On Sam, 2002-03-09 at 23:24, Christopher C. Chimelis wrote: > > On 9 Mar 2002, Michel [ISO-8859-1] Dänzer wrote: > > > I doubt xmms-cdread is broken at all, it works here with the OSS output > > plugin but not with crossfade (it doesn't matter if crossfade uses the > > OSS plugin or its own OSS support). The xmms-cdread code also looks > > sane, though a quick look at xmms-crossfade didn't reveal any obvious > > mistakes either, but it's considerably more complex than xmms-cdread. > > I haven't tried it in a few weeks, but last time I did, I only got white > noise (LE audio needed to be byteswapped/swabbed). Has this changed?
As I said, white noise with crossfade, perfect sound with OSS directly. > If not, then xmms-cdread is certainly broken since the code is just not > there to do the swabbing. True, but when the EQ isn't used, the byte order shouldn't matter as long as it's correctly advertised, should it? The output plugin should swap bytes if necessary. > > Speaking of tumbler sound, has anyone else noticed the following issues? > > > > - tendency to skip, most notably when using APT > > - random hangs (constantly repeating a short part) with SDL apps > > - some apps cause the driver to go into a weird state where all attempts > > to play sound fail, the only remedy is to unload and reload it > > > > Any known workarounds? > > I haven't seen any of the above yet. What do you mean by 'tendency to > skip'? Skip how? Short (up to about a second) pauses under heavy load, but also very short pauses with no load at all when doing certain things, e.g. often when using APT. The same only happened very rarely, under very high load (never with APT normally) on the Pismo. > As for the random hangs, this may be caused by high system loads and > happens on just about all of my machines when I saturate the bus... Again, this never happened with the Pismo, but now it happens in every SDL app I've tried on this TiBook/667. It seems like the SDL audio thread hangs with the sound looping indefinitely until it hits a timeout. Not what I'd expect as a symptom of bus saturation, but my expectations may be wrong. :) -- Earthling Michel Dänzer (MrCooper)/ Debian GNU/Linux (powerpc) developer XFree86 and DRI project member / CS student, Free Software enthusiast
