Closed cryptozoy closed 3 years ago
I tried to reproduce this issue:
Please give a step-by-step instruction how to reproduce your issue - thx.
It happens only under Windows 10. In Fedora 34 Linux everything is normal.
It works with e8e27c9. Thank you.
@spam-receiver Thx for reporting - update is available in unstable now. How does the program behave the other way round? Starting with CH1 mode, can you go up to 24 and 30 MS/s? I cannot check on a win system with real HW, only Demo Mode that starts always with 2 channels. So I leave this issue open 'til your positive feedback.
You right. Now one channel-mode can't go upper 15MS/s after start until on/off any channel. I forgot to check it out.
Ok, but better this way round. But nevertheless I will look for a correct behaviour if I can spare some time.
Tl;dr: It will take some more time...
I created some test builds with extensive tracing output and it seems that the described behaviour is caused by a race condition between the startup default setting (always started with 1 channel in the past, now defaults to 2) and the persistent setting from the last shutdown. This sets always the correct (stored) number of channels, but misses to update the max sample rate limit according to the actual number of channels - it still uses the one from the startup value. The communication between the core and UI of the program is based on Qt's signal/slot mechanism that obviously adds some unpredictable delays.
Under Windows 10 after starting the program, it is allowed to turn on samplerate above 15MS/s until you turn off and turn on again any channel.