Requests with ISL!

[juan_irender]

Hi, Ono.
Yes, here it is.

Thanks!

[juan_irender]

Sorry, Ono .
The scene i posted to show this problem is this I put now!
Use this instead, please!

Thanks!

[OnoSendai]

Thanks for the test scene.
It's getting messed up due to the render setting 'merging' being on, which merges objects.
Open up the scene and set it to false in the render settings.

[juan_irender]

That was it.
Thanks Ono. In fact, in the export options of this C4D scene in Cindigo, in "render options", the "Merge UVs and Vertices" op was checked. I wasn´t aware of this until now, hehe.

Thanks again .

[juan_irender]

A new wood shader.
The cube has applied the Oak wood shader, devised years ago by Larry Gritz as a RMAN shader, and now adapted to ISL, with some tweaks added by my own.
It has over 15 params to play with .
This shader has another child known as "parquet plank shader", which uses the oak shader as a base to generate the wood patterns. But if this shader is a little monster in ISL, then the parket plank one will be a really big monster...

Ciao.

[OnoSendai]

The shader on the cube is amazing, can you post it?

[juan_irender]

The shader on the cube is amazing, can you post it?

Larry Gritz is a master coding shaders, his book "Advanced Renderman" is a must read.
Of course I´ll upload it to the shader gallery, I´m not sure if applied to the standard material sphere would look right.

Thanks.

[CTZn]

I know I'm going to sound mean; the shader does look cool, however, having seen the code I'm finding that it's a lot of fuzz for a little bang

Keep it going anyways ! I would have uploaded my current works if I was happy with it...

[CTZn]

I know I'm going to sound mean; the shader does look cool, however, having seen the code I'm finding that it's a lot of fuzz for a little bang

I'll be more specific: the main aspect is very similar to fract(fbm() * 10.0), and grain would be much better handled using voronoi.

However I'm glad that you brought to me the 3d orientation Juan, it's a must for 3d woods thanks

[OnoSendai]

The shader on the cube is amazing, can you post it?

Larry Gritz is a master coding shaders, his book "Advanced Renderman" is a must read.
Of course I´ll upload it to the shader gallery, I´m not sure if applied to the standard material sphere would look right.

Thanks.

Thanks.
Shader looks good, but it does run *very* slowly.

[juan_irender]

The shader on the cube is amazing, can you post it?

Larry Gritz is a master coding shaders, his book "Advanced Renderman" is a must read.
Of course I´ll upload it to the shader gallery, I´m not sure if applied to the standard material sphere would look right.

Thanks.

Thanks.
Shader looks good, but it does run *very* slowly.

Yes, CTZn, Ono, the shader isn´t optimized at all in respect to ISL. Probably the code can be comprised/optimized, it has a lot of code executing right now. You should understand that my main goal, right now, is to port some powerful RMAN shaders to ISL. First reward is for me, as an exercise for coding shaders in a functional language. Next is for the enrichment of the Indigo community .

Ciao.

[OnoSendai]

Some general comments on the oak shader:

* noise() in indigo is already in range [-1, 1]. So you don't need snoise().

* code like this

Code: Select all

sqrt(y(pringsum())*y(pringsum())+z(pringsum())*z(pringsum()))

is better to write as

Code: Select all

let 
  p = pringsum()
in
  sqrt(y(p)*y(p)+z(p)*z(p))

The first code *may* get optimised to the second, but it's best not to rely on the optimsiations and factor out the common calls yourself.

Code: Select all

pow(clamp(0.8-x, 0, 1), 2)	

Don't use pow(x, 2). use a square() function instead, as the pow(x, 2) code may actually execute the floating point pow function which is very slow.

EDIT: One more thing - it would be good to have proper 'max' values on the shader params.
Thanks for your work!

[juan_irender]

Some general comments on the oak shader:

* noise() in indigo is already in range [-1, 1]. So you don't need snoise().

* code like this

Code: Select all

sqrt(y(pringsum())*y(pringsum())+z(pringsum())*z(pringsum()))

is better to write as

Code: Select all

let 
  p = pringsum()
in
  sqrt(y(p)*y(p)+z(p)*z(p))

The first code *may* get optimised to the second, but it's best not to rely on the optimsiations and factor out the common calls yourself.

Code: Select all

pow(clamp(0.8-x, 0, 1), 2)	

Don't use pow(x, 2). use a square() function instead, as the pow(x, 2) code may actually execute the floating point pow function which is very slow.

Hi Ono, your comments are of much value for me!
I´ll revise the shader code.

Thanks.

[juan_irender]

EDIT: One more thing - it would be good to have proper 'max' values on the shader params.
Thanks for your work!

Without doubt. I´ll post a reference for the parameters role and its optimal values. Maybe I´ve upload this material early, but I think is useful to let the parameter values goes "free" in some degree, at least in the initial steps.

Thanks.

[CTZn]

This shader as an exercise is totally legitimate to me. You are free to update it anytime.

I'm discovering new stuff thanks to you.