# Time Stretching Problem

i’ve got a problem

i’m trying to time-stretch 9 bars voice loop - but i can’t do it properly
i mean it either plays slower or faster

8 bars, 4 bars, 2 bars loops get cut and stretched perfectly but 9 bars don’t

is there any solution?

here’s the example how i cut it:

samplevoice 0900
samplevoice 0902
samplevoice 0904
samplevoice 0906
samplevoice 0908 //and so on

thus this loop plays faster than it should be

Bantai, can you explain this interpolating thingy more in depth, or direct me to the page in the renoise wiki tutorial? I’ve searched, but couldn’t find it in the wiki + my attempts in renoise at what you’re discribing didn’t work, or maybe I’m just doing something wrong. cheers

man i just spent 15 minutes looking through the search function to find something i wrote about this… damn search engines to Hell!

anyway saw this an thought ooh that explains it good!

should mix this image of Mlon’s with Bantai’s directions an put it in da wiki!

thanks for the info’s, will check it now in renissle

after my announcement of the problem of the timestretching in Renoise 1.8
in this post https://forum.renoise.com/t/beatsync/18207

I want to post my timestretch test song.
Am I doing something wrong, or is it Renoise?

song

You must interpolate from 0900 to 09FC (not 09FF) when using a 64 lines long pattern.

-> for X lines in pattern, substract 256 / X in the last line.

thank you for that Tac tic!
makes sense indeed.
nice thing I don’t have to calculate every in between steps anymore with different track lengths.

am i correct in thinking this would not pertain to patterns above 128?

thanks.

above the 256 I think, because then all the offset points in the sample ‘fit’ into the pattern since Renoise calculates 256 offset points into a sample.

hi

you can’t change the pitch without changing the sample playing time unless using the sample offstet.
sometimes i want to keep the tone and slow down the beat.

sometimes we want to change the tone and keep the tempo
the “playing speed” of the sample IS atatched to the pitch of the sample (or vice-versa) we do not have any time-steching

the FF positions of a sample offstep do not have enought precision.
now imagine that you’ve got a 342 seconds of a sample, now imagine you want to match the last event of the sample with another position in the pattern.
excuse me but these is not cool

a 5.7 minute sample is very impractical for tracking, but i do see your reasoning very clearly. there as been many hours, an maybe even days or weeks worth of discussion on this aspect already.

the imprecision is so usefull! man serious!
for me it helps gel things together--------->Oh So Very Well!!
an there is even ways to get past this now with the recording feature.
(tho i haven’t personally used it yet, it looks very promising!)

am i looking at this wrong?

64 line pattern.
256 offset points

256 / 64 = 4

0 to FC truncates 3 points. FD, FE, FF

128 line pattern.
256 offset points

256 / 128 = 2

0 to FE truncates 1 point. FF