Can I separate audio in an mp3 file?

[johnfrei4_1971]

Hello. I inherited a project where the case study videos have music playing in the background under the narration. In one of the videos, the volumne on background music is too high and overwhelms the narration. I only have the mp4 file for the video, which I was able to separate the audio and video into separate tracks and export the audio to a separate mp3 file. Now, I want to separate the audio in the mp3 file into separate tracks (one for narration and one for music) so I can lower the volume of the music track. Can I do this in Audition? Normally, I would scrap the audio completely and redo it from scratch, but I don't have the music file that was used, and I want to preserve consistency across all of the case study videos, which all use this same music as background. Thanks!

John

[SteveG_AudioMasters_]

Your problem here is that you've started with a lossy format, and extracted the part you want to work with into an even lossier one! There are some specialised tools that can attempt the task you want to carry out, but they all work better with  high-quality files. Even if the MP4 is good, the audio should be extracted to a wav file - at least this won't increase the losses.

Audition's tools for working with this type of material only really work with stereo files - generally their use relies on the music being spread across the stereo field, and the narration bring in the centre. With this, it's relatively easy to separate them, at least partially, but if reverb has been used then you tend to get 'ghostly remnants' that are really hard to do anything about. The tools are the Vocal Enhancer and the Center Channel Extractor.

If you do a search using the term 'separate speech from music' you'll find all sorts of on-line apps, and I believe that some of them can work quite well, so it might be worth trying a few of those. Adobe also has a speech enhancer that can help quite a bit when you've got the basics done, although I fully expect you to have some trouble with the music, I'm afraid. Your best bet would be to try to get hold of the music separately and add this again afterwards - you stand a much better chance of working with the speech.

[johnfrei4_1971]

Hi Steve,

Thanks for the reply. I also found the center channel extractor and it seemed to do the job wrt removing the voice narration. The resulting music track sounded OK. As for the voice narration, I was planning on redoing it using one of the TTS voices available in Storyline 360. I will export the audio again as a .wav file per your suggestion to see if that improves the quality of the music track, but so far, the results I got seem satisfactory.