On Sat, Jan 28, 2017 at 07:27:21PM +1100, Damien Zammit wrote: Hi! Last time I spoke with someone from RME (that was probably back in 2012 or 2013), they were unwilling to give us their USB drivers (let's say to maintain their competitive edge) nor to document the protocol (why would they, the driver is free to change it at any point in time). That said, I see three options: 1. Get an EXBOX.UMA. 32in/32out ADAT/MIDI/USB. The company is super helful, provided all the specs and enabled me to write a working driver: https://plus.google.com/+adiknoth/posts/PddEdbm7PNj The product is discontinued, but if you want to polish the driver a bit, I can lend you mine indefinitely. Should be fully functional, only the control software could use a decent UI (I tried HTML5 talking to an HTTP-to-MIDI server in Go but then had no time to finish it - cmdline is ready, though) 2. Reverse-Engineer the RME protocol. Probably lots for work. Basically what folks did when FFADO started, but it's cumbersome to interpret individual bits in a long bitstream just to spot the individual samples. And even then it'd be another six months to write the driver. 3. Talk to RME (I can give you a contact) and try to get hold of their driver sources or a protocol spec. Both is very unlikely, especially since they were repeatedly unwilling to reveal anything that involves the new TotalMix. The other option is to forget about USB2 and look at other connectors like Thunderbolt (ultimately external PCIe) or Ravenna/AVB. And last not least, there's an ALSA driver for Dante somewhere. Speaking of Thunderbolt: if you just want 4xADAT, you can always put an RME RayDat or one of the Marian cards in an external Thunderbolt enclosure. HTH -- mail: adi(a)thur.de http://adi.thur.de PGP/GPG: key via keyserver