Ken wrote:
> 
> Spider wrote:
> >
> > here is the image from .general discussion.
> >
> > //Spider
> 
>   I take it by your example you are implying that by choosing to use the
> filter component instead of the transmit component there comes a point in
> the procedure where it will show it's limitations as clearly evidenced by
> your example.
Both methods has it's pro's and con's. My example was made to show
both(even If I think that I should have used a greater max_trace_level)
The filtering method you showed has the advantage of making the light
coloured as well, and it will provide nicer reflections and light
effects on a background, while the transparence I pointed at doesn't do
this, but instead provides a smoother shadow. I know that the scene
isn't very good for showing this, but it was only a test I made to
reassure myself of the differnces and the results.

> It is particularly prevalent in a single color pigment like
> the red color you choose and is readily identified. I will allow you this
> victory and will not argue the point with you (too much anyway).
but please do. I greatly enjoy and appreciate your comments, since they
are good and has a great deal of experience bahind it.

>   On the other hand if you compare the color of the light shadows for each object
> as it appears on the floor you will notice that the transmit components chosen
> for your textures allow virtually all light to pass through the object unaffected.
> This is indicated by observing the neutral color of the shadow zone around your
> objects. From this we may argue that the light is not behaving in a rational and
> predictable manner.
Yes, That argument is perfectly valid, and should also go in both
directions, since it always is dependant on the scene it is supposed to
fit in.

>  Why is this you may ask. In reply I can only say that it is because the light
> should have the color shaded regions of the object added to the lights color
> as it is travelling thorough the object. In consequence we must conclude that
> using the transmit only solution is not an altogether accurate model and more
> work must be done to resolve this seeming insurmountable problem.
"insurmountable" Gah, you had me bring out the lexicon for this. 
(http://www.nada.kth.se/skolverket/swe-eng.html)
Well, as well, now I know the word, but is it really so?
 
>   Now we must examine our options. A moments thought about what we have
> learned tells us that perhaps we need to add some of the properties of
> each component to achieve our desired goal of realism. Of course this
> is the answer we have been looking for all along.
Good one. I'm not sure of how to do this, since I haven't tried. Do you
have any good ideas on how to work it out for the best effect?
 
> In conclusion:
> 
>  Now we know that the ideal solution will be a combination of both transmit
> and filter components, in a carefully selected balance, to obtain optimum
> material performance levels. A few minor changes to the color map and it
> will be done.
Any tips ?

>   It is then, and only then, that all who will see it, may know the light
> for what it truly is, or truly shall come to be.  :)
Hmm, POV, may it rule forever.

//Spider
-- 
#declare life = rand(seed(42))*sqrt(-1);

Post a reply to this message