Has anyone here noticed an issue with AudioListener eventually dominating the performance graph after running for long periods? I have an app with an AudioListener, and after leaving it running for 6 hours, it gradually slows down… if I then pause it and .remove the audioListener from the scene… fps suddenly goes back up to 60
notice the “linearRampToValueAtTime” is dominating the frame…
That method is used inside three to interpolate audio listener parameters, Inside updateMatrixWorld… And it uses a THREE.Clock deltaT internally to select the interpolant values…
I posted a synopsis of my questions on that chromium bug I linked… perhaps getting some more google eyes on it will yield more information… I’ll keep looking into it… Thanks again, @Mugen87 !