6

I have recently noticed that the order of the VST plugins on an audio track can make a great difference on its sound. I usually use virtual preamps, compressors, distortion and chorus effects, equalizers, reverb and delay. I also work with MIDI tracks.

So are there general rules or tips on how to organize VSTs? How does the order affect the sound exactly? How about track folders, as they already add an effect on all the sub-tracks?

I would like to know if there's a way to arrange VSTs properly, so that every plugin uses its full potential.

7

Are there general rules or tips on how to organize VSTs? (...) I would like to know if there's a way to arrange VSTs properly, so that every plugin uses its full potential.

There's no correct order. It's all about preference, what you want to achieve, and the context.

You'll find a lot of suggestions, like putting time-based effects (like delay or reverb) at the end , and compressors at the start. These make a lot of sense in some scenarios, but shouldn't be taken as absolute rules.

This page, for example, suggests this order:


filter -> dynamics -> distortion -> eq -> pitch -> modulation -> gain -> time

This page suggests this order:


utility -> filters -> intelligent processor -> dynamics -> gain -> time ->

Both are very similar. That pattern is very common for electric guitars (and it's the most commonly recommended pattern), but what about other contexts?

In mastering, if you are going for 0dBFS, you definitely want the limiter at the very end of the chain, so it would be a big mistake to implement this pattern there. In mixing you will want some processes in parallel rather than in series (more on that later), and you might want to compress and eq each effect separately. For synths it's very common to put a filter at or near the end of the chain, after distortion. Guitarists break that pattern all the time.

Stop thinking about "what's correct". If you have a sound in mind, start thinking about what you need to achieve it. If you are not sure or you are experimenting, switch the order of things and see what you like most.

Another important thing to consider is not only the order, but the way you are chaining everything. You can go in series or parallel. In series you connect the output of one effect to the input of the other. In parallel you send the signal to two (or more) effects to process it separately.

Series:

Series

Parallel:

Parallel

How does the order affect the sound exactly?

One effect will process what the previous one feeds to it. Think about a gain and wave shaper chain. With the gain before the wave shaper, the gain will also control how much the signal is distorted, it is also functioning as a "wave shaper amount". With the gain after the wave shaper you only control the volume of the output of the wave shaper, the gain can't affect the amount of distortion induced by the wave shaper.

Or think about a pitch corrector and a pitch shifter. If you put the corrector before the shifter then you can go wherever you want with the shifter, but if you put the corrector after the shifter the corrector will try to correct whatever you are doing with the shifter, will try to put the frequency of the signal where it thinks it should be.

To know what will happen, you also need to know what the effects do. Once you know exactly what the effects do, then knowing what your chain is doing becomes easier. Your chain and order is explicitly and literally telling you "I'm doing this, then that, then that, and finally that".

How about track folders, as they already add an effect on all the sub-tracks?

It's pretty much the same. The only difference is that you are feeding the sum of all those tracks to the effects chain, instead of one just track.


Related questions:

Why does pedal order matter?

Where to apply my effects pedal?

Where should I put delay pedal, before reverb or after?

3

There are basically two classes of effects: linear ones and non-linear ones. From an idealized signal processing view point (discounting noise and quantization), the order of linear effects can be interchanged without a net change in the result. A lot of effects are also time-invariant: start your signal an arbitrary time later, and the result will be the same, just coming late by the same amount of time.

Notable effects that are not time invariant are Leslies and Wahwahs since they come with their own independent timing.

Effects that are both linear as well as time-invariant include a lot of tone adjustments (like a graphical equalizer) and most delays/reverbs.

So the basic setup boils down to how one arranges the non-linear elements in between linear blocks.

One non-linear element are dynamic compressors. Another one are distortions. Do we want dynamic compressors before or after distortion? Depends on our goals.

The net result will be a tone with stabilized volume. Putting the the dynamic compressor before the distortion will feed the distortion with stabilized volume, leading to a constant distortion quality for sustained tones. Putting it behind the distortion will feed the distortion with the raw volume, leading to a timing- and strength-determined distortion, giving a sound quality that changes over time and can be stronger influenced. Both make sense for some settings.

Now reverb generally wants to give your sound ambience. Putting any nonlinear effect like a distortion after reverb will put artifacts in that ambience that "don't make sense". On the other hand, some kind of delay line play might want to suggest chorus effects in which case it might make sense to put the delay lines before some nonlinear devices.

And so on. A graphical equalizer before a distortion device will not do much useful since clean sound has a pretty sparse frequency spectrum, so you approximately only meddle with the balance of the strings, and the changes in tone characteristics (sound getting brighter or duller) really get lost in the distortion. After the distortion device, a graphical equilizer has much more material to work with and thus is more effective for shaping a particular sound.

So basically the key is to experiment and see which combinations lead to useful changes, and which don't.

  • 2
    Lots of great points. Two remarks: 1. linear does not actually imply that effects commute, to use the mathematical term; only linear operators with a common eigenbasis commute. That is indeed given for time-invariant effects, but a wah-wah and a delay will not commute. 2. You're right that a equaliser before distortion won't shape as much and an obviously the result character, but still it's often very useful, for instance boosting higher mids so the low harmonics won't be too dominant in the distortion. – leftaroundabout Nov 18 '14 at 11:34
1

Put generally, EQ in front of a compressor will change the way the compressor engages. This might be a good thing! Say you have an acoustic guitar track that's got some rumble in it. You put it through a comp->EQ and you're fighting it all the way. You put it through a high-pass EQ first to take the undesirable LF away first, and you get a far more musical result from your comp.

In general: do your forensic/surgical EQ pre-comp, do your overall broad shaping EQ after the comp. This rule may be broken constantly, but it's a good place to start.

In general: do reverbs later in the chain than sooner...they're much more fragile and can be dramatically altered by post-processing.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.