You can also think of this as like the Luma output from the Visual Cortex Pre-Encoder Expander but you are able to use it at any point in your patch–not just after Cortex’s compositor stage. Anywhere you resolve an intentional RGB image or have 3 distinct signals where you want to use the perceived luminance/amplitude of a RGB image to feedback/influence another parameter of your patch, this is a direct way to do it + an inverted output. It doesn’t dead-end your patch to do feedback with RGB mixes anymore nor does it require a triple mult.
In the context of using it as the key input for compositing two separate images, you usually want to feed the Y output into a soft or hard keyer module and its output then goes into the VC control of a triple crossfader that fades between the RGB passthrough of LUMA and another RGB image. Consider these rough patch flows:
3x Cadet VCOs → SMX-3 for colorization → LUMA RGB Passthru-> SHUTTER Ch.A
Mix Cadet Ramp outs → SHUTTER Ch. B
LUMA Y → SHUTTER R Switch Input
SHUTTER Out → Encoder
3x Cadet VCOs → SMX-3 for colorization → LUMA RGB Passthru-> Marble Index Ch.A
Mix Cadet Ramp outs → Marble Index Ch. B
LUMA Y → Doorway Source → Marble Index Opacity CV (b/w Ch. A&B)
Marble Index Out → Encoder
The keying effect, instead of being like a screen wipe fed by a H or V ramp, will follow the brightness of the SMX-3’s output into LUMA. You get a much more intuitive luma key that more tightly or loosely fills in darker/lighter areas of the first RGB image with another RGB layer.
I’ve been putting it right before Cadet II and feeding it back into a BSO Crossfader that has the passthrough of LUMA 's RGB inputs and nothing into ch. B for black. Then into Cadet II encoder. Basically using it as a fine-tune VCA so negative space gets defined well.