Do higher TPL or LPBs improve MIDI recording?

Help, Support & Bugs
21

You’ve misunderstood dblue here. See https://forum.renoise.com/t/do-higher-tpl-or-lpbs-improve-midi-recording/45134 please. If you are recording !without! quantization, with the note delay column enabled, a resolution of 1024 PPQN is more than enough in most cases.

Changing the LPB does nothing at all to Renoise’s internal resolution. Renoise’s internal resolution are samples. LPB only changes the resolution the events are written into the pattern - and thus are played back afterwards when played back from the pattern.

So recording a pattern with an LPB of 32, then shrinking it back to LPB 8 does not improve timing.

Instead of trying to look for a solution for a problem, try to describe your problem as detailed as possible please, so we can replicate it here. When there is a bug and it can be replicated, it will be fixed. Else it’s really hard to do something here.

22

So recording a pattern with an LPB of 32, then shrinking it back to LPB 8 does not improve timing.

taktik, but exactly this is NOT true! Really. I really don’t know how to describe it better, maybe due my lack of sophisticated skills in English…

Please try my test case I uploaded and do the following steps:

  1. Load test case song. Disable metronome, disable quantization. Enable compensation. Set latency very low (e.g. 8ms)

  2. Now record the rimshot (key C#3) on track 2 along to the kick, exactly for one pattern length. TRY TO REALIZE WHICH NOTE didn’t had much offset while live playing.

  3. Let the pattern loop play and immediately listen to the recording. Usually you will hear at least one note that totally differs (in position) from what you just played (If not, then just re-record, or try higher latency)

  4. Now set LPB to 32, expand x4 using dblue’s pattern resizer tool. Clear track 2.

  5. Repeat steps 1-3, this time with LPB 32. You wont hear any timing glitch anymore.

  6. Now resize pattern /4 (resizer tool), LPB to 8. The timing is still nice.

23

For a more obvious demonstration about that timing problem, check out my latest update of “TunedShortcuts”:https://forum.renoise.com/t/topic/43378

It’s simply amazing how much better you can record at LPB 32. The shortcut will temporarily expand pattern, and shrink on disabling record again. Could be even better.

Please check this out, to get what I mean. Thanks.

24

So I now conclude here my investigation finally:

  1. Using enabled midi compensation, the recorded input will be placed too late in comparison to the live played.

  2. This effect is the stronger the higher the latency

  3. Using very high lpb >= 32 for recording will fix this problem a bit. Also the “random note too early” disappears

  4. Back shrunken pattern to original lpb still have a much better timing regarding the too late position

  5. This about some milliseconds, so best noticeable if testing with short impulses (like in test case above)

25

I cannot hear anymore any issues (in 3.1 RC1) so far… :slight_smile: Also live quantization again works like it should.