Op 15/01/2024 om 00:50 schreef Samuel B.:
> Thomas de Groot <tho### [at] degrootorg> wrote:
>> Op 13/01/2024 om 02:30 schreef Samuel B.:
>>>
>>> Is Analog even still being published? Some of those stories were great.
>>>
>> It seems to be:
>>
>> https://www.analogsf.com/
> 
> Ah cool. I'm a bit tempted to get a subscription for my mom (which would
> eventually make its way here as well). I should probably check with the library
> before doing that, though :D
> 
<grin>
It is amazing that they still do exist after such a long time (1930, if 
I am correct). Analog was indeed my first introduction to the (serious) 
SF literature.

>>> The default uv mapping is going to be different for every primitive, of course,
>>> and some stretching is guaranteed in any case.
>>>
>> Yes, I noticed that.
>>
>>> I'm thinking that for complex objects, triplanar mapping might be the best thing
>>> to use. (Triplanar mapping in POV-Ray would probably use 3 nested slope maps,
>>> with each axis getting one planar map). There /is/ a blending issue when using
>>> triplanar mapping, but it can be mitigated somewhat. It's one of those things I
>>> haven't gotten around to yet (I was going to use something similar to depict
>>> claymation-style fingerprints on objects), but if I make any headway I'll post
>>> it.
>>>
>>> I'm thinking crackle form x patterns as a basis, but it's going to take a bit of
>>> work...
>>>
>> Exciting info indeed! I shall study these ideas also; I don't like to
>> sit back and wait for others to do the dirty jobs :-)
>>
>> --
>> Thomas
> 
> That's the spirit!
> 
> I managed to get one thing done today: the triplanar mapping
>
(https://news.povray.org/povray.binaries.images/thread/%3Cweb.65a4705751388f9716bed5696e741498%40news.povray.org%3E/).
> It's very basic and doesn't have a sharpness setting (rate of transition between
> normals).
> 
That is really looking good, Sam! I shall have to play with this.

Yesterday, I was exploring different issues more or less related to this 
and I experimented with the lunar crater image inside a cubic pattern, 
thinking that, perhaps, it would do something /similar/ to triplanar 
mapping. Of course it didn't, but it was a useful sidetrack because I 
could use it in an interesting way with a superellipsoid for instance, 
and even with the ovus object (which last one has a strange basic 
uv_mapping which I cannot agree with).

I had not got around to switch to pigment_pattern or image_pattern use, 
but it was on my ToDo.

> The craters idea will have to wait for another day.
> 
Of course. Still, a great progress already and worth a warm round of 
applause!

-- 
Thomas

Post a reply to this message