5/2/2011 4:16 AM
I’ve seen references to “convolution reverb” and “algorithmic reverb.” What’s the difference between them? Is one type better than the other? How do you know which kind is best for a particular project?
JOE PASQUALE, HACKENSACK, NJ
Thanks for your question, Joe!
Convolution and algorithmic reverbs use different technologies. A convolution reverb gets its sound by loading an impulse, which is a sampled “snapshot” of a room’s decay characteristics. One way this is performed is by recording a starter pistol shot and decay. Loading this impulse into a convolution reverb tells it how to process an incoming signal so that it decays in the same way as the impulse. Algorithmic reverb synthesizes a room’s characteristics by creating algorithms that generate a particular number of reflections, subject them to high-frequency damping, add predelay, simulate a particular room size, and so on. Although this is an “imaginary” reverb, the algorithms are based on analysis of the way acoustic spaces affect sound. You can think of convolution reverb as similar to keyboard samplers, while algorithmic reverb is more like virtual analog synthesizers. As for which type to choose, convolution reverb is more “literal,” while algorithmic reverb is more “impressionistic.” Convolution reverb is more flexible in that you can load different impulses and obtain entirely different sounds—even impulses of tunnels, or something like the body of an acoustic bass. However, it’s less flexible in other ways because changing individual parameters—decay time, damping, and so on—is difficult, or in some cases, impossible. Algorithmic reverb limits you to the chosen algorithm, but you have far more flexibility to edit the algorithm’s parameters. Each type of reverb has its own sound quality, and neither one is “wrong” unless it doesn’t sound right in the context of your music. Try different types of reverb, and it should be obvious which works best with your recordings.