Inverting while playing ping-pong backwards
The trick with ping pong loops is to put the loop points at the peaks of the waveform, not at the zero crossings.
Flipping the phase would certainly be an interesting extra option, but it should not totally replace how ping pong currently works.
Oh! Right Thanks for the tip.
you could “fake” this type of operation by “unrolling” the loop - copy, paste, reverse - and flipping phase of the copy.
I normally also use the usual solution for the ping pong clicking problem, that involves placing the loop marker not onto a zero crossing but onto the apex of a peak, it takes some experimentation and experience to find suitable peaks though and not all samples are friendly to this operation. I find your take on this (inversion) very interesting!
Edit: I tried out your idea, it works really smooth…just duplicating the zero crossing defined loop section, reverse, invert, set loop to normal operation. Yay! Thanks for bringing up the question.
Nice to hear, will try that, too!
I thought about this a bit further: for pretty symmetrical waveforms your trick would work (placing the loop point at the peak), but what about waveforms like a saw tooth? If you would invert while playing backwards, a perfect sawtooth ping-pong loop was possible…
Btw. Upcoming bitwig 2.4 also inverts phase for backwards play in ping-pong.