 |
|
|
|
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Op 28/04/2021 om 19:26 schreef Bald Eagle:
> Thomas de Groot <tho### [at] degroot org> wrote:
>
>> I do not want to have you wait: This is the first render with StepNoise.
>> My judgment is: perfection.
>
> .... holy crap! :O
>
Indeed. Was what I thought too.
> I would have never guessed it would THAT good! Thanks for not making me wait.
> That is indeed MUCH better than the turbulated cells experiment. I'm glad that
> it's panning out and we have something to work with as a good base pattern.
>
I still have to test the grain size control and such but we can indeed
ditch the turbulated cells. I was happy and not happy at the same time
with those; especially the larger grains were problematic.
> I'd say the next thing to do would be to have that texture be an entry in a
> marble pattern texture_map, with a quartz texture as the vein.
>
Sure, and don't forget the quartz veins in the granite :-)
> After that, you're going to be the geological expert on whatever else might make
> worthwhile additions.
>
Good. Work enough on the horizon.
--
Thomas
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Op 28/04/2021 om 22:37 schreef Bald Eagle:
> Quick stab at adding the vein.
>
> Didn't like marble or planar, so I coded a Boolean math function for the stripe,
> and let it blend on the sides.
>
My latest experiments with veins was by using agate and agate_turb as
second texture layer. Needed to add a warp{turbulence *z} to hide the
layering of agate. The results were promising if not perfect. I do not
have an example render at hand at this moment.
> Adding a mod to the function will repeat it, but it's too regular.
> I tried warp {repeat x*4} but that repeats the whole pattern, and you get hard
> lines.
>
> I thought maybe I could layer it, but it's a pattern texture, and POV-Ray says
> "No way!"
>
> I also tried to add a specular highlight, and I'm too inept to get that to work
> either.
>
> Nor could I get the void created by using a difference to be anything but black.
>
> It's unclear to me at the present time how to have separate turbulence values
> for the stripe and the base StepNoise pattern, but I guess we'll figure that out
> as we proceed.
>
It's a first step.
> Currently using the sum of two turbulated stripe patterns, so maybe the
> placement of the stripes will have to be generated by a macro and an array of
> plane offsets.
>
> Anyway, this gives the general idea.
>
--
Thomas
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Op 29/04/2021 om 03:17 schreef Bald Eagle:
> And the code:
>
Thanks. Shall play with that too.
--
Thomas
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
alright. Getting tired after a day of tweaking.
Second, overhauled, StepNoise version. Several things learned:
- to be careful with the /scale/ of the grains. This is as good as I
dare to go. Larger minerals become increasingly unbelievable, so one
needs to find the balance (as with the turbulated cells obviously, but
StepNoise does a much better, although not perfect, job).
- the mask for the pigment_pattern needs a /very/ careful adjustment.
The present one is good but could probably be improved further. Good
enough for now.
- white grains are present but hardly visible. In the first version they
were placed at the wrong place within the colour_map, rendering rings
instead of mineral grains. Corrected that but need now to be enhanced a bit.
- sslt would probably increase realism, even in the frosted version. I
am going to look into that.
--
Thomas
Post a reply to this message
Attachments:
Download 'dakotagrains_sn_test2.jpg' (95 KB)
Preview of image 'dakotagrains_sn_test2.jpg'
|
|
| |
| |
|
|
|
|
| |
| |
|
|
hi,
Thomas de Groot <tho### [at] degrootorg> wrote:
> alright. Getting tired after a day of tweaking.
glad you did though, "Flanders Gray (tm)" looks ready for turning into kitchen
worktops. :-) (am interested to see it being used other than in a
demo/reference image)
regards, jr.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Thomas de Groot <tho### [at] degrootorg> wrote:
> alright. Getting tired after a day of tweaking.
>
> Second, overhauled, StepNoise version. Several things learned:
>
> - to be careful with the /scale/ of the grains. This is as good as I
> dare to go. Larger minerals become increasingly unbelievable, so one
> needs to find the balance (as with the turbulated cells obviously, but
> StepNoise does a much better, although not perfect, job).
>
> - the mask for the pigment_pattern needs a /very/ careful adjustment.
> The present one is good but could probably be improved further. Good
> enough for now.
>
> - white grains are present but hardly visible. In the first version they
> were placed at the wrong place within the colour_map, rendering rings
> instead of mineral grains. Corrected that but need now to be enhanced a bit.
>
> - sslt would probably increase realism, even in the frosted version. I
> am going to look into that.
>
> --
> Thomas
I am embarrassed to say that I can't see anything from that far. With respect to
that 10cm grid measures, I feel like you are no longer representing the same
rock? Did you work on it like that or with a closer camera? I do feel like
nothing is wrong with what little I see, though ! I have faith in you ! :-)
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Op 29-4-2021 om 19:21 schreef Mr:
> I am embarrassed to say that I can't see anything from that far. With respect to
> that 10cm grid measures, I feel like you are no longer representing the same
> rock? Did you work on it like that or with a closer camera? I do feel like
> nothing is wrong with what little I see, though ! I have faith in you ! :-)
>
The 'problem' with changing patterns or pattern modifiers, is that
scales get changed too. In a sense, you are right: it is not the same
rock indeed and you need to readjust your frame of reference. Don't let
it fool you: that 10cm grid does not represent anything real; I have
always bee troubled by those quasi-real measurements (I am going to take
out those cm-values and replace them by Martian Claw Units). My believe
is that size/dimensions in POV-Ray are dimensionless as it were and only
meaningful in relation to recognisable objects.
I attach a version with the material scaled by two. Still looking good,
imo, and much better than the cells pattern.
--
Thomas
Post a reply to this message
Attachments:
Download 'dakotagrains_sn_test3.jpg' (91 KB)
Preview of image 'dakotagrains_sn_test3.jpg'
|
|
| |
| |
|
|
|
|
| |
| |
|
|
scaled by four; starting to break down I think, but still believable.
--
Thomas
Post a reply to this message
Attachments:
Download 'dakotagrains_sn_test4.jpg' (84 KB)
Preview of image 'dakotagrains_sn_test4.jpg'
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Thomas de Groot <tho### [at] degrootorg> wrote:
> scaled by four; starting to break down I think, but still believable.
Have you tried doing a sequence of objects, with changing smooth-step thresholds
to see what effect that has? There's also the amount of turbulence applied
which will affect how the underlying pattern looks.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Op 30/04/2021 om 19:57 schreef Bald Eagle:
> Thomas de Groot <tho### [at] degrootorg> wrote:
>> scaled by four; starting to break down I think, but still believable.
>
> Have you tried doing a sequence of objects, with changing smooth-step thresholds
> to see what effect that has? There's also the amount of turbulence applied
> which will affect how the underlying pattern looks.
>
Not yet, but good suggestion. Had something like that on my ToDo list.
Want also to try sslt... and object-scale vs landscape-scale use... Too
many balls in the air atm.
--
Thomas
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
