How to fix harsh default sound quality (when saved as mp3)?

When I export an audacity file and save it as an mp3, upon playback of the mp3 I get a very harsh sound quality that is hard to listen to with the volume turned up.

The default sound settings on earlier versions of audacity worked well for me. Not so with this version.

Can someone tell me how to mellow out the default sound quality?

Thanks in advance.

MP3 rearranges sound elements within a performance to make the file size come out smaller if it needs to. So if you pick a small file size and you have an almost overloading but not quite performance, you could generate a terrible sounding MP3. The MP3 will actually be in overload and clipping for real.

While you have a problem performance on the timeline, reduce the volume by 1dB before you make the MP3. Effect > Amplify: Amplification -1dB > OK.

Then make a default MP3 (128 quality or higher, maybe 256).

Still getting the distortion?


There is one other place to get MP3 distortion. Audacity export quality values only work if you start out with a perfect song. You got the music from an audio CD or you performed it yourself. If you got the work from another MP3, the real quality is the combination of the two MP3 files. So your 128 export of a 128 original MP3 has a combined quality of about 64, barely adequate for stereo.

MP3 files are not an open-ended invitation to post production and editing.


Thanks Koz! I appreciate the suggestions.
My effects menu, including “amplify,” is all grayed out in advance of recording. So I made the 1 db reduction after recording but before making the mp3, and I raised the default mp3 to 256 kb. I’m still getting the clipped sound on the resulting mp3.
The project I’m recording is some guitar play-along practice segments off an instructional video at a guitar site I belong to. Would the problem then likely be with my source, i.e. the site?

I know meter watching is an acquired taste, but you should be watching the sound meters during recording. Audacity made that much easier to do in versions 2.1.0 and 2.1.1. The meters are very large with color changes (if you want them) during the recording process.

Keep the meters in the yellow region (-6 or so) during recording.

As long as the meters never make it into the red and certainly never all the way up, there should be little or no distortion in the capture show. That leaves the source material.

It does worry me that your shows sound good but only the MP3 sounds ratty. Properly made MP3s are indistinguishable from the original show. Even Work-Alike software like Lame does very, very well. You don’t have to have a Fraunhofer-Gesellschaft license to make a good MP3. I bet you didn’t know it was licensed software.

There is one other possibility. MP3 music goes through more processing than listening to a straight song. How many different MP3 players exhibit this damage? Just one?