Meanwhile...

With more experimentation, I'm getting a better understanding of how fog and
turbulence work together. So I made another animation (only two parts this time)
which might be more understandable, visually.

It uses my experimental 'mist' file again (same black pigments, and zero for all
the finish settings.) I added test objects again but in a more useful way. The
camera is a bit higher up this time.

PART 1 is just a slowly changing' turbulence' value. PART 2 has a 'set'
turbulence, but the camera moves.

Some interesting observations, to help clarify my previous comments:

Purely as a 'visual analogy', it looks like fog is composed of two distinct
parts: the overall smooth depth-based brightening that fog imparts to a scene--
like simply adding a variable gray value to all the pixels based on a
'depth-map' -- and the turbulence effect itself. (Of course, both of those
effects are probably part of the same code construct or paradigm.)

The turbulence splotches are not 'brighter' than the overall fog; rather, the
fog areas 'in-between' the splotches are actually erased (another analogy would
be that those areas have variable transparency compared to the overall fog.) The
brighter splotches are just the 'normal' fog brightness. I checked this in
Photoshop by comparing two renders, one with turbulence, one without.

The turbulence effect itself is *apparently* only on object surfaces (again,
that's just my own analogy of the visual results, not the underlying code
paradigm.) The individual splotches do not 'fill the space' volumetrically
between objects, like media would; they stop at object boundaries...which agrees
with the general idea that fog is just a depth-based brightening of each
camera-ray/pixel, like an overlay (but a more sophisticated 'pseudo-volumetric'
one, 'behind the scenes'.)

Turbulence is MUCH more apparent when the camera is 'inside' the fog rather than
'above' it.

When the overall fog is made thinner (i.e., using an increased  'distance'
value), the turbulence splotches DON'T thin out as much; it's actually a better
way to see the effect. The splotches do become more transparent, just not at the
same rate.

Turbulence  *can* be scaled in 'size' (meaning, the sizes of the individual
bright splotches and the overall appearance)-- but it's done with the TURBULENCE
parameter, not OMEGA. I think the documentation is confused on that point.  And
turbulence scale depends on how close an object is to the camera-- objects in
the distance show a smaller-sized turbulence effect on their surfaces (the
splotches become much smaller.) That's an obvious correlation of course, as with
*any* 3D spatial pattern in POV-Ray when applied to grouped objects-- bumps,
bozo, etc. But the pattern itself (or maybe the pattern's spatial position)
changes somewhat with the camera position-- made obvious when the camera is
animated. (It wouldn't be noticible in a typical still image.)

Post a reply to this message

Attachments:
Download 'fog_animation_test_2.mp4.mpg' (3490 KB)