Feature to cut unwanted audio before and after the loop, yet exclude from cutting a few milliseconds of audio before and after the loop

[ultimatebeta07-rgb]

Hi there,
First and foremost, I want to thank the ones who worked on this incredible software, making it available for free.

It was a coincidence that I got to know about his one. An extremely useful software but working on this becomes tiresome and the application has not had any updates for quite a while.

I believe the author / authors of the LoopAuditioneer will take necessary steps to include certain features.

I'm likely aware that the LoopAuditioneer was designed to create samples for GrandOrgue. However, since users like me and several others, hopefully, use it for other purposes, it will be useful to include a batch processing feature which allows to cut unwanted audio before and after the looped audio, but still allows me to dial in certain milliseconds that can be excluded immediately or near to, before and after the loop points. Because I see a change in the audio's waveform when the audio is trimmed almost to the loop points. Therefore, leaving out a few milliseconds could be useful.

I also see that the application does not support 24-bit wav files. As a result, I will have to convert the file to a 16-bit on another software and then import the files here. It would be great to include a feature to convert the file from 24 to 16-bit wav on the go on LoopAuditioneer itself.

I even notice that while working on long sessions like an hour or so, the application starts to malfunction. Like, the features like crossfading using linear, s-shape and equal gain don't behave as they should. It's kind of weird. Then i have to deliberately close the application and restart it again. I think others could have encountered the same.

I hope the author/ authors will meet the necessary changes at the earliest. Once again thanks for this software and all your efforts friend/s.

Regards,
Ultima

[larspalo]

the application has not had any updates for quite a while

Last (latest) commit was done on Feb 20, 2026. Since then I've been on vacation until recently. An actual release (0.13.0) was done 18 jan 2026. Sorry, but in my world this is not qualifying as "not had any updates for quite a while"!

However, since users like me and several others, hopefully, use it for other purposes

Other use cases than what's specified in the opening sentence of the README file is really mostly irrelevant when developing this tool.

a batch processing feature which allows to cut unwanted audio before and after the looped audio

Exporting only the looped audio data as new files in batch mode is implemented in 473d493 but only the loop(s) and no milliseconds before or after will be exported.

Because I see a change in the audio's waveform when the audio is trimmed almost to the loop points.

As this is actually the main topic/reason for this issue, can you really prove it? I suggest that you export a loop with the batch process and then compare the exported audio data with the original in a hex editor! This requires some basic skill and knowledge of the wave file format, so perhaps it's wise to make the loop to export cover almost the whole original file without perhaps just a few samples in the beginning and in the end. Some parts of the header will of course have changed as the duration of the audio data have changed so you need to know how to look for the actual audio data part of the file.

Just performing any trimming action won't actually change any audio data information - other than what's removed. However, doing fades (in/out), crossfades or any other operation like that manually will indeed cause a slight change in the audio data! That's to be expected. Absolutely no change will happen if only the metadata is edited - which is one of the key points of this software!

I also see that the application does not support 24-bit wav files

This is a lie. Any wave file that libsndfile and GrandOrgue can handle should also work in LoopAuditioneer. Other files won't be supported. I've personally processed many, many thousands of 24-bit wave files and also manually verified that the writing is indeed bit true when only the metadata is changed.

It would be great to include a feature to convert the file from 24 to 16-bit wav

I'm contemplating adding this as a batch process. The same with re-sampling, but both types of operations will change the audio data somewhat and that can/will make the loop points not being perfect anymore.

I even notice that while working on long sessions like an hour or so, the application starts to malfunction. Like, the features like crossfading using linear, s-shape and equal gain don't behave as they should. It's kind of weird.

Yes, this is weird. I've not experienced this myself and I've worked really long sessions with this software running on Linux. Performance on any other OS might be different but that's something I cannot really test as I only have/use various Linux machines.

I've had one other report that is related to memory management problems when using the batch export of .tsv files and that "should" have been fixed with b4b4df2 but I've not gotten any feedback from the user reporting it yet.

If you can reliably reproduce the weird behavior please describe exactly what steps will cause the problem so that I might try to investigate it. Just saying "while working on long sessions" is not really that helpful to track down where something might mis-behave.

[ultimatebeta07-rgb]

Hey there,
Thanks for the response mate.

Last (latest) commit was done on Feb 20, 2026. Since then I've been on vacation until recently. An actual release (0.13.0) was done 18 jan 2026. Sorry, but in my world this is not qualifying as "not had any updates for quite a while"! #

But on the site https://sourceforge.net/projects/loopauditioneer/files/ it says, the file was last modified on 2023-10-31 and I have the version 0.9.82. This is the version that I'm using for over 3 months or so. Not sure what you mean by "last commit was done on Feb 20,2026" and I have been having certain issues with this latest version.

Exporting only the looped audio data as new files in batch mode is implemented in 473d493 but only the loop(s) and no milliseconds before or after will be exported.

I'm not sure how it works. In the batch mode I only see this option "Remove sound between last loop and cue" and when I use this feature, only the unwanted audio after the loop is removed. The unwanted part before the loop is just there. I have to remove it manually by cutting it milliseconds by milliseconds and its frustrating. And the maximum I can go is 2000 milliseconds.

As this is actually the main topic/reason for this issue, can you really prove it? I suggest that you export a loop with the batch process and then compare the exported audio data with the original in a hex editor! This requires some basic skill and knowledge of the wave file format,....

Here what I mean is, when the unused part is cut, that is, the cut is actually done exactly at the loop point in the batch processing, there is no difference in sound, the waveform change, i mean, is that the waves channel 1 and channel 2 waves merging in line with each other is displayed differently after the batch processing is performed. Here actually I'm wrong. Because this does not actually kill loops. The loops are preserved. I actually confused the change in line with that of loop being killed. See the waveforms for reference... I'm wrong here... I was just worried that the difference in waveform meant that the batch process killed the loop. but it didn't. I just wanted to leave a few milliseconds before and after the loop points just in case to avoid the sampler I'm using from misinterpreting the loop points are being cut.

But still I can only cut the unwanted audio after the loop and not before the loop point.

If you can reliably reproduce the weird behavior please describe exactly what steps will cause the problem so that I might try to investigate it. Just saying "while working on long sessions" is not really that helpful to track down where something might mis-behave.

Let's say I use the "Audio loop Settings" on a particular sample, after going through the audio pool of samples, I select one loop among the other that really caters to my expectation. Then I save that only loop and then use the crossfade feature to crossfade the ends. I usually use the 'linear' or 'equal power/ gain' to crossfade the ends. Sometimes when using the crossfade, the loop sounds like, especially at the point where the crossfade is performed the audio raises and lowers in volume abruptly, like up and down. I totally understand that the 'duration of the crossfade (in milliseconds)' has to be adjusted for the best crossfade results. But sometimes, when I change the milliseconds duration, the crossfade does not perform as intended. Sometimes, the application even freezes. Then I end up closing the application and perform the same operation after a while, and this time the crossfade works as intended. I'm not sure why this happens.

Again thanks for the awesome work you are doing. But certain improvements to this application would reduce hours of work. It usually takes me at least 3 hours to work on a single patch from recording and editing the samples for a single sound and most of the time is spent is spend on the LoopAuditioneer. Sorry if I'm wrong, since I'm not a tech guy and sorry for the terrible language. English isn't my language either.

[larspalo]

Sorry if I'm wrong, since I'm not a tech guy and sorry for the terrible language. English isn't my language.

No problem at all, english is not my native language either!

Anyway, the binaries on SourceForge are indeed old. As mentioned on the button there the project has moved to GitHub - here! So the finished releases can be found at https://github.com/GrandOrgue/LoopAuditioneer/releases and the most recent intermediate builds that will contain all the latest features not yet released can be found on https://github.com/GrandOrgue/LoopAuditioneer/actions. You'll find more options with a recent build!

Regarding the crossfades I can only say that on audio that is well correlated the linear crossfade type is the only one that produce good results. All the loops detected in samples with LoopAuditioneer will be more or less well correlated... The other types were added just for completeness and, as I understand it, they are mostly intended for less correlated audio anyway. That said - there could absolutely be possible that some things in LoopAuditioneer is not working as they should!

[ultimatebeta07-rgb]

Thanks for the quick response!

No problem at all, english is not my native language either!

Haha. Nevertheless your language is too good!

Anyway, the binaries on SourceForge are indeed old. As mentioned on the button there the project has moved to GitHub - here! So the finished releases can be found at https://github.com/GrandOrgue/LoopAuditioneer/releases and the most recent intermediate builds that will contain all the latest features not yet released can be found on https://github.com/GrandOrgue/LoopAuditioneer/actions. You'll find more options with a recent build!

Ahh, my bad. Extremely sorry. It was my fault. I must have looked into the site more carefully.

Regarding the crossfades I can only say that on audio that is well correlated the linear crossfade type is the only one that produce good results. All the loops detected in samples with LoopAuditioneer will be more or less well correlated... The other types were added just for completeness and, as I understand it, they are mostly intended for less correlated audio anyway. That said - there could absolutely be possible that some things in LoopAuditioneer is not working as they should!

Ahh, I see. Thanks for the insights. I shall look into the latest version. Hope most of my needs will be met with this latest release. Thanks again for the tremendous work that you put in, all that at no cost for us users.

Regards,
Ultima

[ultimatebeta07-rgb]

Hey mate,
I still don't see an option to remove unwanted audio before and after the loop with just one click on the latest version. Am I missing something?

If not, could you please add it as a feature in the batch processing, if possible, in the upcoming update? Thanks.

[larspalo]

Am I missing something?

No, it's not yet in a release but if you go to download the latest intermediate build from https://github.com/GrandOrgue/LoopAuditioneer/actions you should see that it's available since build number 60.

I have still a few things to do before I'll publish a new release, so until then please use a recent build from https://github.com/GrandOrgue/LoopAuditioneer/actions instead.

[ultimatebeta07-rgb]

Thanks. Sorry, but I'm not sure what I'm doing wrong. I downloaded a file from the latest release for windows, but I don't see the option.

[larspalo]

Try https://github.com/GrandOrgue/LoopAuditioneer/actions/runs/22215594529/artifacts/5585931477 instead until a new release is done.

[ultimatebeta07-rgb]

Thanks for the patience and sorry for dodging you with questions.

I have found the option with the release you mentioned. The function does write the loop to a new file but kills the loop points. It's better to keep the loop points and not kill them, in the attempt to writing it into a new file.

Hope you will introduce this feature in the upcoming build. Thanks!

Have you worked on any other applications similar to this? If so, please let me know. I would like to try them as well.

[larspalo]

It's better to keep the loop points and not kill them, in the attempt to writing it into a new file.

It's just a few lines of code more, so I might add them in a future commit to keep the loop that was just exported in the new file. Stay tuned.

Have you worked on any other applications similar to this?

Back in the day when I was still using Windows and just starting to work with samples for GrandOrgue I actually bought a program called SampleLooper. In the very beginning I also tried using programs like Wavosaur, ZeroX seamless looper and Sound Forge. But do you really think that I would have gone through all the trouble to develop LoopAuditioneer if I had found a program that would do what I wanted it to do?

There are of course many other excellent open source audio programs out there, and I myself use Audacity extensively in the pre-production phase. But when the sample is cut and needs further processing for usage in a sample player like GrandOrgue, there's just nothing else I've found that works on Linux like I want it to do. Depending on your particular needs though, you might find something that works better for you - you just need to keep looking for it (or develop it yourself)!

[ultimatebeta07-rgb]

It's just a few lines of code more, so I might add them in a future commit to keep the loop that was just exported in the new file. Stay tuned.

Sure. Thanks for the timely help mate!

Back in the day when I was still using Windows and just starting to work with samples for GrandOrgue I actually bought a program called SampleLooper. In the very beginning I also tried using programs like Wavosaur, ZeroX seamless looper and Sound Forge. But do you really think that I would have gone through all the trouble to develop LoopAuditioneer if I had found a program that would do what I wanted it to do?

Sounds interesting. You are right. No wonder I couldn't find one good program that could just cut audio and perform loop crossfade seamlessly. It's a pity that, I even tried programs from some big and renowned companies, but they all could do some all the other stuff, except looping an audio. What a joke!

And to my surprise, I was fortunate to come across LoopAuditioneer on YouTube and that's how I ended up here. I know it's not some child's play to develop a program. Though being a non-tech guy, I know the hours of work and efforts that go into making this and I'm thankful to you and your coordinates for making this an open-source software. People like you truly deserve due recognition.

There are of course many other excellent open source audio programs out there, and I myself use Audacity extensively in the pre-production phase. But when the sample is cut and needs further processing for usage in a sample player like GrandOrgue, there's just nothing else I've found that works on Linux like I want it to do. Depending on your particular needs though, you might find something that works better for you - you just need to keep looking for it (or develop it yourself)!

You're right. Audacity is indeed a good software and sadly it still doesn't have looping option, as you have mentioned. I wish I knew how to develop a program like you do. Sadly, I know nothing about programming, apart from some basic stuff on the computer.

So, if you are working with LoopAuditioneer you must be a Pipe Organ player. Do you record and produce your own organ sample sets for GrandOrgue? If so, that would be a lot of work and that's really cool. And knowing both music and programming comes with lots of advantages.

[larspalo]

I've committed some code to keep the loop metadata when batch exporting the audio data of loops to new files that you can test. All the action builds for this commit can be found at https://github.com/GrandOrgue/LoopAuditioneer/actions/runs/23164648792 and the particular build for windows is https://github.com/GrandOrgue/LoopAuditioneer/actions/runs/23164648792/artifacts/5952531082 until a new release is done.

So, if you are working with LoopAuditioneer you must be a Pipe Organ player.

Yes, I'm an organist (church musician) by profession. Programming is just a hobby, as is recording even though I do use some of that skill in my professional work every now and then too.

Do you record and produce your own organ sample sets for GrandOrgue?

Yes, at the top of https://github.com/GrandOrgue/grandorgue/wiki/Sample-Sets there's a list of the sample sets I've produced.

[ultimatebeta07-rgb]

I've committed some code to keep the loop metadata when batch exporting the audio data of loops to new files that you can test. All the action builds for this commit can be found at

Hola! Thank you so much mate. I tried it out and it turns out to work well.

Yes, I'm an organist (church musician) by profession. Programming is just a hobby, as is recording even though I do use some of that skill in my professional work every now and then too.

Glad to hear that. The fondness for this instrument, Pipe Organ has been slowly deteriorating over the years. I'm happy that people like you are taking the pains to learn, teach, record these timeless instruments, that would involve a lot of work. Wish I learnt music in a professional manner. I just play by ear and organs are one of my favourite instruments, though I haven't played on an actual organ myself.

Yes, at the top of https://github.com/GrandOrgue/grandorgue/wiki/Sample-Sets there's a list of the sample sets I've produced.

That's a lot of organ sample sets! That too for free!! I just gave a few of them a listen, and it sounds really good for a free sample set. People like me can make use of these sample sets who can't afford to buy the professional Hauptwerk and their sample sets.

Rotterdam, Caen and Menesterol are few of my other favorites among the professional sample sets. Good that you played a part in making these grand, wonderful organs take a newer life form. Thanks for your service.

Is there any means that I could make a small donation as a thanks to your incredible service?

[larspalo]

Is there any means that I could make a small donation as a thanks to your incredible service?

Thanks, but there are many millions of people in this world that would need any donation you can afford more than I do. Be kind to other people, that's enough.

[ultimatebeta07-rgb]

Thanks, but there are many millions of people in this world that would need any donation you can afford more than I do. Be kind to other people, that's enough.

Sure. Understandable mate. That was insightful!