ingo <ing### [at] tagpovrayorg> wrote:
> in news:web.5c54c166738ba468765e06870@news.povray.org Bald Eagle wrote:
>
> > I guess if you wanted that one-pixel wide strip at the center to use
> > the other function, then you'd use the 4-argument version and repeat
> > the first function for the u = 0 case.
> >
>
> Bill,
> on an almost "aside way" you explain me that things can be done simpler

Well, it's not "simpler" - it's just more straightforward for the easily
confused such as me.   :P


> but seriously, I almost exclusivly use the 4 term select as:
> select(query, ignore(=-1), true, false)

Well, whatever works for you, if you use it consistently without problems - go
for it.

> Regarding the pixel in the middle, where does it belong ??? In case of a
> stereo cam, the left or right side of the image, there is no in between?

Nope.  I suppose you could do the calculations and find out where to shift the
camera so that it's "1 pixel" over one way or the other, and then adjust for
that...
Can you check and see if abs (u) > epsilon so that you can filter out a lot of
"0" results...?


> they are indeed rather under documented. It's hard to understand what
> exactly happens, but once I get it I'm willing to do a write up (to be
> corrected by the ones that realy understand).

Sure, I think the idea is to just post some reasonably well commented code, that
we can look at when we want to implement a similar camera structure, and if
necessary, supplement with some renders, drawings, etc.

> An other thing is that they
> seem to be "verbose". When I try to reimplement an existing 10 line of c++
> code cam it seems I need at least 5 times more space. It requires a lot of
> trials and mostley errors as I have a hard time wrapping my brain around
> some things and my brain isn't very structured, more intuitive.

I seem to be able to teach myself what I need to know faster when I try to
explain the problem I'm having, or when I try to condense it down into a macro
or a formula.


> Ray tracing is about rays, objects an cameras and I realy think POV-Ray
> deserves a whealthy choise of ready made cameras. Paul Bourke has defined
> a very nice set of stereo cameras a decade ago that I.M.H.O would be very
> welcome in an official distribution.

I think there's a LOT of Bourke's work that ought to be readily available and
used a lot more, and even implemented under the hood in C++ so that it's not
stuck in super-slow SDL.

> Le Forgeron, as I discovered today,
> has made a set of _/incredible and bewildering/_ set of 'planet mapping'
> camera's.

Jerome is a really intelligent, inventive, and hard-working guy, and I wish I
had more time to sit down and play with all of the goodies he has coded up over
the years.

> The https://www.cartographersguild.com would love them (/him for
> that), they do not seem to have discovered them yet.

Yet another forum to advertise the wonders of POV-Ray in   :)

> Let's not forget
> Fabrizio's (Clodi) OSD camera.

I have not (yet) heard of that.
Aha -
http://news.povray.org/povray.general/thread/%3Cweb.56d38071f91c97bb4e8811590%40news.povray.org%3E/?mtop=406634

(Clodo)

Post a reply to this message