[EDIT]
Case solved. The keyboard was sending high resolution CC Midi in addition to the normal note on/off. When i switched that off in my kbd settings, everything was fine!
Thx to the rescue team here at Dicourse Planet Z
Removed the ORAM referance from the subject as it seems unrelated to the case.
Case update: On my machine (V5.1 kit) i have more XRuns running Pianoteq (PT) than one should expect. On maximum overclock and these PT settings:
I have around 8 XRuns pr minute of hard playing.
Higher settings on PT and lower or no overclocking results in higher amount of XRuns
This performance is sub par compared to other users experiences
[/EDIT]
Hi all
I have switched from Vangelis to the latest Oram staging. And have lots of XRuns when playing Pianoteq. XRuns before on Vangelis was only there when i was not running RPI5 at maximum overclocking. I have done some testing but will report on that when i can sum it up a little more. Now i would just like to hear if i’m alone in this or if there are others?
I can’t help with the answer to that because I don’t use Pianoteq. But I wonder if this could be connected-related to the earlier version which used maximum overclocking by default and the later, but still several updates back, when we discovered that not all RPis could handle that and changed to less overclocking by default. Sorry about the lack of detail - I just have a vague memory of a problem with this somewhat recently, like in the last year.
I then removed other chains (2-4), but no help there.
I checked that i was overclocking, and i was:
I then tried to find solutions on Discourse, and tried the following (maybe in correct order)
Lowering the sample rate of Pianoteq. I lowered it to 16k, and this might decrease the amount of XRuns with 5%? But not solving in any way.
I changed the buffer size in audio settings to 512, this helps, and the XRuns are gone. But the if i reload my snap (se above) XRuns are quite seldom but the Zynthians heart is often orange and read.
I the read the Jofemodo had succesfully run Pianoteq with full settings on Pianoteq
and a 128 buffer (i think), so i tried that - first with a 512 buffer. Extremely many XRuns resulting in long breaks in sound.
Then i tried, just for the curiosity, to run the same pianoteq settings with 128 buffer, and of course everything was horrible about that
Then i went to my workshop and assembled some plywood for something complety different for a while, came back here and wrote this
Perhaps going back to Vangelis and seeing if the problem actually goes away? And if so, does returning to Oram staging bring the problem back - since they’re supposed to be identical…
I am (mostly) on vangelis (with some zynmixer changes thrown in for good luck) and do not experience xruns when playing lots of notes with sustain on PT configured with all features enabled. Is it just PT or do you have other processors / engines?
[Edit] I am using RPi5 with no hardware, US-16x08 at 48000/128 and no overclocking
Its afaik just PT. I will test if its different going back to Vangelis.
Is there any place where i can monitor which processes are using how much CPU. Maybe something is competing with PT, though i do not know what it should be!
Here is the dubug log after updating to Vangelis:
Jan 25 14:15:12 zynthian startx[8569]: Building XImage…done
Jan 25 14:15:12 zynthian startx[8571]: ZynCore: Setting-up 20+4 x Zynswitches…
Jan 25 14:15:12 zynthian startx[8571]: ZynCore: Setting-up 2 x Zynswitches in RBPi GPIO…
Jan 25 14:15:12 zynthian startx[8571]: ZynCore: Setting-up Zynpots => 4 x PEC11 …
Jan 25 14:15:12 zynthian startx[8571]: ZynCore->gpiod_start_callbacks: Callback thread created successfully
Jan 25 14:15:12 zynthian startx[8571]: Started libzynaudioplayer using libsndfile-1.2.0
Jan 25 14:15:13 zynthian startx[8571]: Started libzynmixer
Jan 25 14:15:13 zynthian startx[8571]: Started libzynseq
Jan 25 14:15:13 zynthian startx[8571]: Started libzynsmf
Jan 25 14:15:17 zynthian startx[8571]: WARNING:zynthian_engine_pianoteq.fix_pianoteq_config: Pianoteq detected previous crash (1)
The last line was also there when updating to staging. I do not know if this is significant!
More or less same number of XRuns i must admit:
Jan 25 14:17:29 zynthian startx[8571]: WARNING:zynthian_autoconnect.cb_jack_xrun: Jack Audio XRUN! =>count: 1, delay: 372.0us
Jan 25 14:17:34 zynthian startx[8571]: WARNING:zynthian_autoconnect.cb_jack_xrun: Jack Audio XRUN! =>count: 2, delay: 348.0us
Jan 25 14:17:37 zynthian startx[8571]: WARNING:zynthian_autoconnect.cb_jack_xrun: Jack Audio XRUN! =>count: 3, delay: 333.0us
Jan 25 14:17:41 zynthian startx[8571]: WARNING:zynthian_autoconnect.cb_jack_xrun: Jack Audio XRUN! =>count: 4, delay: 61.0us
Jan 25 14:18:03 zynthian startx[8571]: WARNING:zynthian_autoconnect.cb_jack_xrun: Jack Audio XRUN! =>count: 5, delay: 94.0us
Jan 25 14:18:27 zynthian startx[8571]: WARNING:zynthian_autoconnect.cb_jack_xrun: Jack Audio XRUN! =>count: 6, delay: 72.0us
Jan 25 14:18:36 zynthian startx[8571]: WARNING:zynthian_autoconnect.cb_jack_xrun: Jack Audio XRUN! =>count: 7, delay: 54.0us
Jan 25 14:18:37 zynthian startx[8571]: WARNING:zynthian_autoconnect.cb_jack_xrun: Jack Audio XRUN! =>count: 8, delay: 76.0us
Jan 25 14:18:37 zynthian startx[8571]: WARNING:zynthian_autoconnect.cb_jack_xrun: Jack Audio XRUN! =>count: 9, delay: 131.0us
Jan 25 14:18:41 zynthian startx[8571]: WARNING:zynthian_autoconnect.cb_jack_xrun: Jack Audio XRUN! =>count: 10, delay: 75.0us
But it was not that bad couple of days ago, and i’m on low settings. So i should like to think that my machine could do better?
I really appreciate all the help i’ve got and will get, but i’m doing some woodwork for some time now, and then there is some family stuff later, so i’m not in a hurry
