can descript-audio-codec works below 32kbps outperforms opus at 32kbps?
can descript-audio-codec works below 32kbps outperforms opus at 32kbps?
the results in papers showed descript-audio-codec@8kbps has similar performance for opus@24kbps.
We just released a 16kbps model that you might want to check out!
We just released a 16kbps model that you might want to check out!
thanks. I checked it. and from my test result, it seems the model performs well for general songs or music. but for audio with strong and dense harmonic, the model generate blurred spectrum with strong noise. such as the sound of train whistle or violin.
is there any advice to improve it? thanks
Can you share some audio where you get that sort of noise?
Can you share some audio where you get that sort of noise?
e.g. test with standard EVS test vector (the final part is a train whistle) or mpeg test vector (si01, si03, sm01)