Linux-Based Low-Latency Multichannel Audio System

Bring your custom casings, arduinos, raspberry pi's, axoloti's and other creative hardware projects

Moderators: MattKingUSA, khz

Post Reply
henrix
Posts: 1
Joined: Thu Feb 11, 2016 3:18 pm

Linux-Based Low-Latency Multichannel Audio System

Post by henrix »

Hi there,

Together with my professor, I've developed a multichannel audio system based on the AD1938 audio codec by Analog Devices and the BeagleBone Green (TI AM335X SoC).
The whole project is based on open source software.
To demonstrate the possibilities of the audio system, I've created a surround delay effect with the open source C++ library DSPatch by Marcus Tomlinson.
Moreover I've created an automatic test based on GNU octave to evaluate the audio system characteristis, such as latency, THD+N, DNR, crosstalk and frequency-response.
Image
The full article is published here.
Everybody who's interested and has some questions or feedback, feel free to contact me.
CrocoDuck
Established Member
Posts: 1133
Joined: Sat May 05, 2012 6:12 pm
Been thanked: 17 times

Re: Linux-Based Low-Latency Multichannel Audio System

Post by CrocoDuck »

Dude, I love this. I did some work on blind nonlinear system identification for my master project and I have plenty of Matlab (octave) code l am porting to JULIA at the moment. In fact, I was thinking to release it at some point.

It feels like this calls for some serious open source collaboration. As emerged in this thread, it would be cool to extend current Linux audio diagnostic tools, so that we can drive and keep updated the audio optimization techniques.
folderol
Established Member
Posts: 2069
Joined: Mon Sep 28, 2015 8:06 pm
Location: Here, of course!
Has thanked: 224 times
Been thanked: 400 times
Contact:

Re: Linux-Based Low-Latency Multichannel Audio System

Post by folderol »

Very interesting, and looks useful, but a couple for questions.

The frequency response looks strange for audio work. At the HF end -3dB looks to be about 60kHz, and at LF 5Hz. Isn't this likely to encourage problems with out-of-band input signals?

What are the band limits for THD and noise? Is it weighted?
The Yoshimi guy {apparently now an 'elderly'}
lilakmonoke
Established Member
Posts: 59
Joined: Sun Feb 10, 2013 9:47 am

Re: Linux-Based Low-Latency Multichannel Audio System

Post by lilakmonoke »

something like this would be ideal for my pure data based performance sampler. ive been looking for a similar platform but cant find anything ... i need at least 6 channels and want to avoid an external soundcard. any hints?
User avatar
sadko4u
Established Member
Posts: 986
Joined: Mon Sep 28, 2015 9:03 pm
Has thanked: 2 times
Been thanked: 359 times

Re: Linux-Based Low-Latency Multichannel Audio System

Post by sadko4u »

Hi! This is very interesting. 24 bit ADC with up to 96k sample rate is motsly all what we need.
Is it possible to run this device as a standalone?
Is it possible to apply in-place DSP and what do you use for applying DSP for the signal?
LSP (Linux Studio Plugins) Developer and Maintainer.
Dominique
Established Member
Posts: 68
Joined: Sat May 26, 2018 10:24 am
Has thanked: 3 times
Been thanked: 2 times

Re: Linux-Based Low-Latency Multichannel Audio System

Post by Dominique »

folderol wrote: Sun Mar 20, 2016 10:35 am The frequency response looks strange for audio work. At the HF end -3dB looks to be about 60kHz, and at LF 5Hz. Isn't this likely to encourage problems with out-of-band input signals?
Not necessarily as long we are talking of the analog part of the DACs and ADCs. In all analog audio circuits I know, the phase of the signal begin to change well before the amplitude when changing the frequency. This imply a good practice is to calculate the bandwitch for 1 decade more than what you want. As example, if you want the -3 dB at 10 Hz, calculate the capacitors and other components in order to get 1Hz at -3 dB. The same in the high tones. 1 decade more can be unrealistic for different technical or cost related reasons, but 1 octave more should not be a problem in most, if not all, the cases.

The extra bandwidth you will get will not be wasted because it will give you the right phase between the different frequencies of a complex musical sound, and that even at the low and high end. In the high end, I think most peoples will not ear the difference, but in the low end, it will give you a more precise, clearer and dynamical sound. This is even true with circuits with feedback, because when the feedback loop have less work to do, this always result into a better sound at the output.
Post Reply