Hardware device for experimenters

[j_e_f_f_g]

https://www.kickstarter.com/projects/42 ... y-interact

[giulio_moro]

Hey good to see this has been posted here already. I am one of the developers and was looking for the right section where to post this! Let me add a few more details if you don't mind.

Our lab at Centre for Digital Music at Queen Mary University of London has made a new high-performance embedded platform called Bela (http://bela.io) which is designed for creating digital musical instruments and interactive audio systems.

This platform runs on the BeagleBone Black and comprises a custom hardware and software environment which is open source and LGPL licensed. We use the Xenomai extensions to run the audio thread at a priority higher than the Debian Linux kernel. At the same time we are using one of the two onboard Programmable Realtime Unit (PRU) to read and write to the audio codec, analog DAC and ADC and the GPIOs. All the sensors are sampled at audio rate.
The system allows to have block sizes as small as 2 samples, achieving audio round-trip latencies below 1ms and analog round-trip below 100us.

We've been developing this for the past two years, and we just launched it on Kickstarter today (and it is already funded):
https://www.kickstarter.com/projects/42 ... y-interact

You can either write C++ code, or use Enzien Audio's Heavy compiler to run Pure Data patches.

A few examples of projects can be found here http://bela.io/made-with-bela/ or at the kickstarter page. More projects, examples, stretch goals to come. Tech specs below.

Cheers,
Giulio

Bela tech details:

Runs on the BeagleBone Black (1GHz ARM A8 with 512MB ram), with a custom cape which provides audio and sensor I/O.
Audio: 16-bit stereo audio I/O at 44.1kHz
Audio power output: 2x 1W 8ohm speaker amplifiers (available when powered from DC jack)
Analogue In: 8x 16-bit analogue inputs at 22.05kHz
Analogue Out: 8x 16-bit analogue outputs at 22.05kHz
Digital channels: 16x digital GPIO at 44.1kHz or 88.2kHz
Analogue I/O is also software configurable to give 4 channels at 44.1kHz or 2 channels at 88.1kHz

More info:
http://bela.io
http://twitter.com/BelaPlatform
https://www.facebook.com/belaPlatform/

[analoq]

similar idea: http://www.axoloti.com

[sadko4u]

Audio: 16-bit stereo audio I/O at 44.1kHz

16-bit audio isn't serious now for DSP, you should better upgrade to 24-bit ADDA converter.

Analogue In: 8x 16-bit analogue inputs at 22.05kHz
Analogue Out: 8x 16-bit analogue outputs at 22.05kHz
Digital channels: 16x digital GPIO at 44.1kHz or 88.2kHz
Analogue I/O is also software configurable to give 4 channels at 44.1kHz or 2 channels at 88.1kHz

Currently it's too weak and does not pass to the standards of audio processing.
If you could make all channels at least 24 bit + 44.1/48 kHz sample rate, it could be a killing feature for the low price.
Now it's just a toy based on ARM processor.

[gimmeapill]

Not too impressed either by the HW for serious audio work, but that Xenomai framework is very interesting.
Is there any practical application for audio on x86?