View Single Post
  #1  
Old 2005-12-16, 23:49
jappe jappe is offline
 
Join Date: Jul 2001
Posts: 54
Spectral Fingerprint and Spectral Broker module

The purpose of this module is two-fold:

1. Spectral Fingerprint mode: To visualize the Spectral characteristics of the sound on the input. This could actually be a Spectral Analyzer - but with the possibility to view spectral characteristics relative to other Spectral Fingerprint modules input.

Example:
I have 8 inputs to the main mixer. Before each mixer input, a Spectral Fingerprint module is inserted. The modules simply shows a colored icon. If any icon is similar in color, then I immediately see they collide frequency-wise. When I tweak an instrument in the rack, I can keep an eye on the Spectral Fingerprint modules in order to avoid frequency collision with the other channels.

The icons could be animated to reflect the relative timbre for each point in time as the sound progresses.


2. Spectral Broker mode: Lets all Spectral Broker modules (=Spectral Fingerprint module set to "broker" mode) automatically and dynamically trade frequency space with each other, based on a set of configurable rules.

Example:
My rack is made up of patches for bass, lead synth, chord, choir and vocals. The output from each of their corresponding mixer is fed in to the main mixer. There is one Spectral Broker module inserted before each of the five input channels.
I already know my bass, lead synth, chord, choir and vocals fit pretty well, since I've chosen & tweaked each patch so that the Spectral Fingerprint modules show me they are reasonably different spectral wise. But they would not be perfectly spectrally separated: I would either have to do some manual (static!) EQ:ing to make them fit "perfect" OR I could turn on the "broker" mode for each Spectral Fingerprint module - making them precision-EQ each channel based on "agreements" between the modules regarding which spectral frequency is to be allocated to which Spectral Broker module for each point in time.(dynamically EQ:ing!)

The rules for how the frequency-trading shall be performed can be configured in different ways by the user. A priority based rule could for example divide the frequencies at each point in time with highest priority for the Spectral Broker module that "needs" each frequency most at the time (i.e. the spectral frequency is a vital component in the spectral characteristics for that sound)

So for each point in time, only ONE of the Spectral Fingerprint modules would pass through a specific spectral frequency to the mixer - that frequency would be supressed at all other SF modules.

When I write "frequency" I actually mean a more or less narrow frequency band (filtering with one frequency bandwidth would cost lots of CPU% I guess)

Many SB-modules would probably utilize quite a few CPU%. Enabling the SB's when rendering could be an option then. (perhaps running them in CPU save mode in realtime, with lower quality EQ:ing)

Theres more to say about this but I think I'll spare you from having to read that for now...