Skip to main content
W
Wackyal123
Participant
September 22, 2023
Question

Can I use a gradient to drive both X/Y amount and scale in TileSampler?

  • September 22, 2023
  • 1 reply
  • 240 views

Hi,

I'm currently trying to procedurally texture a creature with scales. I have a mask which I've then blurred to create a nice gradient. This controls the scale of the tiles (which ironically... are scales!)

I then use a distance node, followed by edge detect to give me a decent scale appearance. 

However, I really need the gradient to also control the number of scales so that as the scales get smaller, they increase in volume. 

Is there a method I can use to do this, or a simple function which would give me access to drive the X/Y amount using the same gradient as the tile scale?

Thanks!

This topic has been closed for replies.

1 reply

Cyril Dellenbach
Community Manager
Cyril DellenbachCommunity Manager
Community Manager
September 26, 2023

Hello @Wackyal123,

 

Thanks for the question.

 

Interesting subject, but complicated to obtain a clean result. Back in 2021, Luca gave an answer to a similar thread with a workflow that might interest you.

 

Another solution would be to use multiple Tile Sampler and merge the results together, but this will greately impact the performances of the graph.

 

Best regards,

 

Cyril Dellenbach (Micro) | QA Support Artist | Adobe
W
Wackyal123Author
Participant
September 26, 2023

Hi Cyril, 

Thanks for your reply. I tried the aformentioned suggestion from Luca, but sadly it's too random for what I'm trying to do where I really need a very accurate link between the size and the scale. But thanks for the speedy reply! Very much appreciated.

For anyone else who might see this post whilst searching for a similar thing, I have managed to create multiple tile samplers with different scaled tiles, and I'm using a custom input pattern (an edited diamond shape) then using a mask per tile sample node to drive the "pattern distribution" so that there isn't as much overlap between patterns. It's not ideal as the gradient isn't as gradual and it's quite a heavy graph, but it does "kind of" work. 

Cheers! 

    Terms & ConditionsAccessibility statement
    Using the community Terms of use Privacy Cookie preferences Do not sell or share my personal information ad choicesAdChoices