I’ve always had an occasional problem with phrases not playing correctly. This time I thought it’s time to post on the forum since it’s degrading my quality of life!
Here is an example of a looping pattern with three chord voices triggered with Z02 Z03 Z04: Dropbox - phraseerror.mp3 - Simplify your life
4: almost ok… strange attack suggests something is happening with the envelopes?
I always thought it was due to my ‘soundcard stutters’ (buffers) that I’ve had a problem with, but it turns out this error is also audible when rendering to file. Perhaps even more so, as even the normal Gxx effect seems to get timing issues when rendering.
Is there some workaround or some mistake i’m making? I’m hoping to get to the bottom of this, finally.
version 3.4.2 - win x64
example file (edit: added some missing fade-out effect number)
oddity2.xrns (8.5 KB)
The XRNS doesn’t sound like the MP3 file!? The XRNS file sound different as the notes don’t fade out making the discrepancy hard to hear.
I’ve tried add the missing fadeout, make five patterns and render it out and all patterns sounds the same to me and I can’t hear anything like what’s going on in your MP3 file. Hear my render.
Same for me. It sounds different to me than in your MP3. And every run sounds the same to me.
W11, x64, R 3.4.2. Tested DSound, WASAPI and ASIO.
Yeah, it’s strange. Any sound card settings shouldn’t have effect on a rendered file, so it’s a bit difficult understanding that inconsistency. Perhaps then it’s due to some other setting in preferences.
EDIT: added the missing fadeout effect in example.
Sync mode is internal? What’s you audio settings?
Here is mine:
Yes, internal sync mode and same settings as you (slightly higher latency setting here, but that doesn’t seem to matter).
I have found something interesting when rendering with DirectSound selected as Device Type versus ASIO.
Check out this image where I put the ASIO render on top and DirectSound on the bottom, ASIO consistently render the transients across all patterens but DirectSound appear to be losing transients beginning pattern 3…
See the bottom right cluster where it’s clearly missing transients:
So somehow the Audio device type has an influence on the render
Interesting. I get more consistent results with ASIO indeed! (I preferred directsound as I won’t have to deal with cumbersome stutter/buffer issues.)
With ASIO, the timing issues with glide effect is also gone.
The difference is shocking, if this is indeed true then never use DirectSound when rendering.
Or even when tracking. I blame my lack of groove on this
Thanks a lot for investigating!
TL;DR But tried the above xrns on macos Renoise 3.42 with different latencies (8ms vs 80ms). It appears to me that the sound is now completely dependent on latency (but it should not at all). At some point sub divisions or whatever are wrongly calculated. Maybe this was introduced with that recent bugfix…
Same here (macOS 10.15.7, Renoise 3.4.2). The difference between latency settings is extremely noticeable, both on playback and render.
Yeah I can confirm too, so this is also happening just to a lesser degree with the Focusrite ASIO if you try max the latency.