~ Russell > On Oct 21, 2014, at 1:15 PM, Len Ovens <len@xxxxxxxxxxxxx> wrote: > >> On Tue, 21 Oct 2014, Harry van Haaren wrote: >> >> Hey all, >> I have a HDA ( cat /proc/interrupts tells me so anyway: snd_hda_intel ). >> Its my built-in laptop soundcard, and it works pretty well, down to about 44.1kHz, -p128 >> -n3. That's ~8ms, which is acceptable IMO. >> I've read on mailing list / internet somewhere* that spec for the HDA is pretty >> open-to-interpretation, so I think the exact hardware / chipset would need to be tested >> to get true results. > > I have gotten 48k -p64 -n3 on mine. Jack will crash if I try -p64 -n2, but -p128 -n2 works. > From my reading, HDA is (like AC97) more of a bus spec than a sound spec. It seems to deal with connecting the sound from the HW to the internal bus. The HDA bus runs at 48K (like AC97) but does not expect the HW to be 48k (as AC97 did) and has methods for transporting audio with other sample rates than 48K. > > In my opinion, 128/2 is just on the edge of useful for live (guitarix for example) use. 256/2 is annoying already. (No I do not play play really fast or anything like I get about the same on the same hardware. Using KxStudio, I can occasionally get 64/2. Unstable for any kind of serious live or recording activity. With - rt kernel can do a little better. Agreed... Playing guitar / bass / synth with anything less than 128/2 is not really usable due to its anoying-ness factor for me personally. The 2nd you slap the headphones on anyone else, usually the very first comment is 'um, there's a delay?!'. R _______________________________________________ Linux-audio-user mailing list Linux-audio-user@xxxxxxxxxxxxxxxxxxxx http://lists.linuxaudio.org/listinfo/linux-audio-user
