Well firstly there's going to be some degradation going in to Ableton as it converts it to .wav, then you're going to effect it, limit, and convert it back to 320.

Personally, I wouldn't do it, but I hate MP3's in the first place so maybe you'd be fine with it.

Not sure why you feel you need to limit it, as if they are ripped from other tunes they're already mastered and probably have a ton of compression, which a limiter will only add to if it's too loud.

Technically it shouldn't be too loud as it's a commercially released track - you'd be better off just normalizing based on RMS.

There a great tool called mp3gain which analyses various factors and tells you the difference between tracks and can standardise them to within 1.5 db of each other (that's all that's possible with MP3).

I would just work in wav and normalise the samples if really needed at all and just leave it at that.

i wasnt sure if ableton would be boosting any input by default thats why i was asking(ive put in mastered tracks and it was hitting the red). i could just burn the exported wav since i dont use mp3 cds anyways. sure, ill check out that program.

I've done this. It's fine. Would it be better to just use WAVs in the first place? Sure. But if you just have 320kbps MP3s, it probably won't make much of a difference.

It depends how professional you want the output to be. If you're happy with it sounding slightly rough, ignore what I'm about to say - but for the best results, the following might be helpful:


When you add layers to a mastered track, you'll often see a noticeable increase in the amplitude of the output waveform. The original track will have been compressed so that it 'fills' the dynamic range and adding a sample in Ableton will simply add to that.

As it's very unlikely you'll just be adding a constant loop throughout the track, you'll see sudden changes in the peak amplitude when the parts you're adding start and finish, although it might not sound wrong listening to it.

The resulting waveform (zoomed out) might look something like this:



The wider bits are where you've added a loop or some samples or something, the narrower bits are the amplitude of the original track.


The original track will probably peak at 0dB (or very close to it), so for a start to avoid your output clipping when other sounds come in you'll have to bring down the volume of both original track and added bits so that the overall peak is below clipping level.

But you'll probably also want to even it out, as when you play it out the higher peaks could well trigger a limiter - you won't have any control over the effect that'll have and it's unlikely to be too nice.

You could just put a limiter on with the threshold set to just above the peak level of the original track - that way, only the sections with added parts will be affected and the rest of the original track will be left alone. But the affect on those sections with added parts will be really harsh - remember the track's already been compressed once when it was mastered, this will add some more really harsh compression to the louder instruments.

The only way to really control it is to apply some compression across the whole output, probably still with a limiter keeping peaks to a fairly consistent level. Applying softer compression first will 'spread' the effect of having to limit it across the whole range, so the result won't be too harsh on any particular sounds or sections of the track.

But of course you're still re-compressing something which has already been compressed once, so it's likely to end up sounding a little bit more squashed than the original. Which isn't something you can avoid unless you can get hold of an unmastered version of the original!

___________________
Stu Cox |

re-mastering mastered tracks isnt that of a farfetched idea, if there is room for it, you have to judge by waveform displacement and peaks. compression was however, not part of my plan. you have inspired me to do more tests though stu.