Hello
I am often working with some pre recorded material with my memory palace and am looking to keep it in tact/same position but use the feedback process and was wondering if im missing something and there would be a way to keep the original video in place and only modify the feedback so that when you say rotate only the feedback path rotates and the image would remain where it was.
This would be similar to feedback when done with a TV and a camera when you are recording what comes out of the TV where if you have something displayed on the tv it stays in place and the feedback moves around it. like then the date or time shows up from your camera on screen or if you keyed an image in. the image never moves.
I was going to request something similar in a long post in the v19 thread. The ability to affect the video frame or the framebuffer separately would be really sweet.
Could you mix (or key, or multiply, etc) the original with the Memory Palace’s output? Seems like that could accomplish it, kind of like making your own dry/wet mix in audio with a full dry (the source) and full wet (the effected output) and a mixer.
the scene mode puts stuff in font of the original and only adds one copy at most unless im missing something. While it does have the doesn’t move the original it doesn’t have feedback and its compositing on top of the image instead of behind it.
Oh, I hear you. I’ve thought the same thing about changing the prioritization of the feedback to stay focused on the original, but haven’t tried to chase it down yet.
How about this: do you have a Visual Cortex? Video into the VC, mult the RGB from the video to one channel on the VC and the other to the MP ins. Send the luma from the video to a Doorway, then mult the key to fade between channels on the VC and also to the alpha in on the MP (set the params for alpha keying). For chroma key use a Polar Fringe between the VC and the MP.
It works OK–I just tried it–you could keep the original image and put it in the foreground over the feedback. The one thing that isn’t perfect is that there seems to be a small gap between the original and the first transformation.
@jevergreen thank i tried a similar flow but something was off or sure and it took up to much stuff that was fun to play with in the patch and i suppose i was thinking all in MP.
@Dr_Rek yeah it seams like a feature request for sure at first i thought maybe i dint understand all the modes
I would edit the tittle to feature request but iI can’t edit or is there a place for feature requests?
You know in a pinch you can use the luma from Cortex to drive the MP framebuffer as a separate layer. I took the RGB of a video input from Cortex straight into one of its channels, multed the luma into one MP input and also to the Cortex key input, then sent MP to the other VC channel and keyed them together. With this patch it’s possible freely transform the framebuffer layer separately from the video source. Of course, your framebuffer is then only monochrome or colorized, but it works fine if full color feedback isn’t necessary. Results are actually quite good!
