Tuesday, June 23, 2009

Order! Order!

Now that computer processing power is fast enough that most computers have no trouble hosting dozens of plug-in effects and softsynths simultaneously, the potential for experimentation in the studio has never been greater. I've wasted entire afternoons messing around with obscenely long effects chains to see how far away from the source sound I could get. What a lot of beginners don't realize at first is that the order you place your effects in has almost as much influence on how the effects change your source sound as the effects themselves.

When you're building an effects chain using your DAW's insert effects, every effect you add after the first effect is not processing the original, dry sound anymore. Instead, it's processing the sound as it's been changed by all the effects that precede it in the chain. The effects are chained serially. Because of this, it makes sense to think ahead of time when you're building your chains to put the effects in an order that makes sense. Here are some examples:

• No matter what you're doing, EQ is nearly always the best choice as the first effect in your chain. This is so you can cut out the frequencies a particular track might not need. For instance, let's say you have a nice, high string part. Even though it's in an higher octave, there will still be low frequencies present in the sound. In the context of a mix, all those low frequencies are doing is adding mud. By cutting out what you don't need, you can free up a lot of space in your mix very easily. But this can also have an impact on how certain effects work...

• The above technique is probably most useful when using compression. Let's say you have a synth bass sound. Typically, you'll want to filter out some of the extremely low frequencies that most stereos won't accurately reproduce anyway... 40Hz and below is pretty common. Let's say you want to put some nice, tight compression on the sound too. If you put the compressor first in your chain, those 40Hz and below subs are going to be triggering the threshold of your compressor even though you're going to EQ them out after the compressor anyway. To have the compressor trigger in a way that is more representive of what the final signal is like, EQ the unneeded frequencies first, thus letting your compressor work only on the audio you are actually using.

• If you're using any kind of distortion, your reverb and/or delay should go last in your chain. If your reverb or delay comes before the distortion, you're not just distorting the original sound, you're distorting the reverb/delay too, which generally sounds like crap (but if that's what you're after, go for it!) Reverb and delay are generally done as send effects anyway, so this is not a huge issue, just an example, as you can build chains on your sends too.

• Along those same lines, if you're using any kind of filter effect with some distortion, take time to consider if you want the distortion to come before the filter (resulting in the filter cleanly filtering the distorted sound) or after (resulting in a dirtier sound with the filter itself being distorted too). Hell, try experimenting with a little of both. Maybe a hard overdrive on a drum track, sent through a filter, followed by some nice, warm tape saturation. Many softsynths, such as Cakewalk's extremely underrated Rapture, let you experiment with different filter and distortion chains and the effect it can have on the timbre of a sound can be significant.

These are just a few examples. I highly recommend wasting an afternoon or two building stupidly long effects chains and seeing how switching the order of the effects changes the end result. Even if you don't learn anything (and you will), it's a lot of fun and that's what it's all about.


Anonymous said...

I'm guessing the data size of your audio will play into this too, such as whether your source audio is 16/24 or 32 bit?


nulldevice said...

Whether the EQ should go before or after the compressor is practically a religious issue. I've seen engineers froth at the mouth about that one.

Tom said...

@nulldevice - No doubt. I think the best solution is to do it both before and after. Before to 'trim the fat' and afterwards for any additional sculpting of the tone.

Tom said...

@Jon-Michael - Bit depth can have an effect on the sound quality of audio (this is more noticeable with acoustic material than it is electronic, generally), but it doesn't have a huge effect on the general timbre of effects