I have generated some Mel-spectrograms using librosa to use it for generative adversarial networks(GANs). I have saved the generated spectrograms through GAN in image format(.png). Now I am trying to convert the images back to audio. Is it possible?
Asked
Active
Viewed 576 times
0
-
Does this answer your question? [Can I convert spectrograms generated with librosa back to audio?](https://stackoverflow.com/questions/61132574/can-i-convert-spectrograms-generated-with-librosa-back-to-audio) – Lukasz Tracewski May 29 '21 at 14:17
-
@LukaszTracewski Can you share the code snippet to implement this for spectrogram images – Nishani Kasineshan May 29 '21 at 14:59
-
`S` in that code is your spectrogram image. – Lukasz Tracewski May 29 '21 at 15:10
-
@LukaszTracewski I have saved some spectrogram images in png format. I want to convert it to audio. Is this possible? – Nishani Kasineshan May 29 '21 at 15:16
-
Yes, it is, though it's not a good approach. Read the post and comments. – Lukasz Tracewski May 29 '21 at 16:22
-
Thank you @LukaszTracewski. Can you suggest me a documentation where can I get the idea? – Nishani Kasineshan May 29 '21 at 21:31
-
I am not sure what you are asking. What documentation? – Lukasz Tracewski May 30 '21 at 20:42
-
[amar nabil taabane](https://stackoverflow.com/users/16613331) posted an [Answer](https://stackoverflow.com/a/68692808) saying "i find this for spectrogram .. [https://github.com/sikora507/elgen/blob/master/src/audio%20analysis.ipynb](https://github.com/sikora507/elgen/blob/master/src/audio%20analysis.ipynb)" – Scratte Aug 17 '21 at 22:47
