a MULT is a MULT?

[terp]

Hi all.

I'm coming from Nuke, where a merge:multiply node spits out the same result regardless of A and B inputs.

This doesn't seem to be the case for TG.  I've looked at the documentation, but am still a bit puzzled.
Can someone help me understand why the output of a merge:multiply node would be different depending on which input is chosen?
(it has something to do with diffuse color?)

I'm playing with a voronoi setup from another thread and noticed a power fractal set to gray, being masked out.
As a thought experiment, I thought I could use a gray shader and merge:multiply to achieve the same result.
It wasn't the same depending on which input I chose in the pulldown of the merge node.

If I used a blue node (constant scalar 0.5), the result worked regardless of input chosen.

Another puzzle:
I differenced the two nodes trees (masked gray power fractal vs. my gray constant mults) and cranked the values.
There is indeed a difference, no matter what I choose?   Also strange to me.

Help?

Thanks,
Terp

[Matt]

Check the "Mix to A" parameter on the "Mix control" tab. It defaults to 0.5 because that's a good default in Mix mode, but should be 1 for many of the other modes. Unfortunately the default is 0.5.

[Matt]

If the only information you care about from the shaders you're multiplying is colours or scalars, consider using Multiply Colour or Multiply Scalar nodes. They are also slightly faster to compute, bypassing some overhead incurred by shaders.

[WAS]

I was going to mention the same thing about using just function based nodes. Also, gotta say, awesome use of the mixing for the PF to Voronoi.

[pixelpusher636]

Hi all.

I'm coming from Nuke, where a merge:multiply node spits out the same result regardless of A and B inputs.

This doesn't seem to be the case for TG.  I've looked at the documentation, but am still a bit puzzled.
Can someone help me understand why the output of a merge:multiply node would be different depending on which input is chosen?
(it has something to do with diffuse color?)

I'm playing with a voronoi setup from another thread and noticed a power fractal set to gray, being masked out.
As a thought experiment, I thought I could use a gray shader and merge:multiply to achieve the same result.
It wasn't the same depending on which input I chose in the pulldown of the merge node.

If I used a blue node (constant scalar 0.5), the result worked regardless of input chosen.

Another puzzle:
I differenced the two nodes trees (masked gray power fractal vs. my gray constant mults) and cranked the values.
There is indeed a difference, no matter what I choose?  Also strange to me.

Help?

Thanks,
Terp

Hope this all worked out for you. Good to see another Mac user on the forums. 

[terp]

Thanks for all of your replies!

Matt: The slider was at 1.0, yeah (feature request? Mix control + Merge modes all on the same panel? ).  I hear you about the calculations, which is why I started testing other options out.  When studying this [borrowed] node tree, I was confused why the all-grey-PF was used.  Maybe he wanted the mask functionality(?) which I thought I might be able to make simpler/less-calculation-heavy.

This is my first year in TG (been in Nuke for 15yrs).  I'm trying to understand how data flows down the TG node tree, how to keep track of values (is that PF black really 0.0 or is it -1.0?  -10.43?).  There's a lot of differences to know/experiment with. I'll keep looking for ways to use the function nodes with color/scalar info.

So, for my curiosity, if I change the Merge Shader Multiply to use A's diffuse color vs. Input's diffuse color there is a different result.  Is there some [shader?] data being passed down the tree?  I don't understand this yet.

WAS: I'm studying this setup (taken from a Voronoi rock thread a few years back).

PixelPusher: I'm working interactively on a 10-core iMacPro, yes!  (Thus, trying to simply calculation in node trees. )
I'm hoping to see how TG performs on a Mac Studio later this year(!)

Thanks!
Terp

[WAS]

Cool idea with the PF stuff. This also reminded me of my voronoi maps exported from Blender, so I gave one a try with this. Here is the project for anyone interested. Blender doesn't seem to make a perfect export, but it's good enough.

https://drive.google.com/file/d/1nW3DkEePSxZlWWvXb4TcEVa41Epcrngf/view?usp=sharing