[ardour-users] firewire raid boxes

Christian Schumann schumann at physik.uni-kl.de
Wed Mar 21 08:05:05 PDT 2007

On Wed, Mar 21, 2007 at 11:03:42AM -0400, Justin M. Streiner wrote:
> On Wed, 21 Mar 2007, Christian Schumann wrote:
> > So the bottom line of my experience: It works, if you take care about
> > the transfer limits and arrange your session that way.
> Thanks for the quick response.  I have a 250 GB Western Digital My Book 
> that I'm not using at the moment.  I've thought about putting that through 
> its paces in a test to see how well it would hold up.  What kind of file 
> system are you running on that drive?
I'm using an ext3 fs on it. Here's the test I did: I recorded 26 tracks
directly form the inputs of my Digiface to seperate tracks. Latency was
5.3 ms at 48kHz, 24bit. I tracked for 3 hours straight without xrun,
dropout or hd throughput problem. So I am confident that everything's
working ok.

HW/SW: Sony Vaio PCG-K215Z, 1GB RAM, Debian sarge, 2.6.17-rt8. Can't
tell the jack and libardour versions from here. RME Digiface/PCMCIA

> That's kind of what I figured.  In most cases, this is for recording live 
> shows, so they would be tracked 'dry'.  That way I don't need to worry 
> about the overhead of plugins during tracking.  If there are effects added 
> at FOH during the show that need to be punched up, I could take care of 
> that in the mix or take a split from the appropriate FX buss(es) from FOH 
> as additional tracks if I have the room.
ok, so then my test could be directly applied to your situation. Perhaps
you should also do a test like this (with a sine wave input or similar)
on your hardwar, just to be sure. But I'm confident it's going to work.


Dipl.-Phys. Christian Schumann   |Technische Universitaet Kaiserslautern
Mail: schumann at physik.uni-kl.de  |           Fachbereich Physik
Tel.: 0631/205-4842 (Office)     |     http://www.physik.uni-kl.de 
Tel.: 0631/205-4843 (Lab)        |
Fax.: 0631/205-3902              |    
Post: Erwin-Schroedinger-Strasse, D-67663 Kaiserslautern

More information about the Ardour-Users mailing list