On Mon, 1 Sep 2014, Len Ovens wrote:
> So my idea is that:
> - in time with the word clock a stream of data is sent over the ethernet.
> - it is less than the length of one word clock pulse
> - The data includes:
> x channels of audio formatted as consecutive AES3 (like MADI)
> An end of audio tag
I should add in here any audio related data such as MIDI, transport
control, OSC, etc.
> a data tag that tells how much data is within the frame
> regular data packets perhaps encapsulated in some form.
> - both ends look to the OS like both an audio device and an ethernet port
> - One end is responsible for breaking network data packets apart
> - the other end puts them back together.
A while ago, I got my wife an atom based MB with two cores and it happens
to have to ethernet IFs on it as well. It works okish. This is the one
Intel product that does not have an open grapphics driver. So my wife
finds the video she watchs tends to be slow. She uses it for kareoke and
to watch content in her native tungue. I originally bought it with the
idea it could in future replace my server (two NICs) where display beyond
text is not needed. (in fact, I use serial port to an old Ampex terminal)
Anyway, it is way beyond most small boards and has linux and x on it as
well as an audio IF. This is enough to try to do this with to figure out
if the idea can work and what the transport latency is (the internal audio
IF latency is not very good... -p64 minimum with -n3). But audio device
stuff is down the road anyway.
-- Len Ovens www.ovenwerks.net _______________________________________________ Linux-audio-user mailing list Linux-audio-user@email-addr-hidden http://lists.linuxaudio.org/listinfo/linux-audio-userReceived on Tue Sep 2 04:15:02 2014
This archive was generated by hypermail 2.1.8 : Tue Sep 02 2014 - 04:15:03 EEST