#8 - My friend, the compressor!

Hi friends,

For a location sound mixer, recording dialogue can be a challenge. Dialogue can be very dynamic. The softest dialogue as compared to the loudest dialogue, (by the same artist), can easily have a 30 db swing!

In film school, we were taught how to use a compressor. But back then, I never paid attention to how a compressor could even out the dialogues, so that a soft whispered dialogue and a loud shout can be heard comfortably in the theatre. I am sure, our professors taught us all this, but somehow, I can't seem to remember it. One of my first jobs after passing out was at a place where we did not use compressors while recording music. It was a purist way of recording music. So I spent several years of my early life, working without compressors.

For the first time in my life, I actually realised the importance of a dynamic compressor when I was shooting a television show, sometime in the nineties. I was new to multicam television shows. A 12-channel analogue mixer was par for the course then!

So, on one of the shows, I was the production mix engineer and the host of the show was a lady with a very loud voice. At times, she was soft, at times she spoke in a normal tone, and at times she got very loud. Today, I realise that this is absolutely normal in a television show (or for that matter, it's normal everywhere!), but back then, trying to record this kind of voice using a wireless lapel mic, WITH NO COMPRESSOR IN THE SIGNAL PATH, was a huge task! I had to ride the faders constantly! This got me thinking. There has got to be a better way to do this.

Enter the dbx compressor! I don't know how or where I must have seen it, but I managed to get one stereo unit on my shoot. And lo and behold, my problem was solved! Now the host could scream her lungs out, and I was cool with that!

dbx166

We then shifted from analogue compressor units to XTA’s C2 digital compressors. This was a giant leap for us since it was a digital unit and had a look ahead feature! It sounded great! What we did not realise was that it introduced a latency of a few milliseconds and that was enough to create a flanging effect if all the busses were not routed through similar units. It took us some time to figure out that digital gear always has latency and even a few milliseconds can have an audible effect on audio.

XTA C2

Once I started using compressors, I began to read up about them. Long long ago, I read a very nice explanation for a layman to understand what a compressor does. The person who wrote that article (or whether it was written in a book, I can’t remember) explained the working of a dynamics compressor in a very easy way to understand.

Imagine a kid on a trampoline in his house. The harder the kid jumps on the trampoline, the higher he goes up. If he were to jump very very hard, he might hit the ceiling and bump his head. Our audio signals are like that. The louder a person talks in front of the mike, the stronger the electrical signal coming down the wires. If the person screams very loudly into the mike, the electrical signal coming down the wires can get so strong that it hits the limit of the circuitry. In effect, the audio signal will distort. To prevent that from happening, if we put a reverse trampoline over the kid, as he jumps up and down, he gets protected from hitting the ceiling. How high above his head we put the reverse trampoline is the THRESHOLD. How hard the trampoline material is, becomes the RATIO of the compressor. A softer ratio will allow the kid to move more before the trampoline pulls him back down. A harder material (higher RATIO) will not allow him to move up much at all. This is a very simplified way of understanding how a compressor works.

Over the years, we have moved from analogue consoles to digital consoles. We no longer need to carry outboard compressors and insert them in a channel or a bus using an insert cable. Nowadays, every digital console has a dynamics processor (a compressor) built into every channel. And we also use many many more mics on a show. So when signals are bussed, things can easily go out of hand in terms of levels. A compressor helps the engineer to keep things in control.

What I have realised over the years is it is better to gradually compress a signal. You start by inserting your compressor in your channel at a low ratio, then you add another compressor in the bus (where many mics are combined together) and then you can add a final compressor in your stereo bus. This way, you can tame your signals better, rather than go in for a very strong compression in your channel. This is how I have used compressors and this may not be the best method….but in audio, there are many different ways to do the same thing and another engineer might have a different method.

Another place where compressors can be very useful is when you have to give a dialogue mix. The smaller the loudspeaker on which your mix will be played back, the more the compression that you need.

For example, a film soundtrack when played in a theater, or cinema hall, can have a very wide dynamic range. When the same film is to be viewed on a television at home, on say decent speakers, you need to compress your tracks a bit. If the same film is being watched on a laptop, you need to compress the tracks still further. And if the same film is to be seen and heard on a small mobile phone speaker, you need to have it compressed a lot.

Another factor to keep in mind is that in a cinema hall, the ambient noise floor is low. The AC is quiet. There are no sound leakages from outside. So the ambient noise floor is generally low. As we go from Cinema hall towards the mobile screen, the ambient noise gets higher and higher and the playback speaker gets smaller and smaller. A suitably compressed track will make it easy to listen to your dialogue track, even on the tiniest of speakers.

On many YouTube videos, a professional sound mixer does not get to work on the final mix. And in many such videos, there is invariably a pre recorded song or a music track, added to the mix. While watching such a show, one needs to turn down the volume on the television when the song starts, and push it back up when the song ends and the dialogues come in.

Songs are recorded with large diaphragm mics, one inch away from the singer's mouth. Most musician's instruments are also recorded with close miking techniques. These tracks are also heavily compressed. The mixed song is mastered and has much more energy compared to a dialogue track, recorded with a hidden lapel mike and an overhead boom. A mixed and mastered song can easily be kept 20 dB below the level of dialogue to match loudness levels. Newbies who are mixing audio for a video fail to understand this. They tend to mix by watching levels, rather than listening to the relative loudness between dialogue and music.

Coming back to recording dialogue on location, one important thing to keep in mind is that we recordists cannot (and should not) tell artists to maintain one single solitary audio level. It's their job to act, and if they want to be loud and soft whilst saying their dialogue, it's their choice, because it's their performance. What we can do is leave enough headroom on the transmitters and in our recording chain that we capture the full dynamic range of the artist. Even if editors complain of low levels, ignore them. The dynamic range can be controlled later on in post. It's the job of the post production dept to handle this.

Generally, film location recorders only have a limiter and no compressors in the signal chain. Since we don't have compressors, we need to record a bit low to make sure we don't hit the limiter very hard on a loud dialogue.

And on multicam television shows, where we have big digital mixing consoles with the luxury of compressors on all channels, we should use the channel compressors with gentle ratios and compressors also at the mix bus stage.

Hopefully, my explanation of the working of a dynamics processor (compressor) will help someone understand how a compressor can help in making a mix sound smooth and help one do better audio work!

Until next time,

CC

E-mail: cheerag.cama@gmail.com
Twitter: @cheragc
Blog: www.thesoundblog.in
Website: www.cheeragcama.com