I'm trying to convert an audio file to an FFMPEG waveform, train it in HyperGAN, and produce an image file to convert to an audio file. What I'm asking is if there is a way to convert a waveform image back to an audio file without any loss of quality (basically, if you were to hash the original and the converted one, it would be the same hash), and if FFMPEG is the correct tool to use. Currently I have a batch script to go through all my audio files and convert them to images (ffmpeg.exe -i %1 -lavfi showwavespic=s=1024x800:colors=0971CE %1.png), but I can't convert them back to audio files, which defeats the purpose of what I'm doing.
EDIT: FFMPEG converts a 5mb audio file to a 32kb image, so I don't think it would convert back. Is there anyway to make it more detailed?