Using Compressors When Editing Sound Effects.
By : David Mann
When you start recording your own sound effects and put them into your Digital audio workstation session you will notice that most times they sound a lot less loud than the one you brought from the library. If those sounds don't need to be loud it is not a big problem; you just need to set the volume where you'd like it to be and you're done.
But if those sounds need to be loud like gunshots, explosions, punches, car crashes, etc. either you or the mixer will have to do something about it.
The first thing that you'll want to do if the sound is not loud enough is to normalize it. Normalizing is a process in which the computer searches and finds the highest volume point. Let's say you recorded a door and the highest point in the sound was -7dB. The computer will take the sound and increase all of it by 7dB. This way the sound is exactly the same, only louder and the highest peak in the sound will be the highest possible digital sound level (for 16 bit sound that will be 1111111111111111). Every sound software packages that I know of has a normalizing function. You will usually find it under the Effects menu.
If after normalizing the sound is still not loud enough (which will happen a lot), there are mainly two things sound effects editors are doing to address the issue:
1) Add more sounds, which we call "layering"
2) Use compressors
Let's start with layering. First of all, if you layer just to get more loudness, that's not a good idea. You should only do that if it's also adding to the richness and the character of the sound. For example, if you cut the sound for a building crashing down, it is very unlikely that you can find the perfect single sound in the library that will cover the crash.
The way we do it is to start layering sounds like huge impacts, explosions, stones crashing, glass breaking, stones debris, metal creaking, metal debris, etc.
Let's see how layering making things sound louder…
When you have one normalized sound and, at the same time, you add another normalized sound, what you get is a mix of those two sounds, but also twice the energy (it's like adding 3 dB of gain). If you had those two sounds, each on its own channel, and send it to the mix stage it will play louder but will also make the view meters on the stage jump 3 dB higher.
So now what happens is that you moved the problem to the mixer. This is not necessarily a bad decision. Mixers are very good at dealing with this problem. However, this is adding work for them, especially when two sounds are still not enough and you ended up adding three, four or even more sounds. It is not just the loudness issue that the mixer need to take care of -- it is also having to put all those channels at the same volume with the same effects and EQ (or to his liking, which is not always the way you like it mixed). So the way I see it is if the production budget is not huge, you better do this work yourself. And, when you send a door slam to the stage, it better be one sound.
(Of course it can be 3 sounds that you mix yourself and bounce to one channel.)
If you're asking, "Why not send one week sound and let the mixer deal with the loudness issue on the stage?", my answer is:
There are almost always clients on the stage when a show is mixed. And, they will probably jump off their chairs when something they want to sound huge sounds smalls and puny the first time they hear it. For me, it's also the fact that if you can't hear the sounds at the right level when you're editing you'll not be able to tell if those sounds will actually work well in the mix.
So let's move into compression. There are two main types of compressors:
• Limiter - limits the sound to a specific volume while preventing a harsh clipping sound
• Compressor- changing the dynamic range of the sound.
A limiter is a great tool to add to sound loudness without very noticeably changing the sound. (Of course, if you go too far trying to make the sound louder and louder, you will get to the point that the sound quality will go down dramatically.).
I once had to cut a golf swing for a TV show. I went to the library and picked up a club hitting ball sound that I liked. When I put it in my session it was very weak. The reason was that when you record a golf swing, there is a very short sound when the club hits the ball. This sound is very loud but very short so you can barely hear it. After that initial impact, there is more sounds created by the club and the ball and reflections from the area where that hit happened. Those sounds have a lot more character than the initial very short click. In this case, the initial click was 15db louder than the rest of the sound, so by putting it into a limiter and lowering the extreme peak in 12 dB, I was able to push the rest of the sound 12 dB higher. This made it much louder, and the fact that the short peak was 12 db lower did make a small difference but I could live with that. Yes, in an ideal world, I would send it to the stage untouched and it would have sounded better that way. But, in the real world, the mixers will not be able to mix it that way because it would have clipped and then the mix would have been rejected by the TV station.
My favorite limiter is the Wave L-1 as it is very easy to use and very easy to see how much your peaks are being reduced. By the way, these days a lot of the digital editing software like Pro Tools have a soft clip protection built in, so don't be surprised that they work almost like a limiter when you simply gain things up or push the volume up. Still, I'd recommend using a limiter as they do make sound better.
The other type of compressor is the dynamic range compressor. I don't use that one very often when editing sound effects as this is really better left for the mixers. Dynamic range compressors are mainly used on the dialog and in the music.
Have fun cutting!