up with highly specialized distro of your own that is difficult to maintain that makes Windows maintenance look like a first world problem.

Only if you don't know how to use version control, scripting and automation.

Getting low-latency bit-perfect high res audio out of a Linux box is not easy because the Pulseaudio/Alsa architecture was never designed for it.

Why does low latency matter (except in a live performance situation)?

I could go on, but yes, seems getting Linux to do all the very specific things you seem to need does take some effort. So does tweaking windoes in a way that survives version upgrades.
 

audimus

Senior Member
Joined
Jul 4, 2019
Messages
458
Likes
462
Only if you don't know how to use version control, scripting and automation.
Then it becomes worse than maintaining Windows. Been there done that.

Why does low latency matter (except in a live performance situation)?
In HT applications, you run into lip sync issues in the chain, especially when the delay is data dependent. Combine that with latency of a software DSP or parametric EQ, combine that with the sampling/resampling constraints of PulseAudio/Alsa/etc.

I could go on, but yes, seems getting Linux to do all the very specific things you seem to need does take some effort. So does tweaking windoes in a way that survives version upgrades.

There is not much to go on really unless you can post or point to a link that actually addresses all the issues I have raised and how it can be done in Linux without being a developer himself than just wave hands and claim it can be done. Otherwise, it is just a theoretical discussion.

Somethings like madVR type of processing are not even available in Linux. Maintenance of Windows is a talking point for those that are biased against it but anyone who has lived with Linux knows, a single update on Linux breaks can break things and does frequently. Having used both and lived with them for decades, this is more of an ideological debate than reality.
 

Julf

Major Contributor
Forum Donor
Joined
Mar 1, 2016
Messages
3,004
Likes
3,998
Location
Amsterdam, The Netherlands
There is not much to go on really unless you can post or point to a link that actually addresses all the issues I have raised and how it can be done in Linux without being a developer himself than just wave hands and claim it can be done.

Indeed. I was trying to acknowledge that there is probably no way to address all your issues within the specific constraints you postulate. I guess I, being a developer, have a slightly different perspective.
 

Sal1950

Grand Contributor
The Chicago Crusher
Forum Donor
Joined
Mar 1, 2016
Messages
14,073
Likes
16,609
Location
Central Fl
Then it becomes worse than maintaining Windows. Been there done that.
Somethings like madVR type of processing are not even available in Linux. Maintenance of Windows is a talking point for those that are biased against it but anyone who has lived with Linux knows, a single update on Linux breaks can break things and does frequently. Having used both and lived with them for decades, this is more of an ideological debate than reality.
What are ya'll involved in beyond using your HT's to view movies and TV, plus listen to music. Don't understand all the complications unless it's just become a hobby of its own. These things happen all the time, you like to take a few pictures and next thing you know you have a $50k darkroom. I'm a avid sport shooter and now have a fortune invested in reloading equipment. LOL
 

LumbermanSVO

Member
Joined
Dec 31, 2018
Messages
50
Likes
71
I bought the audio extractor I linked and it arrived today, I plugged it into my Mini with a toslink cable, and only get 2-channels of input no matter what I set the extractor to. So it looks like getting anything more than 20channels of inout isn't trivial.

I'll dig out my thunderbolt HDMI capture card later this week to test with.
 

Ride154

Member
Joined
Jun 11, 2020
Messages
9
Likes
1
Hi Guys, read this thread and this interests me also. Any progress? No expert what so ever but
how about Apple TV > HDMI > TV > HDMI > Mac Mini > usb > okto research dac8 pro > 8 power amps of your choice > speakers?
The dac8 was first made with a dsp inside and as I understand will return with a dsp.
And I’m thinking about using (maybe modifying for a case) hypex plate amps, which also has a dsp inside.
Here I read that Atmos is supported since El Capitan.
https://developer.dolby.com/platforms/apple/macos/overview/
Or am I missing something?
 

digicidal

Major Contributor
Joined
Jul 6, 2019
Messages
1,981
Likes
4,838
Location
Sin City, NV
Hi Guys, read this thread and this interests me also. Any progress? No expert what so ever but
how about Apple TV > HDMI > TV > HDMI > Mac Mini > usb > okto research dac8 pro > 8 power amps of your choice > speakers?
The dac8 was first made with a dsp inside and as I understand will return with a dsp.
And I’m thinking about using (maybe modifying for a case) hypex plate amps, which also has a dsp inside.
Here I read that Atmos is supported since El Capitan.
https://developer.dolby.com/platforms/apple/macos/overview/
Or am I missing something?

You're missing plenty in that chain. Primarily a way to extract the audio and separate the channels for output over USB to the DAC (without any more latencies than you've already got - or one hell of a lipsync delay I suppose). Of course, even assuming that wasn't an issue... you've only solved the problem for lossy (AC3) output of Dolby encoded source material... DTS will still be just as unplayable as ever as will lossless Dolby bitstreams, etc.

Sorry to rain on the parade. Unless I completely misread the charts for it, even Catalina only provides limited support and none for TrueHD which is required to avoid compression. If you're just providing E-AC3 (DD+) or AC3 (DD) then it's a compressed, lossy stream and almost all the improvements in the DAC will be lost. Definitely not an appreciable improvement (in quality, capability, or cost) than most AVR's... especially if you factor in all the potential points of failure in the chain (vs. just a single HDMI cable to the AVR).

That is unless you've found some device which will strip separate audio PCM channels out of an HDMI signal and pass it via USB... in which case you have found something everyone would be very interested in! ;)
 

Ride154

Member
Joined
Jun 11, 2020
Messages
9
Likes
1
As I said ... no expert whatsoever :) but very interested into ditching an AVR which is always the one who gets outdated and plenty of Mac gear in my house and office.
There isn’t a fun little program that can fix this?
I found Loopback from Rogue Amoeba and this fixed my problem with my Cambridge Audio AVR not having a ’party’ mode. Easy routing stereo in to 8 channels and some how overruling my AVR who kept pushing music into stereo mode.
 
OP
DownUnderGazza

DownUnderGazza

Active Member
Forum Donor
Joined
Jul 28, 2018
Messages
130
Likes
301
Location
New Zealand
I've a few tests I've been meaning to try out.
It seems Apple has integrated more Dolby Atmos capabilities across its line. So they have a license to potentially do something useful.

If for whatever reason one chooses to not go with JRiver or similar, there seem to be only two paths.
  1. Set-up basic 5.1, 7.1 or even Atmos (to be tested) via Apple's "Audio MIDI Setup" Utility.
    • This gives you basic channel allocation and relative gain at each output;
    • currently this does NOT give you Bass Management, time delays or EQ (PIR or FIR) - pretty severe limitations;
    • there does not seem to be any opening for a 3rd party to interface with this utility to provide 'advanced' configuration;
    • I did email Rogue Amoeba about maybe them taking on this enhancement / integration role, their response was 'interested' but their current tools are all stereo and not multichannel.
  2. Treat the Mac as a high end Audio DAW (Digital Audio Workstation) and use professionally available tools
    • This allows you to integrate the Pro-level content creation tools used to create our surround sound music and movies
    • There's Dolby (https://www.dolby.com/us/en/professional/content-creation/industry.html) and Auro (https://www.auro-3d.com/buy/) for starters, DTS (https://www.waves.com/bundles/dts-neural-surround-collection) and pretty much anything the pros use.
    • These options, and others, are the real deal but are NOT cheap.
    • There are a wealth of high quality plugins for things like linear phase EQ (https://www.musicradar.com/reviews/fabfilter-pro-q-3)
    • While Rogue Amoeba allows you to use these AU's (Audio Units), it is limited to stereo only
    • Therefore, the only option to integrate all this wondrous Pro-level capability is by running it all through a DAW. Apple's Logic Pro, Avid's ProTools or Reaper come to mind.
    • A proper DAW will allow detailed filtering for bass management, time alignment, EQ across all inputs and channels as well as Up-mixing and Down-mixing,
    • All input sources and output channels can be mapped and routed as desired and overall and relative volume levels can be controlled centrally (using a midi controller or an iPad as a remote control)
    • The down side of course is that these are pro level tools. Stable and reliable once set up, but not for the faint of heart or the weak of wallet.
The second option is one where we recreate a Pro-level multi channel surround processing studio.
Our audio 'quality' is therefore only limited by the quality of the audio interfaces we choose to adopt. Imagine if you will a bunch of top-notch balanced analogue and digital input channels for our various sources and say 16 channels of superb output channels, analogue or digital, for routing to our loudspeakers.
Exemplary performance, mind-boggling flexibility, mind-bending configuration control, and absolute state of the art end result.

Anyway, yep, still pondering my options...
 

digicidal

Major Contributor
Joined
Jul 6, 2019
Messages
1,981
Likes
4,838
Location
Sin City, NV
@DownUnderGazza - but won't you still be in the dilemma of having no realtime decoding and processing in those scenarios? In both you're going to have to rip/re-encode your streams in order to output separate channels sync'd to the video, correct? Hardly a "toss in a movie and use the remote to adjust the volume" scenario. Or my personal nightmare, my wife decides she'd like to watch a movie an hour after I arrived at the office and needs to be walked through the process. With the Oppo, she can go network to NAS and grab a .MKV and have no problems... but I can only imagine if the first step was: "Open the DAW and wait for the workspace to fully load, now go to projects..." :eek:

Technically, isn't the only real obstacle in all of this the fact that HDCP is present (and required) for all HDMI-output content players? So, all we would really need is a paradox: a licensed HDMI device which produces unencrypted separate audio streams in PCM format... which would invalidate said license by it's very existence (and make daddy Intel very angry with you). ;)
 
OP
DownUnderGazza

DownUnderGazza

Active Member
Forum Donor
Joined
Jul 28, 2018
Messages
130
Likes
301
Location
New Zealand
Fair points @digicidal, and yes totally agree that HDCP carries a huge amount of blame for our troubles. Intended to prevent piracy, but in reality thwarting and frustrating law abiding consumers since its introduction...

From my experience with DAW software (albeit limited, but hoping to learn more), audio processing in realtime is what they are built for. They usually include capabilities for controlling the sizes of buffers and setting target latencies. Both essential elements in dealing with 'monitoring' of real time feeds and also for handling 'lip-synch' with video content.

One thought I had is that introducing 'Dante' routing may help simplify getting the correct channels to the correct processing and outputs. Adding Dante allows adding everything from high-end Focusrite 'Red' interfaces for scalable input and output channels. Plus Dante routing software would help deal with picking up the audio streams from iTunes, QT Player or whatever video playing software is used.

And yes, unless the 'wife/gf/partner' is a certified pro-audio technician, then if anything went wrong the WAF (Wife Acceptance Factor) would plummet.

So going the 'Pro-Studio' approach is not without its own challenges, but it did occur to me as a means of completely bypassing the typically technically compromised AV/PrePro performance.
 
Top Bottom