Kinyugo

@mthrok Thank you for the guidelines on how to contribute. About the implementation: 1. Borrowing an idea from other implementations, I think it is possible to define the operations using...

I have also managed to port a 1 to 1 implementation of the inverse. As it is supposed to be an invertible operation we can check the reconstruction error to...

@mthrok I managed to get a working vectorized implementation of the MDCT algorithm. However, I have trouble setting up the development environment on my local machine to open a PR....

Hello @carolineechen. I followed the steps in the contributing guide but ran into some problems installing the package i.e: `python setup.py develop`. While building `sox` I am running into an...

**Full Error Message** ``` /home/kinyugo-maina/Learning/ML/AudioGeneration/torchaudio/audio/third_party/sox/../install/lib/libvorbisfile.a /home/kinyugo-maina/Learning/ML/AudioGeneration/torchaudio/audio/third_party/sox/../install/lib/libvorbis.a /home/kinyugo-maina/Learning/ML/AudioGeneration/torchaudio/audio/third_party/sox/../install/lib/libogg.a /usr/lib/gcc/x86_64-linux-gnu/9/libgomp.so /usr/lib/x86_64-linux-gnu/libpthread.so && : /usr/bin/ld: cannot find -lmkl_intel_ilp64 /usr/bin/ld: cannot find -lmkl_core /usr/bin/ld: cannot find -lmkl_intel_thread collect2: error: ld returned 1 exit status...

The issue is still persistent.

Hello @carolineechen, I will look into creating a draft PR, as I would also appreciate some feedback on some design choices that I made. Godspeed on resolving the issue. Thanks...

Hello ✋🏿 Thank you for your insights. I have made the changes and I can confirm that it works. I see that `torchaudio` took quite a different approach to the...

Thank you for your time. I will stick with `GriffinLim` with a few iterations for now. I might also look into neural vocoding approaches. Feel free to reach out to...

Hello Flavio. It makes sense not have dynamic thresholding. Have you experimented with the effects of clipping on the final sample quality?