Date: Sun, 27 Feb 2022 23:26:53 +0100 From: Florian Walpen <dev@submerge.ch> To: freebsd-multimedia@freebsd.org Cc: Hans Petter Selasky <hps@selasky.org> Subject: Re: New FreeBSD port ALSA to JACK MIDI, a2jmidid Message-ID: <82302170.0ko45tJjV3@z800> In-Reply-To: <586816bd-4332-4a0c-ab4b-b798588514ff@selasky.org> References: <a1f43e5a-5d1f-1980-e771-00a0064f74bb@selasky.org> <d55c36c9-1151-aea4-4a18-37010b4d3fc8@selasky.org> <586816bd-4332-4a0c-ab4b-b798588514ff@selasky.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Sunday, February 27, 2022 9:41:03 PM CET Hans Petter Selasky wrote: > On 2/27/22 21:36, Hans Petter Selasky wrote: > > On 2/27/22 21:29, Hans Petter Selasky wrote: > >> On 2/27/22 18:45, Florian Walpen wrote: > >>> @HPS: Any ideas about the latency introduced by alsa-seq-server / CUSE? > >> > >> Hi, > >> > >> What tools did you use to test? > >> > >> You might want to ktrace the applications in question to see what is > >> going on. > >> > >> Is this something I could test on my computer? > >> > >> --HPS > > > > Did you set real-time priority for alsa-seq-server and did you install > > the very latest version available from ports? Real-time and jitter are not realistic to cause such a delay - plenty of idle CPUs waiting for work. Latest from ports. > > Hi, > > One more idea: > > Can you make a FIFO "man mkfifo" and add that to the alsa-seq-server > instance? Then start jack_umidi on that FIFO aswell. What is the result? Shortcut where the device comes in, if I understand correctly? alsa-seq-server on FIFO, period 384@48kHz (8ms): Average latency: 16.28 ms (781.47 frames) Lowest latency: 15.12 ms (726 frames) Highest latency: 16.60 ms (797 frames) Peak MIDI jitter: 1.48 ms (71 frames) Average MIDI jitter: 0.20 ms (8.85 frames) alsa-seq-server on FIFO, period 192@48kHz (4ms): Average latency: 8.09 ms (388.46 frames) Lowest latency: 7.08 ms (340 frames) Highest latency: 8.35 ms (401 frames) Peak MIDI jitter: 1.27 ms (61 frames) Average MIDI jitter: 0.23 ms (10.55 frames) jack_umidi on FIFO, period 384@48kHz (8ms), same as before: Average latency: 11.93 ms (572.41 frames) Lowest latency: 8.02 ms (385 frames) Highest latency: 16.00 ms (768 frames) Peak MIDI jitter: 7.98 ms (383 frames) Average MIDI jitter: 1.99 ms (95.13 frames) jack_umidi on FIFO, period 192@48kHz (4ms), same as before: Average latency: 5.97 ms (286.44 frames) Lowest latency: 4.02 ms (193 frames) Highest latency: 8.00 ms (384 frames) Peak MIDI jitter: 3.98 ms (191 frames) Average MIDI jitter: 1.00 ms (47.32 frames) Additional latency is a bit better for alsa-seq-server, 4.3ms and 2ms. The results for jack_umidi are unchanged, since all MIDI events have returned by the time when JACK processes the next cycle (4ms or 8ms). The aconnect would be for shortcutting the a2jmidid ports, correct? Have to compile / install first. Regards, Florian
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?82302170.0ko45tJjV3>