Duration of the recoder

[EzioFenix]

I have a question, so between flac and a wav comes out? Or a flac comes in and a flac can come out?

• A limitation of the previous version was the duration of 10-15 min in mp3 if it lasted longer it was damaged. I'm going to check the new version and see if that keeps happening.

• What I did to avoid that, was to use ffmpeg to split the file and then make a merge with it, perhaps using a library of it will help you to fix the duration problem.

• Let me learn python and if you like I will help you, the question is that I work with javascript / node, but now I am learning python, thanks to the fact that you, my dear friend, have encouraged me to get my hands on python

Pd Thank you for hearing my request.

Sorry if the translation is not good, I speak Spanish

[amirldn]

Hello,

• So yes, as of right now a FLAC/MP3 will go in and a WAV will come out in all cases. The system I have in place to record uncompressed audio then convert it into a WAV is quite a hefty amount of my code and may require a substantial amount of rewrite (changing the underlying library I use for that bit) so I will see if I find the time to do so. Possibly the library has a way of exporting into different codecs so I will look into that since that will not need a re-write of my code.

• I have not tested anything longer than 3 minutes so what you are saying here is new news to me, I will try and look into it and see if I can see why this issue is occuring and try fixing it. Off the top of my head, the only thing I could see that would cause this to happen is your machine ran out of RAM since the audio files stays uncompressed while it being recorded it may end up quite large.

• That is a smart idea to use ffmpeg to split the files. Please try with FLAC and see if the error still occurs and if it is not fixed, I will look into including that for long audio clips.

• That is great! I am glad I have inspired you and no worries, your English is perfectly readable :)

[EzioFenix]

I will carry out some tests to see if the error persists, it is that I plan to connect a super resolution module, that its output is the input of your module and that output is again the input of a super resolution module.

If that is to be achieved, you would have perfect audio for teacher videos in times of pandemic.

In addition, everything would be implemented with python and deep learning.

The program I'm talking about is this, read it, it may interest you.

https://kuleshov.github.io/audio-super-res/

https://github.com/kuleshov/audio-super-res

I have not implemented it because:

1- I don't know how to train models 2- I don't know python well for now

[amirldn]

That would be a great addition to program if you are able to work on something like that! I currently am not too familiar with deep learning since I have not learned it at university and haven't looked into it much on my own but these libraries are very useful so thank you for that.