Sorry if this is a dumb question. I am trying to understand a video tutorial on render nodes. At 16m 50s it is explained that when a mix node is set to subtract the order of the noodle inputs matters. But for multiply it does not matter - for mathematic reasons. See video below.
This makes sense to me. So, when I set it to “Add” I would expect the order not to matter. But in reality I get different results for some reason. See video below.
I don’t understand why this happens.
If you set factor higher than one it changes the result because the second input is scaled by the factor and thus the result heavily depends on the exact values of input2.
Having a mix factor above 1.0 is most usually not a thing to do whatever mix mode you use. Actually it is beyond me why it is even allowed in this node as it makes no practical sense.
For some blend modes (especially add, actually) it’s useful because it lets you scale up the value before comping it in. For something like a bright flare you can absolutely use fac=5. Or 10. Or 20.
The factor in the add node works like this:
input 1 + factor * input 2
works as expected
Readability wise I would personally prefer a separate multiply op before the mix as it makes it plain and clear that something is done with the element before the mix. Exposure adjustment in the mix operation can easily hide this and makes debugging the setup more tedious.
Yeah, but unfortunately the compositor refuses to have a single node that will increase exposure by more than 4x (the HSV node’s “value” slider really shouldn’t have a limit, or at least a much higher one). You have to either chain multiple color correct or HSV nodes, or use the separate/combine HSV nodes and stick a math node between them. So we’d have to fix that first. 