26 Jun
2010
26 Jun
'10
11:47 a.m.
Hi everyone,
I'm looking for an app that assists in organizing and previewing
samples/wave files. I vaguely remember that there exists at least one
app for that, but Google fails me. Any idea?
Also, are there sample editing apps on Linux which support "elastic
time" editing, such as Pro Tools or Cubase do? I know Ardour has
timestretching and chopping up regions, but I'm looking for something
more comfortable...
Thanks,
Leonard
26 Jun
26 Jun
1:58 p.m.
On Sat, Jun 26, 2010 at 5:47 AM, Leonard Ritter <paniq(a)paniq.org> wrote:
Hi everyone,
I'm looking for an app that assists in organizing and previewing
samples/wave files. I vaguely remember that there exists at least one
app for that, but Google fails me. Any idea?
Also, are there sample editing apps on Linux which support "elastic
time" editing, such as Pro Tools or Cubase do? I know Ardour has
timestretching and chopping up regions, but I'm looking for something
more comfortable...
Given that Rubberband was really the first library available on Linux
without a license that could do high quality timestretching, its not
really suprising that there are not that many apps which support this
(*). "Elastic time" is really a phenomenally complex thing to
implement - I've heard that when it was added to ProTools they
basically had to redesign the entire program and it was probably their
biggest engineering undertaking since starting the project.
(*) yes, there was soundtouch for several years before, but it uses
WSOLA which generates artifacts on a lot more material than RB does.
and yes, there is one proprietary library available too, but without
source and with a license that makes it hard to use on GPL projects.
2:38 p.m.
On Sat, Jun 26, 2010 at 8:15 AM, Louigi Verona <louigi.verona(a)gmail.com> wrote:
What is elastic time then? How is it different from
time stretching?
fundamentally, its no different. but the term as i've seen it used by
digi, steinberg, cakewalk etc. refers to a more seamless integration
of it into the DAW workflow. i have to admit that when i saw a demo of
the first version of PT to do this, i was totally blown away. just
what digi had done with the GUI alone made me sweat profusely thinking
about how much work it would be. one example i saw was taking a female
vocal line, putting two markers around a particular section of vocals
and then stretching/compressing them in time by moving the markers.
rubberband can do this kind of thing, but in the demo, the waveforms
and everything else move dynamically as this is done. it changes the
basic model of what a DAW is from "something that plays back bits of
audio files in a specified order" to "something that continually feeds
audio through timestretching algorithms with dynamic parameters".
doing this efficiently ... i can imagine how to do it, but sheesh, its
a huge amount of work (perhaps, as usual, more in the GUI than
anywhere else)
--p
2:40 p.m.
On Sat, Jun 26, 2010 at 8:38 AM, Paul Davis <paul(a)linuxaudiosystems.com> wrote:
On Sat, Jun 26, 2010 at 8:15 AM, Louigi Verona
<louigi.verona(a)gmail.com> wrote:
> What is elastic time then? How is it different from time stretching?
oh, one other thing. at least in the PT implementation, elastic time
also relies on pretty good feature analysis of the audio. so you also
have to arrange for this analysis to be done in the background and for
the data to be continually available. ardour sort of does this
already, but only if the user asks for it because its computationally
quite expensive. ardour also only looks for transients by default
whereas PT appears to collect information on more features.
2:56 p.m.
Logic 9 has features as mentioned below too.
Its called the "flex" tool, and essentially does the same as the ProTool
one.
Its got another couple of fancy features when used together with the
selection
tool though, allowing the user to select the piece of audio to move, and it
will then
timestrecht before & after it. Makes for great use if a drummer misses a
hit, or guitarist
strums that little too late...
Nice features... If i ever understand the Ardour code base I might peek
about to see what
can be done... ;-)
Cheers, -Harry
On Sat, Jun 26, 2010 at 1:38 PM, Paul Davis <paul(a)linuxaudiosystems.com>wrote;wrote:
On Sat, Jun 26, 2010 at 8:15 AM, Louigi Verona
<louigi.verona(a)gmail.com>
wrote:
What is elastic time then? How is it different
from time stretching?
fundamentally, its no different. but the term as i've seen it used by
digi, steinberg, cakewalk etc. refers to a more seamless integration
of it into the DAW workflow. i have to admit that when i saw a demo of
the first version of PT to do this, i was totally blown away. just
what digi had done with the GUI alone made me sweat profusely thinking
about how much work it would be. one example i saw was taking a female
vocal line, putting two markers around a particular section of vocals
and then stretching/compressing them in time by moving the markers.
rubberband can do this kind of thing, but in the demo, the waveforms
and everything else move dynamically as this is done. it changes the
basic model of what a DAW is from "something that plays back bits of
audio files in a specified order" to "something that continually feeds
audio through timestretching algorithms with dynamic parameters".
doing this efficiently ... i can imagine how to do it, but sheesh, its
a huge amount of work (perhaps, as usual, more in the GUI than
anywhere else)
--p
_______________________________________________
Linux-audio-user mailing list
Linux-audio-user(a)lists.linuxaudio.org
http://lists.linuxaudio.org/listinfo/linux-audio-user
3:32 p.m.
On Sat, Jun 26, 2010 at 4:57 AM, Paul Davis <paul(a)linuxaudiosystems.com> wrote:>
Given that Rubberband was really the first library
available on Linux
without a license that could do high quality timestretching, its not
really suprising that there are not that many apps which support this
FYI, Qtractor ( http://qtractor.sourceforge.net/qtractor-index.html )
supports this feature and library:
Audio clip time-stretching (WSOLA-like or via
librubberband), pitch-shifting (via librubberband) and seamless sample-rate conversion
(via libsamplerate).
Yes, Linux needs a Logic or Sonar -- and so far Qtractor is the
closest thing to that. It is wonderful!
-- Niels.
http://nielsmayer.com
5:03 p.m.
On 06/26/2010 01:57 PM, Paul Davis wrote:
Given that Rubberband was really the first library
available on Linux
without a license that could do high quality timestretching, its not
really suprising that there are not that many apps which support this
(*).
Rubberband is interesting, but when I had to develop a professional tool to time
stretch voice recordings, the quality was far from acceptable.
In the end, I used Dirac. The quality's amazing. It's not open source, but
there's a free version with a rather permissive license (44.1/48Khz only).
Also, the author has recently added Linux object libs.
http://www.dspdimension.com/technology-licensing/dirac2/
What about adding (optional) support for it in Ardour, as an alternative to
Rubberband?
--
Olivier
5:14 p.m.
On Sat, Jun 26, 2010 at 11:03 AM, Olivier Guilyardi <ml(a)xung.org> wrote:
On 06/26/2010 01:57 PM, Paul Davis wrote:
I've been aware of Dirac for some time. Its not feasible to use his
license in combination with Ardour without creating a fairly complex
plugin APi. Why? His license is a license for the developer, and it
won't permit source redistribution. Since Ardour links against at
least 1 GPL library, it would violate the GPL to distribute Ardour
with Dirac support as a compile+link time option. I'm not against the
idea of a plugin API for time FX backends, I've just got better things
to do at this point in time, and for the foreseeable future.
Given that Rubberband was really the first
library available on Linux
without a license that could do high quality timestretching, its not
really suprising that there are not that many apps which support this
(*).
Rubberband is interesting, but when I had to develop a professional tool to time
stretch voice recordings, the quality was far from acceptable.
In the end, I used Dirac. The quality's amazing. It's not open source, but
there's a free version with a rather permissive license (44.1/48Khz only).
Also, the author has recently added Linux object libs.
http://www.dspdimension.com/technology-licensing/dirac2/
What about adding (optional) support for it in Ardour, as an alternative to
Rubberband?
5:54 p.m.
On 06/26/2010 05:14 PM, Paul Davis wrote:
On Sat, Jun 26, 2010 at 11:03 AM, Olivier Guilyardi
<ml(a)xung.org> wrote:
I see. Actually, I looked at Ardour time stretching code not so long ago. And
what I found was rather tight integration with Rubberband. I was't looking for a
plugin API, but for some simple kind of OO abstraction, that would allow to
quite easily add some new stretchers. So, unfortunately, it looks like the way
to a such plugin API is a bit long indeed...
--
Olivier
On 06/26/2010 01:57 PM, Paul Davis wrote:
I've been aware of Dirac for some time. Its not feasible to use his
license in combination with Ardour without creating a fairly complex
plugin APi. Why? His license is a license for the developer, and it
won't permit source redistribution. Since Ardour links against at
least 1 GPL library, it would violate the GPL to distribute Ardour
with Dirac support as a compile+link time option. I'm not against the
idea of a plugin API for time FX backends, I've just got better things
to do at this point in time, and for the foreseeable future. Given that Rubberband was really the first
library available on Linux
without a license that could do high quality timestretching, its not
really suprising that there are not that many apps which support this
(*).
Rubberband is interesting, but when I had to develop a professional tool to
time
stretch voice recordings, the quality was far from acceptable.
In the end, I used Dirac. The quality's amazing. It's not open source, but
there's a free version with a rather permissive license (44.1/48Khz only).
Also, the author has recently added Linux object libs.
http://www.dspdimension.com/technology-licensing/dirac2/
What about adding (optional) support for it in Ardour, as an alternative to
Rubberband?
7:53 p.m.
On Sat, Jun 26, 2010 at 11:54 AM, Olivier Guilyardi <ml(a)xung.org> wrote:
I see. Actually, I looked at Ardour time stretching
code not so long ago. And
what I found was rather tight integration with Rubberband. I was't looking for a
plugin API, but for some simple kind of OO abstraction, that would allow to
quite easily add some new stretchers. So, unfortunately, it looks like the way
to a such plugin API is a bit long indeed...
it can't be that bad : see libs/ardour/stretch.h which has hooks for
SoundTouch as well.
7:54 p.m.
On Sat, Jun 26, 2010 at 1:53 PM, Paul Davis <paul(a)linuxaudiosystems.com> wrote:
On Sat, Jun 26, 2010 at 11:54 AM, Olivier Guilyardi
<ml(a)xung.org> wrote:
its mostly a function of what parameters are present in a
TimeFXRequest, and how to present them to the user.
I see. Actually, I looked at Ardour time
stretching code not so long ago. And
what I found was rather tight integration with Rubberband. I was't looking for a
plugin API, but for some simple kind of OO abstraction, that would allow to
quite easily add some new stretchers. So, unfortunately, it looks like the way
to a such plugin API is a bit long indeed...
it can't be that bad : see libs/ardour/stretch.h which has hooks for
SoundTouch as well.
27 Jun
27 Jun
4:22 p.m.
On 06/26/2010 07:53 PM, Paul Davis wrote:
On Sat, Jun 26, 2010 at 1:53 PM, Paul Davis
<paul(a)linuxaudiosystems.com> wrote:
Indeed that looks quite simple. When I first searched for it, I think that I
lost myself browsing the Stretcher* classes in libs/rubberband.
--
Olivier
On Sat, Jun 26, 2010 at 11:54 AM, Olivier
Guilyardi <ml(a)xung.org> wrote:
its mostly a function of what parameters are present in a
TimeFXRequest, and how to present them to the user. I see. Actually, I looked at Ardour time
stretching code not so long ago. And
what I found was rather tight integration with Rubberband. I was't looking for a
plugin API, but for some simple kind of OO abstraction, that would allow to
quite easily add some new stretchers. So, unfortunately, it looks like the way
to a such plugin API is a bit long indeed...
it can't be that bad : see
libs/ardour/stretch.h which has hooks for
SoundTouch as well.
26 Jun
26 Jun
8:21 p.m.
On 06/26/2010 07:53 PM, Paul Davis wrote:
On Sat, Jun 26, 2010 at 11:54 AM, Olivier Guilyardi
<ml(a)xung.org> wrote:
Thank you for this pointer. I must have missed that. If I ever find some time
for it, integrating Dirac would quite be the first thing I would like to
contribute to Ardour. It's not very FLOSSish, and I don't really need it right
now, but there's lot of wow and professional potential I think :)
--
Olivier
I see. Actually, I looked at Ardour time
stretching code not so long ago. And
what I found was rather tight integration with Rubberband. I was't looking for a
plugin API, but for some simple kind of OO abstraction, that would allow to
quite easily add some new stretchers. So, unfortunately, it looks like the way
to a such plugin API is a bit long indeed...
it can't be that bad : see libs/ardour/stretch.h which has hooks for
SoundTouch as well.
8:26 p.m.
On 06/26/2010 08:20 PM, Olivier Guilyardi wrote:
On 06/26/2010 07:53 PM, Paul Davis wrote:
Also, there's something funny, it looks like Dirac's author first thought about
making it GPL, then abandoned the idea. Read that: http://www.clearscale.org/ (!)
So, maybe that he could be like the idea of integration into Ardour, and perhaps
loosen the license a bit, dunno..
--
Olivier
On Sat, Jun 26, 2010 at 11:54 AM, Olivier
Guilyardi <ml(a)xung.org> wrote:
Thank you for this pointer. I must have missed that. If I ever find some time
for it, integrating Dirac would quite be the first thing I would like to
contribute to Ardour. It's not very FLOSSish, and I don't really need it right
now, but there's lot of wow and professional potential I think :) I see. Actually, I looked at Ardour time
stretching code not so long ago. And
what I found was rather tight integration with Rubberband. I was't looking for a
plugin API, but for some simple kind of OO abstraction, that would allow to
quite easily add some new stretchers. So, unfortunately, it looks like the way
to a such plugin API is a bit long indeed...
it can't be that bad : see
libs/ardour/stretch.h which has hooks for
SoundTouch as well. 
8:32 p.m.
On Sat, Jun 26, 2010 at 4:03 PM, Olivier Guilyardi <ml(a)xung.org> wrote:
Rubberband is interesting, but when I had to develop a
professional tool to time
stretch voice recordings, the quality was far from acceptable.
No, Rubber Band is better for some sorts of audio than others and
speech certainly isn't its strong point.
Tim Goetze also had an example recently (here or on LAD) which it
didn't handle well -- I've been looking into that one recently and
hope to have some improvements there for the next release at least.
Chris
8:48 p.m.
On 06/26/2010 08:32 PM, Chris Cannam wrote:
On Sat, Jun 26, 2010 at 4:03 PM, Olivier Guilyardi
<ml(a)xung.org> wrote:
Yep, that's right. I did a quite lot of tests, and realized every stretching lib
is better for some sort of sound than others. Also, some are optimized for
speed, not quality.
I've actually been happy with Rubberband for my own music stuff, but when I had
to develop a critical tool for mass stretching of voice recordings, Dirac just
did an incredible job.
As you may know, it seems to be the result of decades of developement by the
founder of Prosoniq, which also produced MPEX.
Rubberband is interesting, but when I had to
develop a professional tool to time
stretch voice recordings, the quality was far from acceptable.
No, Rubber Band is better for some sorts of audio than others and
speech certainly isn't its strong point. Tim Goetze also had an example recently (here or on
LAD) which it
didn't handle well -- I've been looking into that one recently and
hope to have some improvements there for the next release at least.
Also, when I mention Dirac support in Ardour, I don't mean that it should
replace Rubberband. In such apps as Nuendo and the like, the user has a choice
about which algorithm to use. Plus, since Dirac isn't FLOSS, it wouldn't be
installed by default in many cases I suppose. AFAIK, Rubberband and SoundTouch
are still the only free stretchers in the market place :)
But, from my tests, I believe that you could improve things a lot in regard to
voice processing.
I can recommend that you take almost any piece of voice, stretch in the 80%-120%
range, and compare the Dirac result with Rubberband's.
--
Olivier
27 Jun
27 Jun
4:58 a.m.
Paul Davis wrote:
Given that Rubberband was really the first library
available on Linux
without a license
<pedantic>
Rubberband does have a license.
</pedantic>
Erik
--
----------------------------------------------------------------------
Erik de Castro Lopo
http://www.mega-nerd.com/
5:06 a.m.
On Sat, Jun 26, 2010 at 10:58 PM, Erik de Castro Lopo
<mle+la(a)mega-nerd.com> wrote:
Paul Davis wrote:
oops, that was a typo not a thinko. i wasn't for a moment trying to
say that it didn't have a license. i was trying to type "without a
GPL-conflicting license" and failed.
Given that Rubberband was really the first
library available on Linux
without a license
<pedantic>
Rubberband does have a license.
</pedantic>
5062
days inactive
5063
days old
linux-audio-user@lists.linuxaudio.org
18 comments
8 participants
participants (8)
-
Chris Cannam -
Erik de Castro Lopo -
Harry Van Haaren -
Leonard Ritter -
Louigi Verona -
Niels Mayer -
Olivier Guilyardi -
Paul Davis