|
From: Jonathan W. <jw...@ph...> - 2011-04-19 23:18:41
|
Hi Rolf > Hardly any problems on all rates up to 96kHz. > We're now trying to take it to the next level: running at rates up to 192kHz. > Unfortunately this does not work at all: ... > : > ... but before I go on, I was just wondering if anyone has ever managed to > get FFADO running with a DICE (or any other) device at high samplerates > (176.4 kHz & 192 kHz) and if there? Although it's about as different from a DICE device as it can get, I've had MOTU interfaces running at 192 kHz. I've also had my experimental driver for the RME FF400 running at 192 kHz during testing, although I haven't run that for extended periods yet. > If yes, any suggestions, do's/dont's? Nothing that's not common for ffado setups in general. Things like: - ensuring jackd has the required permissions to run with realtime scheduling, - minimising (if possible) the number of devices sharing the interrupt with your firewire controller, - avoiding vendor binary graphics card drivers if possible Since you have solid behaviour at 96 kHz though I would guess that you've already seen to most (if not all) of these. About the only other thing which springs to mind is to question what kind of latency you're trying to push out of the system. If you're running with very low latency (that is, low value arguments to "-p" and "-n" when starting jackd) then perhaps this may be the reason your system chokes at 192 kHz. To this end it might be helpful if you could post your jackd command line and/or the jackd settings you're using. Keep in mind that in terms of latency time, moving from 96 kHz to 192 kHz while keeping your "-p" and "-n" fixed will halve your latency figures. Regards jonathan |