[linux-audio-dev] Viable multi-channel audio I/O boards for Linux ?

Mark Knecht mknecht at controlnet.com
Fri Nov 1 14:41:01 UTC 2002

   You are correct! I'm sorry. I went back and looked at the email I
received. He actually said you could buy two PCI cards and link them under
ASIO. My mistake. I'm was not at all attracted to that solution for the
obvious reasons of cost, syncing, heat and interrupts on more cards. Yuck.

   However, if you're correct and RME is just moving standard 1394 packets,
then even with their own protocol you could just use a standard 1394 hub
between a single adapter card and two breakout boxes. That could work and
would only run about $50. It would require that RME be able to keep track of
multiple boards on a single 1394 bus, but that's no big deal really. I
should ask Matthias at RME. Worth looking into for this application for
sure. (P.S. - I do 1394 for a living...)

   I'll have to look in my notes about the competitors to the AI-3. There
are a couple now that are lower priced I think. RME supports 96K across
pairs of ADAT lines. I wonder if they support 16 22K channels across a
single channel? ;-)


-----Original Message-----
From: linux-audio-dev-admin at music.columbia.edu
[mailto:linux-audio-dev-admin at music.columbia.edu]On Behalf Of Paul Davis
Sent: Friday, November 01, 2002 10:57 AM
To: linux-audio-dev at music.columbia.edu
Subject: Re: [linux-audio-dev] Viable multi-channel audio I/O boards for
Linux ?

>And the DigiFace/MultiFace are very nice products. It was my understanding
>from one of RME's vendors that at least under ASIO both the DigiFace and
>MultiFace support multiple breakout boxes with a single PCI or cardbus
>adapter card. I was keeping this in the back of my mind for the day when I
>can dump Pro Tools completely. Does this work under Linux or has anyone

its simply false. there is a single 1394 connector on the back of the
PCI card and a single connector on the cardbus dongle, so you can only
connect a single breakout box to one interface unit. whoever told you
this either knows something that i don't or they are misinformed/confused.

this is the main area i am aware of where the MOTU units look more
appealing than the RME ones. SMPTE sync is the other one.

>These RME products use an RME proprietary interface that looks and smells
>like 1394, uses 1394 cables, but isn't 1394, so you can't hang a 1394 audio
>drive on it like I use on my systems. (I do not know if there is any danger
>in making a mistake when you plug the wrong device into their cable...)

it is 1394. it just uses a proprietary mid-level protocol for the
communication. you don't have any access to the 1394
firmware/controller on the card, but if you did, you could write a
driver for it and use it as a general 1394 interface. alas, RME saw
the (commercial) danger of putting everything in the external box and
having you use a $50 1394 interface. as a result, we get this from
them and for the same reasons, the MOTU stuff from MOTU.

>All of this seems pretty high end for what I thought the requestor was
>originally looking for at 22KHz, but I know of no answer for that market
>specifically. I think the AI-3 or one of the other devices similar to that

that's true, i forgot the AI-3 units. they are a good deal for this
kind of thing.

>I wonder if there's a business to be made in developing hardware for apps
>like this? I'm looking for a good business idea like that.

given the tiny size of the computer audio market as it is, it seems
unlikely to me. what i think would be a great success is a properly
priced 24 channel A/D-D/A converter. +10/-4 switchable balanced
inputs, ADAT I/O for the digital side or some 1394 standard once it
emerges as an option. that would be great, and would have an instant
home in many studios of many different sizes. 16 is enough for many
people though, and the creamware and rme solutions might have this all
sewn up.


More information about the Linux-audio-dev mailing list