Zoli posted Jul 14 '17, 09:58:
Yesterday i made a samplerbox with a rpi3. I installed the 2017-07-10 release.
I can import soundfonts from usb, its great, the latency too. I use it with an Alesis VI49.
Thank you for your work!
Unfortunately i have an issue. Some sound get stuck regularly (not the same, always different). It sounds after i let go the corresponding key. The sound only stops when i push the touched key again.
It happens very often, about every 1-2 minutes.
This is never happened with the same keyboard, when i used with a DAW on my notebook.
Anybody have solution for this issue?
Petar posted Jul 18 '17, 01:12:
I had the same bug using SamplerBox for a gig at the Folle Journée de Nantes last February.
Fortunately, it happened a couple of times only during the rehearsals, not the concerts.
A fix for this would be great. I would really appreciate to be able to use SamplerBox live again.
Zoli posted Jul 18 '17, 20:39:
I found this tip, in a youtube video comment section:
"To avoid "sticking" of keys, it is necessary to switch the USB controler in the USB1.1 mode. Open cmdline.txt file on your SD-card, then add dwc_otg.speed=1. Like this:
dwc_otg.lpm_enable=0 dwc_otg.speed=1 console=ttyAMA0,115200 kgdboc=ttyAMA0,115200 console=tty1 root=/dev/mmcblk0p2 rootfstype=ext4"
Now, i have no time to try this. But i'm afraid, the usb 1.1 mode will affect the samples loading time :/
HansEhv posted Jul 19 '17, 21:33:
Well, loading will be damn slow, I'd say unusable. I tried because of similar issues, however my problem is somewhat longer ago and disappeared when changing from pyaudio to sounddevice - which is in the new SB's like the one you seem to use. You can check to be sure however.
There is an in-depth discussion on http://www.raspyfi.com/anatomy-of-a-pi-usb-audio-quality-and-related-issues-on-pi/
This locates the issue within the PI firmware handling of audio. There is also a hint that the issue reveals itself when using USB1.1 devices behind a 2.0 hub (the PI one). Thus the advice to set the hub to 1.1 mode.
In other words: it might be influenced by using a 1.1 usb dac (audio device) or 1.1 midi interface or so. Principally no problem as those speeds are high enough for their purpose, but here it may be a problem.
Just a hint; I cannot check this :-(