Local SDR testing script

 

https://drive.google.com/file/d/1GC9pwch0WQXZXwBNTz_QnXE_UyxdKmQF/view?usp=sharing by Dill

https://drive.google.com/file/d/1BeqNw3TnRTDMwnoQGMbOqwrcGRwe4Zht/view?usp=sharing GUI by zmis (but it scores a bit lower for some reason):

 

“Here's a handy little python script I made using the help of Ai that can calculate the SDR of a track based off of the actual instrumental or vocal of the song.

 

You can do python sdr.py --help for an explanation on how to use the script.

 

You just need numpy and scipy for it to work, and python ofc!

 

I'm not sure if you would like to pin this or not, but I've been using this script to help me improve my separation methods.

 

https://github.com/ZFTurbo/Audio-separation-models-checker/tree/main

Based on MUSDB18-HQ dataset”

 

“Q: Why SDR goes <0 in silence parts? (song_006)

 

A: SDR and SISDR behave weirdly when 1 of the input is silent, and that's why log_WMSE was made: https://github.com/crlandsc/torch-log-wmse/

 

Interestingly, L1freqMag metrics is giving same results than some users here (1296 a bit better for instrumentals, 1297 a bit better for vocals).” jarredou