I record music using multiple (and sometimes many) tracks in audacity projects. Sometimes, I have made mistakes with the recording levels and, consequently, gotten clipping on those particular tracks. But, even when ALL of my tracks are recorded well and there is NO clipping on any of them, it still ALWAYS clips (and VERY BADLY!!) when I play back the tracks together/simultaneously. It sounds fine to me when I monitor this multi-track playback through my headphones. What is going on here, and will this be a problem when I export these recorded songs to either WAV files or mp3files and other people play them on their own computer/audio systems? What is the point of having this great multi-track recording software if you always get clipping when actually using/recording multiple tracks?
When playing multiple tracks you have to reduce each track gain accordingly.
e.g. if you have two tracks peaking at 0db, you’ll have to turn both track gain sliders
down by 6dB to ensure they cannot clip when played simultaneously.
When making an mp3 of your project you have to leave headroom of 1 or 2dB to avoid clipping,
as the mp3 encoding process can increase peak volume by 1 or 2dB.
That’s perfectly normal!
Mixing is done by summation. With digital audio data the samples are summed and analog mixers are built-around summing amplifiers. Anallog mixers have a level control for each track, plus a master control so it ends-up more like a weighted average instead of a simple summation.
Audacity has level sliders for each track (to the left of the waveform) or you can use the Amplify effect with negative “amplification” (attenuation). But Audacity doesn’t have a master level control.
Audacity itself used floating-point internally so it won’t clip, but you can clip your DAC (digital-to-analog converter) when you play it back at full playback-volume.
If it goes over 0dB and you export as regular (integer) WAV, it will be clipped to 0dB.
One “foolproof” solution is to export as floating-point WAV, which can go over 0dB without clipping. Then re-import the floating-point mix and run the Amplify or Normalize effect to bring-down the level before exporting to your desired format.
What is the point of having this great multi-track recording software if you always get clipping when actually using/recording multiple tracks?
Pros often record at MUCH lower levels and they also leave headroom in the mix. Then after mixing the levels are brought-up as part of a separate mastering stage. And pro DAW software does have a master level control, and meters on every track as well as a master meter.
If you mix a track with an exact copy of itself you’ll get +6dB (double the amplitude). If you invert one copy before mixing you are subtracting (adding a negative) and you’ll get dead-silence.
Many thanks to Trebor and DVDdoug for the information. It is very helpful. I figured that lowering the individual/track gains was one way to compensate for the clipping caused by multiple tracks, but I also knew that, as the number of tracks gradually increases, there would be CONSTANT re-mixing (i.e. volume/gain adjustments) to make. And that seems like it would potentially be a lot of extra work. Yes, master controls/sliders would certainly make the whole process easier. But, as long as I know what I have to do to deal with these limitations, I can compensate with some extra time and effort being invested.
Just one more question on this general issue of record/playback levels. I have noticed that I will sometimes (or even often) get a recorded track that looks either normal or even low when looking at the track waveform, but when I look at the playback level meter it is clearly showing lots of red as it clips near or at the 0 db point on the meter. Why do they seem to be giving conflicting or contradictory information?
Thanks for the help,