By Decklyn
Intro
--------------------------
So I’m stuck on my tune, wore down and out of inspiration.
As deadly habit has suggested, this is the perfect time to stop putting out music and give back to the community a bit, so today that’s what I’m doing.
This tutorial will cover a slightly more advanced compression techniques – parallel compression. It assumes that you know the basics of what a compressor does, and I also assume that you know at least a little bit about working in Cubase. Anything like Group channels, which some people may not use, I try to walk through click by click though.
What is Parallel Compression?
--------------------------
Parallel compression, in its most basic form, is a compressed signal running parallel to a dry signal. This is generally obtained by using a send loop with a compressor, so that we can control the amount of signal being sent to the compressor. It’s essentially the same thing as running a reverb effects unit via a send.
Why use Parallel Compression?
--------------------------
Straight compression on an insert reduces dynamic variance, so it can really sap the sense of dynamics from your track. By bleeding in some of your original signal, you get the best of both worlds. You get a generally loud signal from the compression, but you get the dynamic qualities of your original signal as well.
Also, single band compression looks at the loudest peak in the frequency range. If your singer, for example, hits a hard “P”, this will trigger a spike in the mid range frequencies. This will cause the whole signal to be compressed. What happens, then, is that the high frequencies/sibilance, get compressed when there is a spike in the low frequencies. This is undesirable. We then get the offbeat pump and breathe from the release of the compressor when the highs snap back up. By running compression in parallel to the original signal, we still maintain the highs, even when the compressor is working hard.
We can also see the usefulness of this on drum tracks, where a hard kick or snare will squeeze the hats in a very undesirable way. By using parallel compression, we can ensure that our hats stay clear in the mix.
How to Apply Basic Parallel Compression using an Insert
--------------------------
There are primarily 2 ways to apply PC. We’ll be looking at some percussion in a track that I’m working on and applying it both ways. Before we start compressing, we need to deal with a few things though.
As we have multiple drum tracks, we first need to group them into a group/bus. So start by making a new group. Right click as shown and select “New Group Channel Track.” Select “Stereo” and hit okay.
We will call it “PC”. Click on the new groups name and rename it.
Now send the output from each channel that you want included in the processing to our group. To do this, click on the channel in the Project view to select it, and then click on the output, and select our PC group, as shown. I drew some red marks to show you.
You can just select some of the elements, if you like. I normally would only take elements that I want to clearly emphasize. Ghost hits I would probably leave. However, in order to show the effect, I’m going to use all of the elements in the percussion track.
Now, as mentioned, there are two ways of applying PC. The first is reliant on your compressor, but it is the simplest way to do it. (This is also a great trick for mastering – try the following over your master bus).
We’ll get back into Cubase and put a compressor on our PC group via an insert. I will assume that you know how to use inserts. I’ll be using the PSP Mixpressor. (Later I will recommend a couple free VST compressors with a mix control).
Now play with the compressor settings and presets, with the mix on at 100% until you have a nice sounding setting. Then set the mix setting (circled below) back to lower than 50%. The compressed signal is going to be louder by average, so we want our original signal to be perceptible in there, but re-enforced by the compressed signal.
Check the audio sample out now. The first loop is uncompressed. The second pass is PC using this method.
http://www.epicacademy.com/downloads/de ... orial2.mp3
How to Apply Basic Parallel Compression Using a Send with Any Compressor
--------------------------
Don’t close your project yet! We’re going to remove our compressor from our group now. Again, I’m going to assume you know how to do this. So our group channel should be empty now like so:
Now we are going to apply parallel compression as it is used in hardware. I don’t think rack mount compressors generally have a “mix” knob! So we’re going to do this in a more classic way – using sends.
First, make an FX Channel Track similar to how we made the Group Channel Track above.
Make it stereo, and select a compressor in the Plug-in field. Here we’ll use the URS 1970 console compressor, a nice sounding (expensive!) vintage modeled single band compressor.
Then, go back to our Group Channel Track. Now create a send to our FX Channel Track (it will be entitled with the name of the compressor you chose).
Get back into your compressor by hitting the little e button (circled) and play with the settings. To get a bunch of bite, I selected a longer attack and a fast release. You can solo the FX track to hear the effect of the compressor.
Check this audio sample. It's the same file as the other one. The sound is so similar between insert and send methods that I couldn't justify making another file.
http://www.epicacademy.com/downloads/de ... orial1.mp3
Notes on Basic Parallel Compression
--------------------------
The examples are subtle – and intentionally so!! If you want your cake, and you want to eat it too, you’ve got to be cautious. You can probably mix in a BIT more compressed signal than I did, but you want the compressed signal to be tucked under your original signal, not the other way around, otherwise all your work is not really worth it, as you’re smashing the dynamics again. Onward to the more dynamic New York Trick!
The New York Trick
--------------------------
If you’re listening to examples and sort of wondering if this is all really worth it, then you may want to check out the New York Trick, which will build on the previous examples.
The New York Trick is parallel compression with rather heavy EQ applied to the compressed signal. The addition of EQ to the compressed signal just helps bring out some more energy in the percussion, and gels it together.
This is very easy to do now that we have the PC set up. Bring up the FX track by selecting it and clicking the little E back in the project view.
Using your favorite EQ, boost 6-10db at about 100hz, and 6-10db at about 10khz. I used the Cubase EQ.
You may need to adjust the level of the compressed signal again so that it sits under your original signal.
Here is an audio example of this in practice. The first half is dry, the second half is with the New York Trick applied. I think I might have snuck the signal a little bit too far under, but you can definately still hear the difference.
http://www.epicacademy.com/downloads/de ... orial2.mp3
Phasing Problems
--------------------------
There is one pitfall to watch out for here. And that is phasing.
Plugins can add a small ammount of delay to a signal when it is being processed. Most software environments account for this, so there is no need for consideration here. However, if you are experiencing phasing using sends, just drop another copy of your compressor on the insert channel of your dry signal, and turn the ratio to 1:1 and/or the threshold all the way up. The compressor won't do anything to the signal other than delay it so it has been delayed the same ammount as your original signal.
I hope this guide helps someone!
Tune that the sample is from:
http://dubstepforum.com/viewtopic.php?t=18925
