On Sun, 27 Apr 2008 22:56:08 +0200 Rene Herman <rene.herman@xxxxxxxxxxxx> wrote: > On 27-04-08 22:47, Sergei Steshenko wrote: > > > On Sun, 27 Apr 2008 22:30:46 +0200 > > Rene Herman <rene.herman@xxxxxxxxxxxx> wrote: > > >> Makes no sense. Setting the sampling rate has no meaning outside of the > >> action of playing or recording. > > > > ALSA is needed only to playback or record using a soundcard (a number of > > soundcards in more complex cases). > > > > And it is my choice which sample rate on the card(s) to use - it is me who > > decides on quality <-> CPU load <-> memory bandwidth relationships. > > > > So where in the above you see no sense ? > > In the decoupling of setting the rate and playing/recording. Please reread, > I already explained well enough. For me it makes perfect sense to decouple playback/recording on the one hand and sample rate on the other. This is because I am using a modern OS with multitasking, and I know that each task may deal with audio data having various sample rates, while all the applications need to access the same sound device. For this exact reason, to avoid mess, I'd prefer to mandate the device sample rate (within its physical capabilities, of course) and use SW tricks (resampling) to let the applications really communicate with the device. > > >> Note that mixing is not (historically) an expected feauture either > > > > ??? > > > > Very strange. I well remember SoundBlaster (16 ?) - it had a mixer in Windows. > > You are now confusing "a mixer" -- that part of a soundcard with the sliders > and the switches and where you can mix in analog inputs -- and stream mixing > in the sense of dmix. I was ofcourse talking about the latter and I can > assure you your SB16 didn't do stream mixing. It doesn't matter in this case, since there was possibility to mix, say, analog inputs from CD, line, microphones and digital PCM streams - all coming to the same card. It doesn't matter, since I don't care in which adders and in which multipliers (CPU ones or sound card ones) mix = k1 * signal1 + k2 * signal2 + ... kN * signalN operation (mixing) is performed. I as end user do _expect_ to be able to mix. I do not want to know as much as possible the internals of my computer - even though I assembled it myself. While I'm working with desktop apps all sound sources and sinks are equal to me, the whole machine is an audio processor for me. > > Rene. > Regards, Sergei. ------------------------------------------------------------------------- This SF.net email is sponsored by the 2008 JavaOne(SM) Conference Don't miss this year's exciting event. There's still time to save $100. Use priority code J8TL2D2. http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone _______________________________________________ Alsa-user mailing list Alsa-user@xxxxxxxxxxxxxxxxxxxxx https://lists.sourceforge.net/lists/listinfo/alsa-user
