search

renoise / tools

Scripting tools written by the Renoise team
MIT License
285 stars 43 forks source link

Rubberband: Misses a channel sometimes #11

Closed GoogleCodeExporter closed 9 years ago

GoogleCodeExporter commented 9 years ago 
Timestretching a stereo sample sometimes produces a mono sample.

It's not that random. I haven't found the exact pattern yet, but this is 
reproducable:

1) new song: BPM 128 / LPB 4 / 64 lines
2) Load "Break Amen.xrni"
3) Timestretch Amen sample to a half pattern, ie. 8 beats: left channel only
4) Timestretch Amen sample to a full pattern, ie. 16 beats: stereo

Original issue reported on code.google.com by marvint...@gmail.com on 15 Jun 2010 at 2:06

GoogleCodeExporter commented 9 years ago 
Issue reported by several people. 
http://www.renoise.com/board/index.php?showtopic=25255

Original comment by marvint...@gmail.com on 15 Jun 2010 at 2:09

GoogleCodeExporter commented 9 years ago 
Might be rubberband bug. I will dig into this today.

Original comment by ClyS...@gmail.com on 12 Jul 2010 at 12:09

GoogleCodeExporter commented 9 years ago 
That reproduction method is not reliable. Does not work every time, works some 
times.

May be some sort of race condition. Not sure exactly. Need to study that a bit 
more.

Original comment by ClyS...@gmail.com on 13 Jul 2010 at 7:47

GoogleCodeExporter commented 9 years ago 
[deleted comment]
GoogleCodeExporter commented 9 years ago 
sorry, had to delete that comment. different behaviour on different song... so 
NOT reproducable.

Original comment by plasmani...@googlemail.com on 19 Jul 2010 at 10:30

GoogleCodeExporter commented 9 years ago 
It appears to be Rubberband bug, the source sample and parameters are identical 
on both calls the one creating broken and the one creating functional file.

Original comment by ClyS...@gmail.com on 19 Jul 2010 at 11:29

GoogleCodeExporter commented 9 years ago

Original comment by ClyS...@gmail.com on 19 Jul 2010 at 11:29

GoogleCodeExporter commented 9 years ago 
Setting the Crispness to anything apart from 4 seems to fix the stereo channel 
problem.

Original comment by mark.bry...@gmail.com on 19 Jul 2010 at 12:09

GoogleCodeExporter commented 9 years ago 
Might be temporary solution to change the crispness default for now.

Original comment by ClyS...@gmail.com on 19 Jul 2010 at 12:25

GoogleCodeExporter commented 9 years ago 
Upstream bug: http://code.breakfastquay.com/issues/3

Original comment by ClyS...@gmail.com on 20 Jul 2010 at 9:17

GoogleCodeExporter commented 9 years ago 
Running on the commandline the following always gives me a reproducable 
behaviour:

rubberband.exe -c 4 --precise --time 4 amen.wav amen_out.wav

Suspected "--no-threads" might solve it, but it's really a bug in rubberband 
and not the Renoise tool.

A small suggestion would be to move the "if stime > 100 then" check into the 
"if res == 'Stretch' then" check, so i don't get bothered by a warning message 
when i press the "Cancel" button. Load one of the included chipsound 
instruments, open the timestretch dialog while it's selected and cancel it.

Original comment by beatslau...@googlemail.com on 28 Jul 2010 at 8:48

GoogleCodeExporter commented 9 years ago 
A bump and request to either fix this one or ignore it - close the report. This 
bug is open since a few months now.

ClySuva: could you close it please?

Original comment by mueller....@googlemail.com on 21 Apr 2011 at 7:52

GoogleCodeExporter commented 9 years ago 
Ah yes, upstream claims it's fixed, I will see if it's true and update the 
package accordingly today.

Original comment by ClyS...@gmail.com on 21 Apr 2011 at 8:56

GoogleCodeExporter commented 9 years ago

Original comment by ClyS...@gmail.com on 3 May 2011 at 10:18