Skip to content

Small python program that converts images to spectrogram and then creates an audio file based on it

Notifications You must be signed in to change notification settings

nschmidtg/image_to_audio

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

4 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Image to audio

Python package to convert image to audio. The system interpretates the images as raw spectrograms and synthetize the audio using customizable iFFT parameters.

Install

pip install -r requirements.txt

Process an image

python main.py --image_path=<PATH_TO_THE_IMAGE> --hop_size=512 --out_name=<NAME_OF_OUTPUT_AUDIO>.wav --sample_rate=16000

  • The higher the sample rate, the higher the content of high frequencies, but the length of the output audio gets shorter
  • The higher the hop_size, the higher the resolution and longer the length of the audio file (it slows down the playback). Very slow hop_size will lead into harmonic artifacts (could or could not be wanted...).

I recomend using images with large resolution (~4000px x 3000px), a hop size of 512 and a sample_rate of 16000 to obtain ~1.5 min of audio.

Test

TODO

About

Small python program that converts images to spectrogram and then creates an audio file based on it

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages