Experimental audio effect based on autoconvolution

VST, AU, AAX, CLAP, etc. Plugin Virtual Effects Discussion
RELATED
PRODUCTS

Post

I have zero knowledge in programming and such so sorry if my comment doesn't make sense at this point, but hopefully a good samaritan is gonna compile it into a .exe & .dmg installers just like Paulstretch! :party:

Post

The examples sound great :tu:
I get This error:

Code: Select all

Traceback (most recent call last):
  File "C:\Users\Shpitz\Desktop\autoconvolution\2xautoconvolution.py", line 310,
 in <module>
    process_audiofile(input_filename,options.output,options,2 if options.keep_en
velope else 0)
  File "C:\Users\Shpitz\Desktop\autoconvolution\2xautoconvolution.py", line 93,
in process_audiofile
    subprocess.call(cmdline)
  File "C:\Python27\lib\subprocess.py", line 524, in call
    return Popen(*popenargs, **kwargs).wait()
  File "C:\Python27\lib\subprocess.py", line 711, in __init__
    errread, errwrite)
  File "C:\Python27\lib\subprocess.py", line 948, in _execute_child
    startupinfo)
WindowsError: [Error 2] The system cannot find the file specified

Post

Bump?

Post

Sounds very good. It reminds me of the blur tool in FL Edison.
Can you describe what it does from a sonic perspective?

Post

Absolutely fascinating effect Paul. Many thanks for sharing.

Is this sort of analogous to convolution with a constantly changing impulse, using the area around the 'playback position' as the impulse? Been doing that kind of thing with Fscape here. I'll check this out tomorrow.

Post

@shpitz Got the same error, but on OS X. Fixed it by installing https://libav.org for the avconv program.

Getting very interesting results with this, need to experiment some more.

Post

Spent a good 3 hours trying to get this up and running. I'm generally half decent with command line stuff (i.e. old DOS navigation skills and the odd exe with -switches) but I'm absolutely stumped. Is this within reach of the layman right now, or do you realistically need to be a programmer to get it going?

Running Windows 10 x64. Installed Python 2.7 and Scipy (via the Anaconda distribution), but running 2xautoconvolution.py (even without the switches/arguments) always produces this error:

Code: Select all

  File "2xautoconvolution.py", line 4
    <!DOCTYPE html>
    ^
SyntaxError: invalid syntax
Any insights? Should I just leave this alone for now. :)

Post

That "doctype" line is not in the raw text 2xautoconvolution.py file in github, I think you downloaded the whole HTML page instead of just the .py file, or copypasted something extra without noticing.

Post

pottering wrote:That "doctype" line is not in the raw text 2xautoconvolution.py file in github, I think you downloaded the whole HTML page instead of just the .py file, or copypasted something extra without noticing.
Very kind of you! Was wondering why HTML was involved. Can't believe I didn't spot that when I had a brief look at the .py file in Notepad. I just immediately assumed Notepad couldn't display .py files correctly when I saw something different to the program in there. Chrome seems to download... well, let's say it's not the .py file :lol:

Copied the program out and pasted it into 2xconvolution.py via notepad and that gets it going nicely. Had the wrong version of Anaconda as I was getting errors regarding quotes, going from Python 3.5 back to Python 2.7 (Anaconda distribution) as per toddhisattva's post seems to have solved that.

Getting the same error as shpitz now, and can't for the life of me figure out how to install libav on Windows for the avconv program. Any noob hints? Everything else seems to be working correctly as running the .py with the -h switch disaplys help as expected.

Post

cron wrote:
pottering wrote:That "doctype" line is not in the raw text 2xautoconvolution.py file in github, I think you downloaded the whole HTML page instead of just the .py file, or copypasted something extra without noticing.
Very kind of you! Was wondering why HTML was involved. Can't believe I didn't spot that when I had a brief look at the .py file in Notepad. I just immediately assumed Notepad couldn't display .py files correctly when I saw something different to the program in there. Chrome seems to download... well, let's say it's not the .py file :lol:

Copied the program out and pasted it into 2xconvolution.py via notepad and that gets it going nicely. Had the wrong version of Anaconda as I was getting errors regarding quotes, going from Python 3.5 back to Python 2.7 (Anaconda distribution) as per toddhisattva's post seems to have solved that.

Getting the same error as shpitz now, and can't for the life of me figure out how to install libav on Windows for the avconv program. Any noob hints? Everything else seems to be working correctly as running the .py with the -h switch disaplys help as expected.
Dropped avconv.exe into the folder containing 2xautoconvolution.py and that seems to have got me further forward. Problem I'm having now is that the program can't create the temporary wav file. The temporary folder that should contain tmp_input.wav is created, but it's always left empty.

Has anybody successfully got this running on Windows? Googled around and it looks as though the program is using \\ rather than \ which seems correct for Windows. I also redownloaded the .py directly rather than using my version pasted into Notepad as I also read that Notepad can switch characters in a way that breaks certain programs. No luck... Current error message as follows.

Code: Select all

D:\vst\notvst\autoconvolution>python 2xautoconvolution.py -b 5 -l 3 -k -o D:\sou
nd\redlight128autonconv.wav D:\sound\redlight12816.wav
Input file: D:\sound\redlight12816.wav
Output file: D:\sound\redlight128autonconv.wav
Using temporary directory: d:\temp\python\tmpryocmz2xautoconvolution
Traceback (most recent call last):
  File "2xautoconvolution.py", line 310, in <module>
    process_audiofile(input_filename,options.output,options,2 if options.keep_en
velope else 0)
  File "2xautoconvolution.py", line 97, in process_audiofile
    with contextlib.closing(wave.open(tmp_wav_filename,'rb')) as f:
  File "D:\installers\Anaconda2.7\lib\wave.py", line 511, in open
    return Wave_read(f)
  File "D:\installers\Anaconda2.7\lib\wave.py", line 160, in __init__
    f = __builtin__.open(f, 'rb')
IOError: [Errno 2] No such file or directory: 'd:\\temp\\python\\tmpryocmz2xauto
convolution\\tmp_input.wav'
Any hints appreciated :)

Post

what do i need to do to get this working inside Logic X ?

is there a AU plugin that can load python stuff ?

or do i have to install libraries or something, i'm a total newbie about this stuff
If your plugin is a Synth-edit/synth-maker creation, Say So.
If not Make a Mac version of your Plugins Please.

https://soundcloud.com/realmarco

...everyone is out to get me!!!!!!!

Post

realmarco wrote:what do i need to do to get this working inside Logic X ?
You'd be need to be an audio-savvy Python 2 developer (that knows Scipy and libav.org for the avconv program). From what I've read here, it seems that Python could create a stand-alone app. I wonder if it can also (easily) make a plugin?

A few up-for-a-challenge KVRers have tried to get it working, but only sandbags and asksol have reported success (in OSX). They didn't say if they managed to make a stand-alone executable or just ran the script in the Python IDE.

Anyone made a standalone yet?

I'd even be happy if someone wrote a step-by-step how to guide.

So I'm sat here waiting for a kindly Python Developer to snake by, and since I've already waited for Godot, this will be easy :lol:

Post

double post.
Last edited by Pytchblend on Wed Jan 06, 2016 8:22 pm, edited 1 time in total.

Post

Pytchblend wrote:
realmarco wrote:what do i need to do to get this working inside Logic X ?
You'd be need to be an audio-savvy Python 2 developer (that knows Scipy and libav.org for the avconv program). From what I've read here, it seems that Python could create a stand-alone app. I wonder if it can also (easily) make a plugin?

A few up-for-a-challenge KVRers have tried to get it working, but only sandbags and asksol have reported success (in OSX). They didn't say if they managed to make a stand-alone executable or just ran the script in the Python IDE.

Anyone made a standalone yet?

I'd even be happy if someone wrote a step-by-step how to guide.

So I'm sat here waiting for a kindly Python Developer to snake by, and since I've already waited for Godot, this will be easy :lol:
From memory I ran it as a command-line tool.

Turning this into an AudioUnit plugin would be a significant task I think. In the first place you'd either have to embed the Python interpreter into your plugin or port the code to C or C++. I'd probably go the former route, it's been done before (https://github.com/guibarrette/PyoPlug) but I suspect it's not without issues.

Then you'd have to figure out what it was going to do as a plugin ... either loading a file or making it into some kind of buffer effect (I have TimeFreezer in mind here as a model) and build the GUI to go with it.

Without a lot of effort I don't think you'd get a super amount of value over just running it via the command line :-)

m/

Post

Not counting to see this as a plugin. Seriously doubt…

BUT having this compiled as a standalone program like Paulstretch would be great though. Too painful to use as a music tool at the moment.

Post Reply

Return to “Effects”