Op 19/12/2019 om 23:27 schreef Alain Martel:

>> "Bruno Cabasson" <bru### [at] cabassoncom> wrote:
>>> "And" <49341109@ntnu.edu.tw> wrote:
>>>> Alain Martel <kua### [at] videotronca> wrote:

>>>>>> Your cloud do have some great properties. Although it is 
>>>>>> unexpected simple(the
>>>>>> pov-code). The fluffy appearance is what I want. And the coverage 
>>>>>> is moderate.
>>>>>> But in fact the real geometry shows the top and bottom are 
>>>>>> truncated by the
>>>>>> boundary(800m~1200m). I will think how to solve this in the 
>>>>>> future, but if you
>>>>>> have idea I think it will be nice.
>>>>>>
>>>>>
>>>>> You may use a planar pattern to hide that. Just scale and translate it
>>>>> properly, and add a map or use an alternate wave type to control 
>>>>> how it
>>>>> need to fade near the container's edges.
>>>>
>>>> How do you combine these two(The original and the planar pattern)? I 
>>>> can't spend
>>>> time to try it right now, but I wish to look your code (even a 
>>>> simple example).
>>>
>>> I think the planar (or equivalent) pattern can be used as an 
>>> additional density
>>> function (that will multiply with the main one). I did use this for 
>>> some other
>>> clouscapes atempts.
>>
>>
>> Multiply is a straightforward idea that combine these functions. I 
>> have doubts
>> about it because it seem change the original pattern (unnatural) 
>> ....but I
>> hardly ever try it. You said that you have used it. Is the effect fine?
>>
>>
>>
>>
> 
> The planar pattern start at 1 on the X-Z plane and drop to zero 1 unit 
> above and under it.
> 
> There is no distortion involved.

Also, do not forget the spherical and boxed patterns which work 
according to the same principles as planar!! I use them regularly for 
atmospheric or cloudscape business.

-- 
Thomas

Post a reply to this message