Open atskler opened 1 year ago
The log files: 11433-konsole-output.txt 11433-mixxx.log
The problem does not happen at 44.1k.
In any case which results in a crash, we need a backtrace from you: https://github.com/mixxxdj/mixxx/wiki/Creating-Backtraces Thank you
i'd love to work on this issue or help and collaborate with whoever is working on it
@Aryan-Mishra24 I don't think anybody is working on this right now. As a first step towards investigating/fixing this, you can you try to reproduce this issue as outlined in the bug description.
In any case which results in a crash, we need a backtrace from you: https://github.com/mixxxdj/mixxx/wiki/Creating-Backtraces Thank you
Here it is:
Thank you, could be Rubberband, though I'm not sure since it could be mixxx fault passing invalid pointers or too large buffer sizes.
```cpp
Thread 15 (Thread 0x7fffc6e746c0 (LWP 4738) "mixxx"):
#0 0x00007ffff72d4929 in RubberBand::FFT::inverse(double const*, double const*, double*) () at /lib64/librubberband.so.2
#1 0x00007ffff72e5483 in RubberBand::R3Stretcher::synthesiseChannel(int, int, bool) () at /lib64/librubberband.so.2
#2 0x00007ffff72e6427 in RubberBand::R3Stretcher::consume() () at /lib64/librubberband.so.2
#3 0x00007ffff72ec080 in RubberBand::R3Stretcher::process(float const* const*, unsigned long, bool) () at /lib64/librubberband.so.2
#4 0x0000000000a8e5a2 in EngineBufferScaleRubberBand::deinterleaveAndProcess(float const*, long, bool) (flush=false, frames=246, pBuffer=
I also noticed the sudden pitch change recently and I always use 48 kHz. IIRC my personal branch is based on 2.4
I suppose a condition block for 96kHz sample rate is the way to go we'll need to make sure that the engine switch is done safely to prevent crashes or playback speed issues. This might involve stopping the playback, switching the engine, and then resuming the playback It's hard for me to Locate the source file responsible for handling the Keylock/Pitch-B Engine
Apologies for taking 4 hours to reply it was midnight in India when i got the ping but luckily I'm an early bird
@Aryan-Mishra24 Don't worry :tea: no one is expecting this to be resolved in hours
What is the path to the files for the keylock feature
The original ControlObject [Master], keylock_engine
is created in src/engine/enginemaster.cpp but the value changes are computed in src/engine/enginebuffer.cpp
Just search src/ for "keylock_engine"
bit off more than i could chew , choked, coughed it out, made the outline for the " engine switch is done safely to prevent crashes or playback speed issues. This might involve stopping the playback, switching the engine, and then resuming the playback" but then got lost as to how to properly implement that functionality and what are the relevant functions to that operation @ronso0 was right " no one is expecting this to be resolved in hours" and neither should i
Bug Description
At samplerate 96kHz changing the engaged Keylock/Pitch-B Engine during playback speeds up the playback or crashes Mixxx, whatever happens first. Sometimes many switching required for the crash to happen.
The issue on video: https://www.youtube.com/watch?v=CY3o2nKhyCs
Version
2.4
OS
openSUSE Tumbleweed