On Tue, 9 Jan 2018 10:40:47 -0800 (PST) Len Ovens <len@xxxxxxxxxxxxx> wrote: >On Tue, 9 Jan 2018, Will Godfrey wrote: > >> On Tue, 9 Jan 2018 11:26:19 -0600 >> "Chris Caudle" <chris@xxxxxxxxxxxxxxx> wrote: >> >>> Going through the network to jack adapter layer adds additional latency, >>> so I'm not sure exactly what the purpose of running separate jack servers >>> at low latency would be compared to just running a single server with >>> higher latency settings. >>> >> I seem to remember hearing somewhere that the jack server can't make use of >> multiple cores, but surely multiple *severs* could each be on their own core. > >Introducing a network layer adds more latency than creating a jack client >with buffer that talks to two jackd servers. > >I believe jackd2 can use more than one core for non-dependant strings of >clients. That is, a jack aware application could split two jack strings by >calling itself two clients and buffering audio between them. However, the >purpose of jackd is to provide known latency. Adding workarounds to make >use of more cores destroys that. > > >-- >Len Ovens >www.ovenwerks.net Interesting. Maybe I miss-read Jonathan's original post, but I was under the impression that he was using three instances of Yoshimi that had their own MIDI streams, rather than a general audioIn->audioOut. If these were ALSA MIDI then they would be buffered at that level rather than the audio. Does that make a difference? Would a single jack2 server be able to put the audio on different cores? -- Will J Godfrey http://www.musically.me.uk Say you have a poem and I have a tune. Exchange them and we can both have a poem, a tune, and a song. _______________________________________________ Linux-audio-user mailing list Linux-audio-user@xxxxxxxxxxxxxxxxxxxx https://lists.linuxaudio.org/listinfo/linux-audio-user
