Dynamic Signal Processing - Posted on 2014-05-31 by Jin Huang

Talking about the dynamics of a song, there are four types of processors we can use to control the signals: compressor, limiter, expander and gate. As compressor reduce the dynamics of the signals, expander does the opposite which can be handy when you want to separate things more. Limiter is more like a compressor, but the ratio is set high and is only involved when strong transients are trying to poke through. Gate allows only the signals above certain threshold to pass, which is very useful for reducing background noises; this is the same idea as the noise gate on a guitar pedal. Now let’s see some of the details.

Date of Session: Apr 16, 2014

Time in: 11:00

Time out: 13:30

Lesson 14: Dynamic Signal Processing

Of the four types of dynamic signal processors, apparently compressors are the ones that we are using all the time. Back in the days when compressors was not yet to born, even a vocal track needs to be recorded three times for the soft, normal, strong parts. Audio engineers will separate different parts in different track so that the gain can match each other. Now with compressors, they bring up the weaker part of the song while compress the peaks according to a defined ratio, which makes the work a lot easier, especially for vocals and drums with large dynamic ranges. 

There are several important things to remember when applying compressors. First of all, compressors cannot be used merely to turn up volume because it’s reducing the dynamic range at the same time. Second, if the attack is set too fast, transients and bottom end of the signals will be lost, and this is not what we want for drums. Third, compression is multiplicative; once your signal is over compressed, you are losing the dynamic range and cannot take them away. Also, don't compress classical music. Different genres of music require different compressions, but when it come to classical music, no compression absolutely because the conductor is the person who builds up the music; a compressor will take a way the conductors work.

Lionel showed me the differences of two most widely compressors: the 1176 and the LA2A. The most significant difference between the two is that 1176 is hard knee compression while LA2A is soft knee compression. Hard knee compressors are good at catching things that are really fast like drums while soft knee compressors are better for something slower like vocals so that you don’t really hear when it’s working. Lionel has build some 1176s with his brother Dave, and he’s going to build some LA2As too. If Lionel has both compressors, he prefers to use them in serial. The LA2A will be set in a lower ratio to do most of the compressions, and the 1176 set in a higher ratio to be used as a limiter.

In terms of serial compression, it’s one way to use multiple compressors in serial, which means that the output of the previous compressor goes into the input of the next compressor. On the other hand, parallel compression is what happens when two compressors are used in parallel. That’s usually when we like the uncompressed signal of the sound, but we also like the compressed signal as well. Parallel compression is a solution of having both uncompressed and compressed signal together.

One more thing, depending on the components, different compressors will sound differently. This is not only applied to compressors, but also applies to EQs, pre-amps, mixing consoles, monitors, everything. Some equipment tends to add colours to the sound whereas some tends to be very transparent. What we want to do is to put more colours when tracking so that the recording is interesting with some unique characteristics; however, when it comes into mixing, keeping things transparent will be the key. Of course we can mix with equipment that adds colour to the mix, but in that sense we need to kind of taste what we can get from it along the way. 

Lionel also talked about how important it is to have a playlist that travels with the engineer. That was more about the translations of a room, so I’ll be talking about it in the chapter of acoustics and monitoring. This time I brought Lionel a reference track to ask how I can achieve that kind of clear sounding vocal, which also has something to do with another reference track I brought to Lionel. They are basically the same idea, so I will write about them in a latter blog when I’m talking about vocal treatments and mixing. That’s it for compression, more to come.