gwc don't respect the FHS policy path
This should be fixed in git; I'm planning to make a new release soon.
Bug in button
However some operations fail if you try to process an entire file, due to a bug where the selection is off by 1 sample. It should be an easy fix, but I was too busy to let it hold up the new release. For the record, these were fixed a few years back.
Patches are welcome. Or (preferably) pull requests against the stable branch at https://github.com/alisterh/gwc Otherwise I'm not likely to look at this in the foreseeable future, due to other commitments, sorry.
maxint.c:37:5: warning: incompatible implicit declaration of built-in function 'printf'
gwc don't respect the FHS policy path
For the record, as per the README: autoreconf -i configure is a generated file, so it is bad practice to include it in git. make dist packages it up in the source releases. When I was cherry-picking from master to create the stable branch, for some reason I missed the commit that made this more prominent in the README. I have added it now.
cannot test stable branch. https://github.com/AlisterH/gwc/tree/stable There is no configure files to building program in linux?
I have committed a fix to my stable branch (https://github.com/AlisterH/gwc/tree/stable) - an error message if the selection is too close to the beginning or end of the file. I will probably make another release in the next few weeks. It would be nice instead to fix it to work even if the selection reaches the beginning or end of the file, but I don't know that it is worth the effort. Users should just avoid trimming the beginning and end of their recordings until they have finished cleaning them....
I have committed a fix to my stable branch (https://github.com/AlisterH/gwc/tree/stable) - an error message if the selection is too close to the beginning or end of the file. It would be nice instead to fix it to work even if the selection reaches the beginning or end of the file, but I don't know that it is worth the effort. Users should just avoid trimming the beginning and end of their recordings until they have finished cleaning them. I determined that: - the other two dethunk implementations...
Testing cut, copy and paste. working in Linux.
And I see that the insert silence tool does the same. Actually that only seems to occur testing in WSL, and there are also issues with cut/copy/paste there, which use some of the same underlying functionality, so perhaps it is a Windows problem.
Thanks very much for the report - we definitely need to avoid crashes, and even though there is no reason to apply the estimate tool to an entire file, it should work anyway. The crash isn't actually caused by trying to operate on the entire file - it occurs even with just a short selection that extends to the end of the file. And I see that the insert silence tool does the same. I have fixed a number of bugs where gwc tried to read one sample past the end of the file, and I think this is the same...
Bug in button
"What I was trying to find is some soft of function which would only apply every 70ms (or any other time interval) for the 3ms of noise"-- that is essentially what the click detector is doing in your case. It is an algorithm that detects impulse events (with a sensitivity you can change), and applies audio correction only to the impulse event itself. Most of your 70ms audio frame remains untouched. And I see you then used denoise. That's the exact approach I would have used. I know it seems intuitive...
"and you are expecting too much" Truth. I admit. I managed to remove them using repetitive use of decklick and denoise. What I was trying to find is some soft of function which would only apply every 70ms (or any other time interval) for the 3ms of noise. And that I could then use for any video done with this camera. But if not possible, forget about it. Related to documentation, indead the one you mentionned is better and more explicit (?). The one included with the software is a bit like "to start...
Upload a longer sample if you like, with some of the audio you are trying to capture, not just noise. I suspect your clicks are just too bad, and you are expecting too much. But it is hard to comment without a proper sample. Regarding documentation: 1) there are various pages where people have described their workflow (e.g. the tutorial link on the left side at http://gwc.sourceforge.net/ - there are a number of links to other software there too). But I think people tend to cover the complete workflow...
Thank you for your help, both of you. I don't mind you close the subject if you think so. Indeed I tried the setting you suggest, Jeff, on the whole document and the click is reduced to a quite inaudible amount (ie, I can still detect it but it is less a nuisance). I'm just wondering if it would be possible to totally remove it, by example as only cleaning only every 70ms... Other idea: since the click is produced by some mechanism (probably) always behaving the same very next to the microphone,...
removing repetitive mechanical/clockwork noise
Hi Jeff, good to see you still keeping an eye on things. Ffroggy - I'll close this ticket as it seems your problem is solved. Is that right?
Just for a quick attempt, I used the GWC declick. The FFT detector didn't work on the sample, (either trying weak or strong click removal). But, uncheck the "use FFT detector" in Settings->declick, and then the weak declicker worked very well on the noisy camera sample. Hope that helps... Jeff Welty On Sunday, May 26, 2019, 9:37:03 AM PDT, ffroggy <ffroggy@users.sourceforge.net> wrote: [support-requests:#9] removing repetitive mechanical/clockwork noise Status: open Group: v1.0_(example) Labels:...
removing repetitive mechanical/clockwork noise
(wish) arts support
Closing as I believe arts was dead even before this ticket was opened!
Window layout
It appears that this was applied some time in the distant past :)
This issue has been raised elsewhere as well - patches would be welcome. Re my comments, for the record: Maybe my Alsa configuration changed or something. Probably I changed from outputting directly to stereo only hardware e.g. by "hw:0,0" to something which uses dmix e.g. "plughw:0,0" or "default". ... it encodes them as joint stereo, playing twice as fast This seems to be OK at the moment; hopefully I fixed it.
Fixed in (at least) the latest version.
Make gwc build correctly on VIA C3-based machines
Note that gwcbatch is now installed with the latest version of gwc, and it doesn't have a .pl extension (in-line with Debian guidelines). However some operations fail if you try to process an entire file, due to a bug where the selection is off by 1 sample. It should be an easy fix, but I was too busy to let it hold up the new release.
Note that this is now available in the recent fork at https://github.com/clixt/gwc/. Unfortunately Jan did that work before seeing that I'd been working on gwc, so it doesn't incorporate my fixes from the new release. I don't currently intend to merge that particular change, because I don't think it works well enough to be better than smart use of the "zoom selection" and "zoom extents" tools (noting that IIRC in the new version I have made some slight improvements in terms of what is selected after...
Note that this is available in the recent fork at https://github.com/clixt/gwc/. Unfortunately Jan did that work before seeing that I'd been working on gwc, so it doesn't incorporate my fixes from the new release. I don't currently intend to merge that particular change, because I don't think it works well enough to be better than smart use of the "zoom selection" and "zoom extents" tools (noting that IIRC in the new version I have made some slight improvements in terms of what is selected after...
Seems that gwc also has some problems of removing cracks/noise that is only present in one channel. I don't know about when this ticket was filed, but now it should be fine as it only operates on one channel if you make a selection in only one channel. Note that there is a recent fork at https://github.com/clixt/gwc/ with support for playback of only one channel. Unfortunately Jan did that work before seeing that I'd been working on gwc, so it doesn't incorporate my fixes from the new release. My...
gwc does not store its window size and position
Fixed in the latest version.
I have updated the icon support and the latest release should install a set of icons for menus etc.
Menu Icon please
save as separate tracks
GWC's track marking is the best I have seen for multisong recordings. I had an idea at least wavbreaker's had always been better, but whatever... I assume this ticket is out-of-date, as there is the option "File>Split audio on song markers".
Everything should be installed in standard locations in the latest version. Please report if I am wrong.
Incorrect install locations
denoising default settings are different from values in quic
First, to clarify what is described: - Ogg and mp3 read support are broken, but don't have any gui controls. - Ogg and mp3 export are always compiled in, and just need to be configured in gwc as they use external binaries. Yes, ideally we would find them automatically, instead of configuring it manually, and true, they may not be available, but if we check for their presence and hide the controls, how will the user know these features exist? I am open to any sensible patches, but otherwise these...
controls for unavailable operations should be disabled
Ogg and mp3 read support are broken, but don't have any gui controls. Ogg and mp3 export are always compiled in, and just need to be configured in gwc as they use external binaries. Yes, ideally we would find them automatically, instead of configuring it manually, and true, they may not be available, but if we check for their presence and hide the controls, how will the user know these features exist? I am open to any sensible patches, but otherwise these are a "won't fix" for me - it isn't worth...
Fixed in latest version - all instances of Khz are now kHz etc.
small typos: Khz --> kHz
It is very simple to install and use, except that it does not provide gui and sound support out of the box. Typo fixed - I had left out "does not"!
Re Windows support, I only recently noticed that Microsoft and Ubuntu have developed something called "Windows subsystem for Linux", in an effort to increase the relevance of their operating systems. This is sort of a cross between Colinux and Wine - they have built a compatibility layer for the Linux kernel, which allows you to install the Ubuntu userspace on Windows. In other words, if you have a 64bit version of Windows 10 you can run unmodified Linux binaries on it. It is very simple to install...
Re Windows support, I only recently noticed that Microsoft and Ubuntu have developed something called "Windows subsystem for Linux", in an effort to increase the relevance of their operating systems. This is sort of a cross between Colinux and Wine - they have built a compatibility layer for the Linux kernel, which allows you to install the Ubuntu userspace on Windows. In other words, if you have a 64bit version of Windows 10 you can run unmodified Linux binaries on it. It is very simple to install...
Re Mac, people certainly used to run gwc on OSX - but I don't have a Mac so I can't...
So - reimplement undo/redo and get pulseaudio working on Windows, and you are away...
FWIW if someone wanted to do it and had the right skills I don't think it would actually...
thats sad a lot of software could use gnome wave cleaner's tools.
Gnome Wave Cleaner fore os
Gnome Wave Cleaner fore os
No other OS's, sorry. This source code is linux only, it's old by even linux standards....
e Wave Cleaner needs something like auto heal
Already done, 15 years ago. View in spectral mode, select area to correct, apply...
e Wave Cleaner needs something like auto heal
Gnome Wave Cleaner fore os
On my system it displays mono files as described here, but it doesn't play them....
On my system it displays mono files as described here, but it doesn't play them.