PaulStretch3 preview(12) release

VST, AU, AAX, CLAP, etc. Plugin Virtual Effects Discussion
Post Reply New Topic
RELATED
PRODUCTS

Post

Preview 11 up, links in the original post. This just has the bug fix.

2017-11-17 (3.0.0.preview11, Xenakios fork)
-Fix bug with multichannel input files in shortening mode

Post

Woah !
Well done, thanks :-)

Post

Kudos Xenakios! Your updates are appreciated.

BTW, here is a demo I made: https://www.youtube.com/watch?v=H83B1azvhrQ

Post

brijawi wrote:Kudos Xenakios! Your updates are appreciated.
Thanks for the video. There seems to be some kind of a GUI glitch, there's an additional group box with the "Time stretch/playback" text at the top of the window that I am not seeing here...Is that maybe because of the screen capture software or does the GUI look like that for you always? Which OS are you using?

Post

I attached an image of the gui. I'm on Windows 10.
You do not have the required permissions to view the files attached to this post.

Post

brijawi wrote:I attached an image of the gui. I'm on Windows 10.
Thanks, that looks OK, I don't have to start hunting for a Windows 10 specific bug, then... :phew: (Which would have been a complete pain, I don't have Windows 10 here...)

Post

Good! This was working very well. Once again, thanks for the updates!

Post

Hello, Linux user here on Linux Mint 17.3 (based on Ubuntu Thrusty ==> 14.04). When I launch the executable from the terminal I get the following error:

./PaulStretch3: error while loading shared libraries: libwebkit2gtk-4.0.so.37: cannot open shared object file: No such file or directory

Then I went to the source to compile it myself in JUCE, but also JUCE (Projucer) complains about the same missing library. I digged a bit and finally I discovered that I cannot get the libwebkit2gtk-4.0 to be installed in my OS (Mint 17 supported up to libwebkit2gtk-3.0).

Any suggestions (aside from "update your system")?

Thanks in advance,

Alberto

Post

AZZIN wrote: Any suggestions (aside from "update your system")?
I realized recently the issue with the libwebkit2gtk exists, but I haven't yet tested the possible solution that exists for it. There's a #define that can be used for JUCE to remove the dependency on the web browser stuff. (The PaulStretch3 application of course doesn't even need it, but the WebBrowserComponent is enabled and built with JUCE by default.)

edit : In any case though, I have no particular plans to keep older operating systems working...Some other change in the build settings/dependencies might later again become a problem, even if this libwebkit2gtk issue can now be solved...

edit2 : Looks like I was able to remove the libwebkit2gtk dependency. The .jucer file is updated for that in the git repo, maybe for Linux I should add the makefile too, so the Projucer doesn't need to be separately built just to get the makefile...But when trying to run the PaulStretch3 binary on Puppy Linux I immediately got a new dependency issue...It can probably pretty safely be said the code just isn't that compatible with old/stripped down Linux distros. (And it was never even intended to be compatible like that.)

The Linux build without the libwebkit2gtk dependency :

https://drive.google.com/open?id=1HHvDp ... tSCM2A6NSB

You may need to set permissions to allow executing the binary.

Post

Xenakios, thanks for the very quick feedback. I downloaded the latest executable.
The GUI starts now and it looks GORGEOUS, congratulations you've done a big work.

Unfortunately when I load a sound file (both mono or stereo) and try to play it I got the message:

./PaulStretch3: relocation error: ./PaulStretch3: symbol _ZNSt13random_device7_M_initERKNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEE, version GLIBCXX_3.4.21 not defined in file libstdc++.so.6 with link time reference

It looks like something related to the version of libstdc++ on which you build it (and possibly I don't have). Is that the dependency issue you got?

Thanks,
Alberto

Post

AZZIN wrote: It looks like something related to the version of libstdc++ on which you build it (and possibly I don't have). Is that the dependency issue you got?
Yeah, the dependency issues I got after resolving the webkit thing were about the C++ library. (It needs a pretty new one with C++14 and even newer things included. That does appear to be present in Ubuntu 16.10, though...So it's not the absolute latest what is required.)

And I can pretty much guess that if that dependency problem somehow gets solved, yet another one pops up...(I can try to test if there's some way to build everything statically into the binary, but I don't even know if that's supported with JUCE and FFTW...

edit : Sorry, this is something that is starting to take a ton of time and effort to sort out, I'd rather work on the upcoming plugin version's code than these build issues on Linux. :( You can use the "ldd" command to see what shared library dependencies the binary has, there are quite a few...

Post

Hey Xenakios,
Just discovered this! This is great!!!! I've been using paulstretch for years now and always prayed that some one will lift the glove on that one. And its you! the father of cdp frontend and hoursglass! such a pleasure!
(now for the feature requests....)
Are you planing on adding a way to automate these parameters?
BTW
testing on windows10 and for now running smooth!

Peace

Ofer

Post

ofersmi wrote: Are you planing on adding a way to automate these parameters?
Probably not for the standalone version...The plugin version has automation working as far as it can be expected to work. That is, there is lots of latency for the parameter changes even with smaller FFT sizes and it only gets worse with larger FFT sizes. There isn't really a good way around that as the processing happens in blocks the size of the FFTs. (I could maybe look into allowing to disable the prebuffering which might shave off some of the latency/randomness involved with the parameter changes...)

Regarding the plugin version, that has slowly progressed but I haven't tested it with many different hosts yet.

Post

Xenakios wrote: Probably not for the standalone version...The plugin version has automation working as far as it can be expected to work. That is, there is lots of latency for the parameter changes even with smaller FFT sizes and it only gets worse with larger FFT sizes. There isn't really a good way around that as the processing happens in blocks the size of the FFTs. (I could maybe look into allowing to disable the prebuffering which might shave off some of the latency/randomness involved with the parameter changes...)

Regarding the plugin version, that has slowly progressed but I haven't tested it with many different hosts yet.
I was just about to eagerly offer alpha testing time, then I realised you'll put it up when you're happy!

Post

CinningBao wrote:
Xenakios wrote: I was just about to eagerly offer alpha testing time, then I realised you'll put it up when you're happy!
I today looked at the code again after some time off from it and was able to make improvements. Some kind of release might not be too far off...What operating system and DAW app are you using? (Sorry, I don't remember, you might have mentioned that at some point...)

Post Reply

Return to “Effects”