Calling .stop() on a sound whilst the browser window is blurred does not actually stop the sound about sound HOT 6 OPEN

Thanks for the detailed bug. That makes sense. Will look at it soon

from sound.

I'm having some issues creating a fork right now, but this snippet seems to have fixed this issue in sound v4 in my project:

import { webaudio } from '@pixi/sound';
webaudio.WebAudioInstance.prototype.stop = function(){
    if (this._source)
    {
        this._internalStop();
    }
    this.emit('stop');
}

from sound.

I'm seeing this same issue, though my repro steps are slightly simpler

1. call pauseAll()
2. call stop() on a specific sound instance
3. call resumeAll()

I believe what's happening is that when pause() is called on a sound, it's _internalStop() is getting called, which is nulling this._source, so when stop() gets called while the sound is already paused, it checks if this._source exists, it doesn't, and then nothing happens. I think if we were to emit the 'stop' event any time stop() gets called, regardless of whether _source exists, Sound should trigger _onComplete() and clean up any instances still in the _instances array.

I've only looked at the v4 code, but here's the instance stop:

and here's the _onComplete:

from sound.

I'm having the same issue, is there an ETA for the fix? Version 5.2.3.

Thank you!

from sound.

I did some digging and problem seems to be lying in refreshPause calls.
@nicolasalt Can you try changes from PR #270 a few times locally? It might be a candidate for fixing the issue.

from sound.

I've tried but your PR is on top of pixi/sound v6.0 which depends on pixijs v8, while I'm using pixijs v7.2.4.

from sound.