Spectral cut-off when converting to MP3

Question

So I recently tried exporting a .flac file to .mp3 in Audition, and upon looking at the spectrogram of the converted file, there's this odd, unwanted cutoff at about ~16khz:

This was weird, because when I exported the same file with the same exact settings in a different program (Audacity), there was not as much loss in high frequencies as when I used Audition. Here it is for reference:

Can anyone please tell me why this happens, and how I can export my files in Audition to look more like the second one? For reference, both above files were exported at mp3 320kbps, 24-bit, 44100hz, so I do not think this is an issue with my export quality being too low, or something like that.

Thank you for the responses in advance!

SteveG_AudioMasters_ · Answer

MP3 coding is complex, and no two encoders do it the same. Most cut off quite a bit above 16k, because the algorithm they use says that this is a good place to get what's known as a 'coding gain' - which is how MP3 files end up being quite a bit smaller than the files they are based on. This is the masking process at work - basically reducing the parts of your audio that the algorithm it uses thinks that you can't hear, to a much lower bit depth. Most encoders also think that anything above 16kHz is fair game for this process, because adult ears have a much reduced ability to hear them.

Audition uses the Fraunhofer encoder, and currently there's no option to change this - although in theory there's the option to now, as MP3 isn't 'protected' any more. A lot of people prefer the sound of the LAME encoder, which is possibly what Audacity uses, but if you want to use that, at present you'll have to find another app that uses it.

Sign up

To post, reply, or follow discussions, please sign in with your Adobe ID.

Sign in to Adobe Community

To post, reply, or follow discussions, please sign in with your Adobe ID.

Scanning file for viruses.

This file cannot be downloaded