This is what sslt looks like in the Granite_21 macro (beta 1.7). Works 
well and/but slow (took >5 hours).

However.

A granite would not look like this. This is more like a granite-textured 
kind of pottery. What is missing is the internal structure of minerals, 
especially in the ears. I am pondering the problem and think into the 
direction of a media, or maybe, a special kind of isosurface (I am 
thinking of my rings of Saturn in 2017), or again, a rand_in_object fill 
with small particles.

-- 
Thomas

Post a reply to this message

Attachments:
Download 'granite_21_sslt_test.jpg' (43 KB)

Preview of image 'granite_21_sslt_test.jpg'

Thomas de Groot <tho### [at] degrootorg> wrote:
> A granite would not look like this. This is more like a granite-textured
> kind of pottery. What is missing is the internal structure of minerals,
> especially in the ears. I am pondering the problem and think into the
> direction of a media, or maybe, a special kind of isosurface (I am
> thinking of my rings of Saturn in 2017), or again, a rand_in_object fill
> with small particles.


Wow this looks extremely promising. I think you are on the verge of a major
accomplishment!

I visited a local granite supply store again recently to look more carefully at
real granite. If I may make a humble suggestion relative to the current phase of
your project based on what I saw it would be this: The samples that I observed
that demonstrated SSLT only seemed to have SSLT properties in the veins, or
perhaps certain types of crystals but not others. The same was true of the
"sparkly" and/or iridescent elements. Perchance that condition was unique to the
granite samples that the business happened to have in stock at that time. I have
no idea whether or not a larger set of samples would reveal that that is a
universal rule.

Also, I noticed that some granite is like the Dakota Red as far as scale, for
example a 1" x 1" piece is all you need in order to know what the pattern looks
like. There were some samples with a much larger scale (although the individual
crystals were probably similar in size) and by that I mean that you would need a
much larger piece in order to get an idea of what the pattern looked like. In
some cases a 3- or 4-square-foot or even larger piece would be necessary.

I promise I'm not trying to complicate your project, I'm just mentioning what I
saw.


Kind regards,
Dave Blandston
Suggested motto: "With POV-Ray anything is possible, but nothing is easy"

Post a reply to this message

Op 4-7-2021 om 10:41 schreef Dave Blandston:
> Thomas de Groot <tho### [at] degrootorg> wrote:
>> A granite would not look like this. This is more like a granite-textured
>> kind of pottery. What is missing is the internal structure of minerals,
>> especially in the ears. I am pondering the problem and think into the
>> direction of a media, or maybe, a special kind of isosurface (I am
>> thinking of my rings of Saturn in 2017), or again, a rand_in_object fill
>> with small particles.
> 
> 
> Wow this looks extremely promising. I think you are on the verge of a major
> accomplishment!
> 
He he! We are getting there, slowly indeed. ;-)

> I visited a local granite supply store again recently to look more carefully at
> real granite. If I may make a humble suggestion relative to the current phase of
> your project based on what I saw it would be this: The samples that I observed
> that demonstrated SSLT only seemed to have SSLT properties in the veins, or
> perhaps certain types of crystals but not others. The same was true of the
> "sparkly" and/or iridescent elements. Perchance that condition was unique to the
> granite samples that the business happened to have in stock at that time. I have
> no idea whether or not a larger set of samples would reveal that that is a
> universal rule >
Yes, you are absolutely right. The point is indeed that only some of the 
minerals composing the rocks are transparent/-lucent. Quartz and 
calcite, which form most of the veins, but also some of the more 
colourful minerals of the plagioclase class (the pink or grey ones). 
Others are perfectly opaque. So, getting this in a (3D) sslt scheme is a 
challenge indeed. I have been experimenting with additional filter 
values in the vein's colour maps but that is difficult and highly 
controversial at the moment.

The 'larvikite' which you showed in another post, the one with those 
iridescent minerals, is a special case of 'monzonite' 
https://en.wikipedia.org/wiki/Larvikite . It resides at the fringe of 
the granite field. It would be a challenge in its own right of course. 
Sam Benge has done some perfect modelling of jaspers which would come 
close, but I would prefer to concentrate on the 'classic' granites which 
are already given enough trouble by themselves ;-).

I can tell that, at the moment, I have five different granites ready, 
based on Daniel Mecklenburg's original colour maps: Dakota Red, North 

In due time, I shall post them. My philosophy at the moment is that - as 
those originals are commercial names and products, possibly not even 
'real' granites but also gneisses or other metamorphic rocks - I 
transform them to my own understanding into my own 'real' granites, 
keeping the name, but differing quite possibly from the original stuff. 
With a little utility provided by Bald Eagle, this has been made much 
easier now and will be made much easier also for the users themselves 
who want to design their own granites (which is all the fun of course).

> Also, I noticed that some granite is like the Dakota Red as far as scale, for
> example a 1" x 1" piece is all you need in order to know what the pattern looks
> like. There were some samples with a much larger scale (although the individual
> crystals were probably similar in size) and by that I mean that you would need a
> much larger piece in order to get an idea of what the pattern looked like. In
> some cases a 3- or 4-square-foot or even larger piece would be necessary.
> 
Yes. I /think/ that the user will be able to play with those pattern 
changes and pattern scales, using the provided pigment_pattern mask 
arrays, and variations arrays. They are very powerful and subtle means 
to model the granite one wants. In fact, each granite can be modelled 
into almost an infinity of different variations with only very little 
changes to the arrays. I need to expand part of the future documentation 
to cover this particularly fascinating aspect of the macro.

I am not finished to be surprised by my own macro indeed. ;-)

> I promise I'm not trying to complicate your project, I'm just mentioning what I
> saw.
> 
And rightfully thanked you are. These comments help to hone my own 
understanding of the whole process indeed.

-- 
Thomas

Post a reply to this message

Thomas de Groot <tho### [at] degrootorg> wrote:
> So, getting this in a (3D) sslt scheme is a
> challenge indeed. I have been experimenting with additional filter
> values in the vein's colour maps but that is difficult and highly
> controversial at the moment.


What you and Bald Eagle have accomplished so far is already awesome so anything
beyond that will be extra-awesome!


Kind regards,
Dave Blandston
Suggested motto: "With POV-Ray anything is possible, but nothing is easy"

Post a reply to this message

Op 05/07/2021 om 04:13 schreef Dave Blandston:
> Thomas de Groot <tho### [at] degrootorg> wrote:
>> So, getting this in a (3D) sslt scheme is a
>> challenge indeed. I have been experimenting with additional filter
>> values in the vein's colour maps but that is difficult and highly
>> controversial at the moment.
> 
> 
> What you and Bald Eagle have accomplished so far is already awesome so anything
> beyond that will be extra-awesome!
> 

<grin> we are certainly getting there, and try to sprinkle a bit of 
extra-awesome over the whole if possible.

-- 
Thomas

Post a reply to this message

Thomas de Groot <tho### [at] degrootorg> wrote:
> What is missing is the internal structure of minerals,
> especially in the ears. I am pondering the problem

Yeah - it's the old "POV-Ray renders _surfaces_ not volumes" problem/trap that i
fall into from time to time.

> and think into the direction of a media,

I think that might lead to an unintentional detour: the pursuit of a PhD in
media

> or maybe, a special kind of isosurface (I am thinking of my rings of Saturn in
2017),

That has potential, as if it were done right, you'd have a lot of separate
surfaces, but then you'd have spaces in between - no idea if that would be
problematic or not.  You'd also have to match up the texture with the isosurface
domains.

> or again, a rand_in_object fill with small particles.

That would be super tricky and take for-ever.
Exposure of the set of points for crackle to SDL, allowing for "one-pass"
placement of "particles" would be what we'd need, and I can still see that as
being "slow" - but probably acceptable to achieve such a difficult effect.

Sam Benge used Voro++, and I'm wondering if maybe there might be an external
library that could generate a Voronoi subdivision of a volume.  Though I don't
like that idea very much.

I've taken several looks at the source code for generating the crackle pattern
over the years, and it's long, somewhat complicated, and would be a project on
its own to convert to SDL.

I'm in favor of the isosurface idea, as it seems to be the fastest, most
reliable, and most direct approach.

RL has me running around doing non-POV-Ray stuff (Damn you, RL!) so I just have
less time than usual...   :(

Post a reply to this message

Op 5-7-2021 om 13:09 schreef Bald Eagle:
> Thomas de Groot <tho### [at] degrootorg> wrote:
>> What is missing is the internal structure of minerals,
>> especially in the ears. I am pondering the problem
> 
> Yeah - it's the old "POV-Ray renders _surfaces_ not volumes" problem/trap that i
> fall into from time to time.
> 
>> and think into the direction of a media,
> 
> I think that might lead to an unintentional detour: the pursuit of a PhD in
> media
> 
>> or maybe, a special kind of isosurface (I am thinking of my rings of Saturn in
2017),
> 
> That has potential, as if it were done right, you'd have a lot of separate
> surfaces, but then you'd have spaces in between - no idea if that would be
> problematic or not.  You'd also have to match up the texture with the isosurface
> domains.
> 
>> or again, a rand_in_object fill with small particles.
> 
> That would be super tricky and take for-ever.
> Exposure of the set of points for crackle to SDL, allowing for "one-pass"
> placement of "particles" would be what we'd need, and I can still see that as
> being "slow" - but probably acceptable to achieve such a difficult effect.
> 
> Sam Benge used Voro++, and I'm wondering if maybe there might be an external
> library that could generate a Voronoi subdivision of a volume.  Though I don't
> like that idea very much.
> 
> I've taken several looks at the source code for generating the crackle pattern
> over the years, and it's long, somewhat complicated, and would be a project on
> its own to convert to SDL.
> 
> I'm in favor of the isosurface idea, as it seems to be the fastest, most
> reliable, and most direct approach.
> 
I'll ponder a bit more. I have couple of other things/ideas to address 
presently.

> RL has me running around doing non-POV-Ray stuff (Damn you, RL!) so I just have
> less time than usual...   :(
> 

Yeah... Often not possible to avoid.

-- 
Thomas

Post a reply to this message

> Thomas de Groot <tho### [at] degrootorg> wrote:
>> What is missing is the internal structure of minerals,
>> especially in the ears. I am pondering the problem
> 
> Yeah - it's the old "POV-Ray renders _surfaces_ not volumes" problem/trap that i
> fall into from time to time.
> 
>> and think into the direction of a media,
> 
> I think that might lead to an unintentional detour: the pursuit of a PhD in
> media
> 
Using SSLT and a scattering/absorbing media that match the texture could 
help, but probably not in all cases. That would be the quicker to parse, 
and probably also the faster to render.

Post a reply to this message

Op 05/07/2021 om 20:51 schreef Alain Martel:

>> Thomas de Groot <tho### [at] degrootorg> wrote:
>>> What is missing is the internal structure of minerals,
>>> especially in the ears. I am pondering the problem
>>
>> Yeah - it's the old "POV-Ray renders _surfaces_ not volumes" 
>> problem/trap that i
>> fall into from time to time.
>>
>>> and think into the direction of a media,
>>
>> I think that might lead to an unintentional detour: the pursuit of a 
>> PhD in
>> media
>>
> Using SSLT and a scattering/absorbing media that match the texture could 
> help, but probably not in all cases. That would be the quicker to parse, 
> and probably also the faster to render.
> 

Thanks for that Alain. I agree and it would probably have my preference. 
I seem to remember I did some testing along that line in some distant 
past (or even universe)... Shall have to revisit my nooks and crannies 
somehow. <sigh>

-- 
Thomas

Post a reply to this message

In terms of isosurfaces, I was thinking about something like this.

-- 
Thomas

Post a reply to this message

Attachments:
Download 'isosurface_functions_test-onion_01.jpg' (164 KB)

Preview of image 'isosurface_functions_test-onion_01.jpg'