Metallic paint

Hi, I’m back after yeeeeears.

Everybody loves metallic paint, isn’t so?
Spoiler: it’s PBR and two tricks.

Paolo

hi res

metallic_rgb.jpg
 
Last edited:
Metallic paint instruction

To get a good metallic paint you need:
  • An inner object with the PBR material applied
  • An outer shell of Membrane slightly larger than the inner object
  • An outer outer shell of Glass slightly larger than the Membrane shell
See the .jas file for a better understanding and use the included PBR material, Membrane and Glass.

Scale the texture image in the PBR material to obtain the desired metallic particles size.

Some colors need the Emissive property to be multiplied others need to be reduced, use the existing nodes accordingly.

You may need this texture

and this too
 

Attachments

  • Metallic paint.zip
    52.7 KB · Views: 162
First I'd like to say welcome back, and thanks for sharing,
you obviously worked hard to get a great metallic paint material.

I also like to delve into the Node system, but I often don't
have a clue what will happen, especially with the Math Nodes.

For me it's mostly trial and error, with a lot of errors.
I found if I go back and unplug one thing at a time and
test render, many things don't have any effect on the final render.

You're setup is brilliant with the interaction between the outer shell's
Index of Refraction set to 4.0, and the inner ball's Directional Node.

I did work out a very streamlined procedural version that
shows the real gem you found with the Directional Node.
I still don't know how it works, but it does work pretty well.

One unfortunate problem with the Directional Node I
found when trying to animate rotation, the dark side
with the Noise/Image nodes rotates as well.

Note: the jas file is slightly too large for some reason...

Metalic Paint test.jpg


metalpaintmat.gif
 

Attachments

  • Metal Paint Mats.zip
    101.6 KB · Views: 144
First try, but my Alpha attempt can be simplified A LOT.

The most significant tactic is that it uses a single Noise Node modified by a Gradient Filter Node to get a few really bright sparkles and a lot more dim ones. These are masked by multiplying the Noise output with a State Node -> I-N value ( I-N gets its value by calculating which rendering vectors are most directly pointed at the camera as a 1.0 value ).

Right now, it relies more heavily on the PBR Nodes's Emissive value's colors than I like. It mixes the Base Color attribute using the State Node's Position value with the Directional Node's Colors. I still haven't found good methods of manipulating the Specular highlights using Falcon and PBR Nodes as effectively as you can manage them under the Cheetah renderer with Material Nodes.

I'm not sure how soon I'll get back to this, but I will post my version of the Material when I'm happy with it.

Metallic Paint alpha.jpg
 
First I'd like to say welcome back, and thanks for sharing,
you obviously worked hard to get a great metallic paint material.

I also like to delve into the Node system, but I often don't
have a clue what will happen, especially with the Math Nodes.

For me it's mostly trial and error, with a lot of errors.
I found if I go back and unplug one thing at a time and
test render, many things don't have any effect on the final render.

You're setup is brilliant with the interaction between the outer shell's
Index of Refraction set to 4.0, and the inner ball's Directional Node.

I did work out a very streamlined procedural version that
shows the real gem you found with the Directional Node.
I still don't know how it works, but it does work pretty well.

One unfortunate problem with the Directional Node I
found when trying to animate rotation, the dark side
with the Noise/Image nodes rotates as well.

Note: the jas file is slightly too large for some reason...

View attachment 38953

View attachment 38954

Thanks ZooHead, your node is actually much cleaner than mine.
I'm currently using it to do some experiments, like adding some Emissive property to make the brightest particles glow more.

I must confess that I'm not fully aware of what I'm doing when messing with the Node system.
However I really like what serendipity can create.
 
First I'd like to say welcome back, and thanks for sharing,
you obviously worked hard to get a great metallic paint material.

I also like to delve into the Node system, but I often don't
have a clue what will happen, especially with the Math Nodes.

For me it's mostly trial and error, with a lot of errors.
I found if I go back and unplug one thing at a time and
test render, many things don't have any effect on the final render.

You're setup is brilliant with the interaction between the outer shell's
Index of Refraction set to 4.0, and the inner ball's Directional Node.

I did work out a very streamlined procedural version that
shows the real gem you found with the Directional Node.
I still don't know how it works, but it does work pretty well.

One unfortunate problem with the Directional Node I
found when trying to animate rotation, the dark side
with the Noise/Image nodes rotates as well.

Note: the jas file is slightly too large for some reason...

View attachment 38953

View attachment 38954

Thanks ZooHead, your node is actually much cleaner than mine.
I'm currently using it to do some experiments, like adding some Emissive property to make the brightest particles glow more.

I must confess that I'm not fully aware of what I'm doing when messing with the Node system.
However I really like what serendipity can create.
First try, but my Alpha attempt can be simplified A LOT.

The most significant tactic is that it uses a single Noise Node modified by a Gradient Filter Node to get a few really bright sparkles and a lot more dim ones. These are masked by multiplying the Noise output with a State Node -> I-N value ( I-N gets its value by calculating which rendering vectors are most directly pointed at the camera as a 1.0 value ).

Right now, it relies more heavily on the PBR Nodes's Emissive value's colors than I like. It mixes the Base Color attribute using the State Node's Position value with the Directional Node's Colors. I still haven't found good methods of manipulating the Specular highlights using Falcon and PBR Nodes as effectively as you can manage them under the Cheetah renderer with Material Nodes.

I'm not sure how soon I'll get back to this, but I will post my version of the Material when I'm happy with it.

View attachment 38956

Hi MonkeyT, can’t wait to see your Node system since I have already understood that it contains useful ideas.
 
I must confess that I'm not fully aware of what I'm doing when messing with the Node system.
However I really like what serendipity can create.

I'm in the same boat, but luckily MonkeyT has a greater understanding of the Node system. ;)


I'm not sure how soon I'll get back to this, but I will post my version of the Material when I'm happy with it.

View attachment 38956
Nice! Can you rotate the ball without it going "dark".
 
Same old 'not so clean technique'

Based on image texture:
Cannot build with nodes the same random distribution and blur
Spark sample1.jpg
Spark sample2.jpg


Shell in a shell in a shell object:
Cannot merge the emissive colors dots without disrupting them (tried everything)
Schermata 2022-12-06 alle 08.36.07.png


Directional or I-N not involved yet:
Quite sure I should
 

Attachments

  • Sparkling guitar.zip
    394.6 KB · Views: 144
  • Spark sample2.jpg
    Spark sample2.jpg
    21 KB · Views: 150
Shell in a shell in a shell object:
Cannot merge the emissive colors dots without disrupting them (tried everything)
View attachment 38972

Directional or I-N not involved yet:
Quite sure I should
You are so very close to your answer.

Single Material:
The Sphere on the left is your multi-object technique. The one on the right is a single Node Material.

Sparkle_12-6.png

Emission works by being additive. If the value is 0 there is no change, the higher the value is, the brighter values are added. The patterns you creating are mostly black: so ADD them together. Black will stay black. Colors added to black will be the colors you calculated.

BTW, your gradient usage is beautiful.

Specular level is much the same (0 to 1 values), so build a map using I•N and MultiplyAdd. t = the multiplication value, -1 in this case, so the highest specular value will be on the edges after you multiply by -1, then add the B value to get the straight on Specular look to appear as you want it. This invert the I•N values, going from Black edges to White edges: {0,1} inverts to {-1,0} then adding shifts the values back up the normal scale {0,1}. The values in this example still needs a little work to match the original technique, to dim down the outer edges.
SparkleNodes_12-6.png


Really good work. Thanks for posting the material!
 

Attachments

  • SparkleNode_12-6.zip
    52.6 KB · Views: 143
MonkeyT… I learn so much from your node work insights. Little by little I am getting more predicable results with my experiments.

Thanks so much for your postings, you are a valuable resource!

Cheers,
gsb
 
Back
Top