Just follow up a discussion that was initiated by Millolab in another thread
Please find below a simple convolutional filter, where you can achieve a bokeh from a self-drawn image.

Input Channel0: Connection for Video
Input Channel1: Connection for Bokeh Image
Input Channel2: (Optional) connection for B&W depth map

I did limit the Bokeh Image to 100x100 pixels. Image is multiplied and added to the Video in a typical convolutional way. But I did set 50% grey to zero influence. Hence below...

This time I converted the following laplacian noise shader into a DCTL Fuse. Would be great to get some feedback if everything works fine.
td_LaplacianFlowNoise.fuse

I try to convert the following Shadertoy into a DCTL Fuse.

Works so far so good as long as you use the resulting image as a still image. As soon as you move it, it starts to flicker and the whole image appears noisy with fine grain.
Maybe there is some trick as I had to deal with beloved textureGrad, dFdx and dFdy code. Would be happy if somebody finds a workaround.
The following video should show the issue - first one is the original image.

td_EndlessTexture.fuse

Please Note: A...

The other day I stumbled upon this:

and found this on Nukepedia.

Shared the whole thing with good friend danell and while he's investigating into making a Plugin or a Fuse out of this, I managed to make it into a macro.

ml_RefineMatte.setting

If someone wants to join in the development of this, you are more than welcome! :)

p.s. do not expect miracles. It has its own limits... :) :)

Flow Manager

After some rethinking I re-did the UI of my scrip Flow Manager and thus wanting to post it here before pusing it to Reactor!

Tag nodes as FullQ or Proxy. FullQ should be your main node (eg a slow depth blur) and Proxy should be your substitute node (eg a fast gausian blur).
By switching between these you can now speed up your flow by temporarly disable heavy nodes and replace them with light once.
Before rendering out, with a press of a button, switch back to the heavy but...

I used this shadertoy in my YT video and it was requested to release it as a fuse. So I built a fuse from the FragmentShader file. The parameters are not limited and may need to be adjusted.

I have only one Nvidia system and have tested this fuse under Cuda and OpenCL (Nvidia).

Have fun

ProteanClouds.fuse

So, my second attempt to build up this fuse was successful. The first time, I couldn't find the reason why I received strange outputs. This time I was better organized with my template. But the mod function seems to work differently compared to WebGL for Shadertoy which gave me a headache.

td_CampFire.fuse

There are already some implementations concerning Voronoi Noise. I thought I should try to create and modify one which includes diverse options to play with.
Hope you have some fun...

Voronoise_Extended.fuse

Please note, that the compression artifact from YouTube will not display the fine grainy noise correctly.

The following fuse is a modified version of the Fiery Waves Shadertoy:

Would be great to get feedback about compatibility with different graphic cards.
Fire_Wire.fuse

Due to a recent discussion about seamless texture for 3D objects I made a DCTL fuse, which transforms different 2D textures into seamless (non-mirrored) textures.
You could also make a macro, but the DCTL fuse should be much faster. The idea is simple - you copy a texture image four times and put their midpoints to each corner of the whole image. You increase their sizes equally and blend the overlapping parts into each other. You can see the results in the following video.

Hope the DCTL fuse...