The Blend Image with Feedback Node is missing a few parameters, I’m proposing a second node!!!
An absolutely essential parameter for feedback in my opinion is hue shift.
Also a “framerate” is very useful. So that not every frame is filled up with feedback, which allows you to create graphs etc.
There are also a bunch of parameters on “Blend Image with Feedback” that aren’t very useful and can comfortably be scrapped to save space. Transforming the foreground image is redundant, and can easily be done with another node, similarly: width, height, colour depth etc.
An improved feedback would be soooo useful in my opinion.
The original method of creating image feedback is both the clean and advanced one ;). Blend Image with Feedback is mostly a simplified version that does the basic trails and scaling. As color depth directly affects smoothness, the breaking point of feedback, and isn’t easily set elsewhere without a blend node - it is absolutely important.
When that is said, the basic “hard” approach for creating more advanced feedback can look something like this:
This enables other displacements of the feedback image, here I’m using “Bulge Image”, but any translation will do to offset the image to enable the feedback. This is also the approach you want to use for diffusion/reaction kinds of effects as you can blur/sharpen with each frame iteration.
As far as framerates goes, this should also be a viable approach. You might also have to hold the original image value to sync it up, but that depends on your composition.
Magnesson - thanks for the composition - my first attempt at “hard” feedback is here: https://community.vuo.org/t/-/6437 - subsequent to that I have made others which are probably better. I think you have the image routed into the FOREGROUND of the blend image node - therefore unsuitable for some applications - ie feedback on the alpha channel behind an image.
What I would love is a simple 2d transform image node! Scaling the feedback doesn’t seem possible with “resize image” so to acomplish this I did the feedback with layers.
Looking good! The Hold-node approach also enables feedback of everything, not just images. In theory you don’t need anything but an image layer, and then you can feed that back to itself. I think I’ve done this before, but trying it now it seems that it breaks the composition.
You can however use an image layer or a 3D square instead of a blend node to achieve more or less the same results with fewer nodes. This also enable multiple feedback paths at the same time if that is desired. Fractal-like patterns may emerge!
