I wasn’t sure which category this post fit in. I searched and couldn’t find a topic on it already. So I apologize if this is miscategorized or redundant to an existing topic.
I am not a developer. I am interested in developing a dedicated hardware system for Harrison Mixbus32c that is highly optimized to run just that program and perhaps a couple of other software programs (such as Pianoteq and Hydrogen, either in plugin format or as the standalone native app).
I don’t need the computer to do anything else other run these programs and serve as a hub for a single or dual display (does not need to be touchscreen), a hardware digital mixer control surface, external USB DAC/ADC and perhaps an occasional USB midi controller.
How trivial and practical would it be, for example, to install Elk Audio OS onto a RPi4 with the HiFiBerry for DAC/ADC as a system dedicated to running Harrison Mixbus32c with recording/playback from an external USB drive? Or if that would be underpowered for this purpose, do the same with a System76 Meerkat + USB Audio card?
I just wanted to “look before I leap.” I have significant (but still amateur) experience with tuning vanilla distros for low latency audio (ref: https://pop-planet.info/forums/threads/pro-audio-pop.249/post-2299), but am finding that I really need to just create a dedicated hardware platform if I want any sense of reliability and maximum performance.
Thank you! This project looks very exciting and a step forward for Linux-based audio. I am hoping that it can have a positive political influence on upstream optimization of Linux OS design for low latency audio work, e.g., similar to Apple OS X circa 1999 (ref: https://www.cse.unsw.edu.au/~cs9242/10/lectures/09-OSXAudio.pdf)
Without having used Harrison Mixbus, from what I see it is a closed source commercial DAW based on Ardour.
Given that, it is not in its current state usable with Elk OS:
In very few words, the principle of Elk Audio OS is that you use our (headless) DAW Sushi, which directly interfaces with our custom audio driver. It can host VST2/3/LV2 plugins, with very high performance, under the following two main conditions:
That the plugin is cross-compiled for the platform. So, you need the source code.
That the plugin does not have GUI dependencies. Instead you interface with it over our DAW’s remote OSC/gRPC API.
As far as Mixbus goes, neither of these two main points apply.
Nor could Sushi replace it, since it is a live DAW: it doesn’t have audio/midi recording, editing and playback functionality.
As it stands, Elk Audio OS is not really a Desktop Audio PC replacement. See it instead as the heart for a standalone instrument!
We keep a comprehensive repository of all open-source plugins that have been compiled for the platform, in case you haven’t come across the link before, here it is: https://github.com/elk-community
Thank you Ilias! Is this mainly for live performance, then, or do you see an application for this in a studio, e.g., as an outboard audio interface? In other words, is the focus of this project on live sound generation/synthesis, and the expected signal chain midi controller/signal > [Elk Audio OS hardware] > DAC > analog outs? Are there other signal chains that you expect to be relevant? Such as midi controller > DAW > USB midi out to [Elk Audio OS hardware] > USB digital audio back to DAW?
Is there a discussion of how this project differs or diverges, for example, from Zynthian? https://zynthian.org/
The most important distinction compared to Mod Duo, Zynthian, Bela, etc, is that Elk is not primarily a maker platform. With Elk, you can create an instrument with the features and performance of an industrial-scale musical device.
We have development kits for purchase, and encourage the open-source community to develop with and for these. You will see that both the software is open-source licensed (GPL3 Affero), and the hardware is open-source to the point that you can manufacture your own if you want.
We certainly do not discourage that makers and musicians tinker with the platform, in fact many wonderful ideas may come out of such use. But the main objective is that the platform is used to develop finished, commercially viable, complete instruments.
It can be a multi-effects device processing audio input, a synthesizer, drum machine, drum-brain, etc. But a developer/team will have to use the platform to make that, it is neither of those things out of the store, only a development kit, towards making either.