Gearslutz.com - View Single Post - Lexicon reverbs: a brief bestiary
View Single Post
Old 8th September 2012   #280
Lives for gear
 
Michael Carnes's Avatar
 
Joined: Mar 2008
Location: Salt Lake Valley
Posts: 723

Thread Starter
Quote:
Originally Posted by seancostello View Post
224, 224X, and 224XL have the Constant Density Plates. The other algorithms all have density that increases with time, although this can be controlled by the DEFINITION parameter in some algorithms.
That's an interesting topic, but not really for the reasons we think. I'm sure Sean is more than a little aware of this, but even the non-constant-density reverbs hit a brick wall in terms of the number of reflections. The Nyquist theorem is in full force. At 44.1, you quite simply can't have more than 22K reflections in a second, no matter how many allpasses or how long the tail.

The group delay that we moan about (for good reasons--it is the cause of many reverbs sounding metallic) also brings along some real benefits. Even though the maximum density of the reverb is reached pretty quickly, the pseudo-reflections line up against one another in different patterns as the sound ages. Although the tail appears to have a pretty regular spectral signature--typically the highs die away faster--in a local sense the frequency response is much more irregular. It's not truly chaotic, but it has a certain chaotic sense.

In the same way that the frequency response is irregular in a moment-to-moment sense, the phase of various frequencies relative to one another is also changing rapidly (if they're going through a typical network of long allpasses). This is a very helpful feature in maintaining mono/stereo compatibility, and also is more natural-sounding as long as group delay issues are managed well.

So in a certain very real sense, all reverbs are constant density after a couple-hundred milliseconds or so.
Michael Carnes is offline  
Reply With Quote