Question about Chris's compressor and fade in-out plug-ins

I’m trying to create smoothe fades between sections of audio that Chris’s Compressor has and hasn’t been applied to, but I can’t get it right. I’ve tried the Fade By DB plug-in, but it doesn’t do what I need it to do, because there’s no way that I know of to tell what the correct volume level of the compressed audio is, since it’s different every time. When I use the compressor plug-in, I usually set the ratio to 0.5, the Attack and Release width to the minimum values, the Attack and Release exponents to the maximum values, and the floor to -96 DB, while leaving the noise gate falloff unchanged. Any ideas on what I can do to get smoother fades between compressed and uncompressed sections, etc?

Usually Chris’s Dynamic Compressor would be applied to the entire track.
I guess that you have some specific reason for not doing that. If you could describe the particular job that you are working on and what you are trying to achieve then we may be able to suggest an effective method for achieving it.

Hi Steve. Thanks for the reply. I’ve got some recordings of the band I’m in, and I don’t want the music to have compression, but I want the stuff between the songs to have it. The recordings were made on a Yamaha Pocketrak recorder, on a low gain setting. They’re good recordings, but I want the background stuff between songs to be louder than it is, and applying the compressor to the selected audio makes this possible, but then there’s a sudden spike in volume, and a sudden drop when the compressor isn’t applied any more. Therefore, I need to fade the compressed sections in and out. I’ve been using Fade By DB to do this, but it’s not getting smoothe fades, because I don’t know what DB level to start from.

There’s probably several ways to do this, but this is how I’d approach it.

  1. Duplicate the section(s) that you want to compress. (shortcut: select the section then press Ctrl+D) Allow a bit of extra space at the start and end of each of these selected regions. Drag each of these sections onto track 2 and delete any empty tracks as you go.
  2. Select track 2 and apply the compressor.
  3. If you have used Chris’s compressor on the entire track 2 then it will join all of the sections together which you don’t want, but you can easily split track 2 up again with “Edit menu > Clip Boundaries > Detach at Silences”.
  4. On track 1, select the part that you want to replace with the compressed version (this should be a bit shorter than the equivalent section on track 2 as that was given a bit extra in step 1). Then “Split Delete” to remove that section (shortcut: Alt+Ctrl+K)

You should now have something similar to this:
tracks000.png
5) Using the “Fade In” and “Fade Out” effects on the overlapping sections, create cross-fades from track 1 to track 2 and back to track 1.
You should now have something similar to this:
tracks001.png
6) When you have done this for each section you can export your finished masterpiece and it will be automatically mixed down into a single stereo file.

I’d suggest that you practice this method on something unimportant before doing it for real. Also, be sure to keep a backup copy of your original unedited recording.

Thanks, Steve. I was hoping that there would be a less time consuming way to do it, but I guess there isn’t one. I’ve actually tried a similar method before, but I didn’t like sitting there an waiting for Audacity to catch up on longer files…I only have a 1.73 GHZ CPU and a gig of RAM, which isn’t much…

I’ll see what else I can do. I have also downloaded the Amp Fade plug-in, which may be useful. Since I’ll be dealing with two tracks, would Auto Duck work?

There probably is, but as I said, that’s the way that I would approach it.
A small shortcut would be to miss out step 4 and then in step 5 you only need to apply a fade-in fade-out to the audio clips in track 2. This will save a fair bit of time, but you need to be aware that where the clips exist in track 2, they will be a bit louder than expected because they are also playing in track 1. You can compensate for that by turning down the gain slider a little on the left end of track 2.

Hi all,
I just started using Audacity, so this may be a noob question, not sure. When I uses Chris’s compressor (1.2.6) over the entire file, I get a ‘fade out’ at the end. Is this normal? I’m working around it by adding a bit of extraneous audio at the end so that will get the fade and I can just cut it off, but I don’t know why the compressor would be doing a fade-out to my audio. Any ideas?
//Dan

Yes it’s ‘normal’ for the end to be a bit unpredictable.
The issue is that the effect changes the ‘gain’ (the ‘level’) of the audio not only based on the current level, but also based on the level of the audio that is coming up next:

If there is a very loud bit coming up, then the effect will start to reduce the level of the track before it gets to the very loud bit so that you still get a sense of the loud part being louder than what came before. Conversely, if the sound coming up is very quiet and the current level is not too loud, the effect will increase the level in advance of the quiet part, so that the quiet part will not be too quiet, but still quieter than what came before. This “anticipation” is called “look ahead”, and it enables the effect to keep the level within reasonable bounds (“reducing the dynamic range”) without causing sudden obtrusive changes in volume.

The problem at the end of the track (or end of the selected audio) is that the effect has no idea of what would come next if there was something. It therefore just assumes a level in its code. If it guesses too high, then there will be a bit of a fade out. If it guesses too low, then the level will tend to increase at the end. The best workaround for the issue is to do what you are doing, and pad the end with a bit of sound at an appropriate level, then remove it again after processing.