I’ve recently started to try blender composition features. I’ve already have some experience with After Effects. So basicaly I’m trying to mimic my AE way of working on blender.
Here it goes: I was trying to make a “stilish” (non-realistic) blood splatter fall on an old piece of paper.
1 - I took a splatter video. Then I made a blood texture. Then I took a paper image.
2 - I managed to get the blood texture to the splatter video (first image bellow).
3 - I was satisfied with the blood splatter comp. Now I simply needed to apply that comp in front of the paper image. I tryed to find a way to mix that comp with the paper image, without rendering it and using an “image sequence node” (to save some time and disk space). See second image below for what I wished.
4 - In AE I’d use a precomp. But couldn’t find a way of doing it on blender (maybe appending or linking from a different scene, or blend file? I’m not sure). So I took another direction and solved it as seen on the third image below.
Though I found a way, the question that stays is: “Is there anyway of making precomps on blender? Or something similar to it?”
Thanks for the help. But groups aren’t still what I’m looking for. They actually organize things and seem to be pretty usefull for presets.
What I’m looking for is a way of using the results that appear on the node “composite” (add > output > composite) as an input node.
If you look on the image below, the “node group” only lets me use the same inputs and outputs that I already had.
So far I realized that blender renders what is in the “composite node”, so it is generally the last node of all.
So if I want to use the results of the “composite node”, I have to render the results. Then I have to add an “image sequence node” and load the sequence I’ve just rendered and then keep on compositing.
I want to skip the rendering and loading the sequence part. I’m looking for a way to use the results of the “composite node” directly in my composite (look at the second image on my first post - its a fake screengrab I did to show my idea).
I’m not sure if i understand your problem. Because your image “what i ended up with” is already the solution. You don’t need to pre-comp or pre-render anything. You just have different “node-flows”, like in this image:
So basically, instead of using the composite node as an input, you just plug that what you would have plugged into the composite node into a mix- or alpha-over-node, and then continue with that new “node-flow”.
If things get too crowded or confusing, then you can use groups.
if you want to save just the blood splatter as a video, you need to save it as a PNG sequence with RGBA and Premul enabled. Route it to a File Output node, and blender will save the snapshot in a file for each frame. Using File Out(s) in conjunction with Comp Out allow you to save multiple images per frame.
Do not include a composite output node in a group. Bad things happen if you accidentally have multiple comp out nodes in a noodle.
As sebastian_k says, just pipe the image of your textured splatter to a new node or nodes “downstream” and use the Composite node only for final output. To view intermediate “pre-comps” you can use the Output>Viewer node, which will display the node it’s attached to in a UV/Image Editor window set for Viewer Node. You can preview many kinds of composite elements this way, including Alpha channels and various Matte types, as well as progressive composite results.
Viewer nodes add some time to node processing, and it’s easy to have them accumulate in complex node work, so try to be Spartan in their use.
PapaSmurf’s suggestion is the way to get the “pre-comp” stages into files if needed.
I see your problem. With your setup, you might need a second AlphaOver-Node to get the desired output of the group. Or even more simple: When you mix the key and the texture, try to switch the inputs:
