Op 20/04/2021 om 20:06 schreef Bald Eagle:
> Thomas de Groot <tho### [at] degrootorg> wrote:
>> to use the albedo keyword however,
>> has been abandoned for the time being as too difficult apparently to use
>> in any way.
> 
> Interested in hearing why - it's always good to document perceived failures as
> much as successes.
> 
In a nutshell (I want to investigate more of course) I fail to tweak the 
different values (following the rules) without drastically changing the 
colour/aspect of the texture too. Something fundamental escapes me.

>> In a next stage, the use of more elaborate macros are planned, including
>> sslt possibilities for instance.
> 
> I simply figured that with something that had transparency like quartz, it might
> be a good thing to include.
> 
Yes, and it certainly needs more work.

>> Bald Eagle has proposed a first version
>> of such a macro and I am going to study it and see how to integrate
>> things.
> 
> As always, let me know if there are any questions.   The code was mostly to
> demonstrate the rudimentary principles and stimulate further discussion and
> development.
> 
Of course.

> Now that some of this has percolated in my head for a bit, I think I see
> Maurice's concern more clearly - the distribution of the texture elements are
> too uniform.  Yes, there are different grain sizes and colors, but there is no
> layering, veins, or variation of scale throughout the texture.
> 
Yes. I tackled the grainsize differences but I am still not happy with 
the aspect overall.

> Should we have a color_map .map file, or should we have a texture or material
> map for something as complex as a granite?   Different grains are different
> minerals, with different colors, iors, sslt attributes, finishes, etc.  Maybe
> make materials for each individual mineral and then combine them into the final
> granite.  More than simplistic, less than explosively complex.   Put some
> comments in where we're leaving excessively complicated things out of the stock
> new granites, and leave those as avenues to pursue for those wishing to craft
> more complex materials.
> Part of what I'm thinking is to be able to select from a list of "complete"
> materials and use them combinatorially, like in Jonathan Hunt's "Pebbles.pov"
> 
Again, yes. I am now exclusively working on the Mohogany granite and I 
want to explore those avenues you suggest. I realise that I have taken 
on a major investigation :-) and that is ok. I am not bound to a 
deadline. I certainly would like to come up with a credible granite 
material and I am beginning to get "obsessed" by it as it were. Slow 
work of trial and error... discussing the issues helps to focus on the 
weak and strong points, and eliminate dead-end issues.

> Going even further along that line of include file development, it would be
> beneficial to create an array of default versions of all the macro textures so
> that the user could address and select them _numerically_.  A macro in the
> include file could be called to instantiate such an array if desired, so that
> it's not generated each time the include file is called.
> 
yes indeed.

> I'm thinking that the variation of the granite will probably either have to be
> controlled / achieved with layering, or by elevating the level of the underlying
> "granite" pattern with a pattern map or function or some other approach.  Here
> at the outset, it's not so crucial to actually achieve the goal, but to define
> it, so that we can see what theoretically should be done, and then decide what
> compromises and tricks can be used to approach that without undue complexity.
> Set out some must-have attributes for what, geologically, a proper granite
> should embody, and then at least we have something to shoot for.
> 
following what was written higher on. The "depth" aspect of the rock is 
crucial but also difficult to achieve.

> I also think that the white that got added into some of the granite texture
> examples seems "off".   Like they are clouds and are floating over and apart
> from the underlying stone and are not a part of it.  Not sure what should be
> done to address that, or if you see it as a problem.
> 
Imo, the white represents the quartz veins that cut through the 
granites. In the original code they came out as mere "clouds" of grey; I 
brought them back at least as something looking like veins (the top 
texture). See the examples I provided.

> 
> Anyway, that's what was bouncing around in my head.
> 
Quite a lot to have bouncing around in there, sir. ;-)

Your comments are much appreciated and keep me on track.

-- 
Thomas

Post a reply to this message