At Wed, 8 Oct 2003 17:44:41 +0200 (CEST), Jaroslav wrote: > > On Wed, 8 Oct 2003, Takashi Iwai wrote: > > > At Wed, 8 Oct 2003 16:44:02 +0200 (CEST), > > Jaroslav wrote: > > > > > > On Wed, 8 Oct 2003, Takashi Iwai wrote: > > > > > > > > I'm still not sure, if we should handle these things in the driver. > > > > > Hardware is not designed in this way and we should describe hardware > > > > > in the driver as most accurate as we can. > > > > > > > > IMO, the master volume is a feature we mostly need. > > > > in the case of emu10k1, it's much easier to implement this on the > > > > driver (as a digital attenuator) rather than implementing it outside > > > > the kernel space (except for the lack of GPR space). > > > > > > Ok, but add more channels to "Master" so we have still a chance to control > > > values independently. > > > > now i'm thinking of a hierarchy like > > > > master (mono?) > > > > front(l/r) rear(l/r) center/lfe ... > > > > outputs... > > > > the advantage is that you can adjust all volumes with a single > > control. adjusting the level for each channel is done by the volumes > > below it. there is still a problem that the sensitivity is different > > between front and surround volumes (the former is ac97 and the latter > > is the digital attenuation). perhaps it's solved when we introduce dB > > hints. > > > > or, yes, instead of the top master, we merge all the lower layers as > > the master volume (either multi-channels or "master front", "master > > rear", etc.) > > > > perhaps we should create some templates of basic mixer designs suited > > to different hardware implementations. the confusion we're facing now > > seems to come from the different implementation even though the same > > control name is used. > > Yes, but again, why to try to solve these things in the kernel space? > There is no point to do it there and also we can modify simple API to use > hints written in lisp to handle the special cases - like covered in > this discussion.
the template i meant above is the basic designs of control naming, and i don't mean that this is the kernel issue. (well, in the case of "master" volume, we cannot handle it simply by abstraction but need a new control, though.) i believe it's just a question which is easier to implement. i agree that almost all cases can be solved by a higher abstraction, and i prefer that, too. Takashi ------------------------------------------------------- This SF.net email is sponsored by: SF.net Giveback Program. SourceForge.net hosts over 70,000 Open Source Projects. See the people who have HELPED US provide better services: Click here: http://sourceforge.net/supporters.php _______________________________________________ Alsa-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/alsa-devel
