Dear POVCOMP competitors
Following the postponement of the deadline of the POVCOMP and after reading
the comments posted here, I've decided to write a series of short
"tutorials" that aim to help the participants to improve their images (if
necessary). Several topics will be addressed in the weeks to come such as
concept, composition, modelling, texturing, lighting etc. The term
"tutorial" is used loosely in fact, as this will be mostly general advice,
hints and examples taken from the vast world of 3D (or photography).
Comments, corrections and discussion are welcome of course.
#1: References in 3D
While POVCOMP is a POV-Ray competition, its objective is also to demonstrate
that our beloved raytracer is a fantastic tool, able to create fantastic
images, to the large majority of people who have never heard of POV-Ray.
These people will be seing these images in magazines or at computer trade
shows: they will not compare the POVCOMP images to other POV-Ray images, but
to the sort of CG imagery that is now commonplace in movies and advertising.
The POVCOMP pictures will "compete" with the best, and, as we know, the bar
has been raised a lot since the early days of POV-Ray.
The following links are examples of recent 3D work taken outside the POV-Ray
community. It's just a very small sample, but it covers a lot of ground.
These works are for most of them created by professional graphic artists
using high-end expensive tools, and able to spend a lot of time on their 3D
work. Is it possible to do images like these with POV-Ray? There's no good
answer to that. Let's say that from a rendering perspective, POV-Ray can
really take you far, or at least far enough for its own limitations not to
be such an hindrance when it comes to create great images. These limits are
mostly in the modelling and texturing (but external tools like Moray, Wings
or Blender can take care of that), in the post-process features (forbidden
in the POVCOMP anyway) and, somehow in the rendering speed (since it's pure
raytracing). However, POV-Ray has also its own strengths - the SDL itself,
macros, functions, isosurfaces or mesh instanciation, for example - and
POVCOMP participants can build on that.
But above all, an impressive image isn't just a question of computing
horsepower. If the images below are good, it is also because they are good
in many things that are perfectly accessible to the POV-Ray user, such as
concept, composition or lighting. So have a look at the eye candy below, and
if you have entered (or plan to enter) an image in the competition, try to
look for images thematically similar to yours and see how those graphic
artists approached the topic. Look for the similarities and for the
differences. What did they do that you can't do in POV-Ray? What did they do
that you *can* do in POV-Ray?
- Graphic experiments
- POV-Ray and Poser computer images
Thanks Gilles, I think this is very cool what you are doing.
In fact, way back in an earlier thread, there was a little skirmish over
whether you would be entering or not and you announced that you would
not be. It shocked me at first but then my immediate reaction was to
want to respond with a proposal that maybe you could act in some sort of
technical advisory capacity for participants. But the idea never got
past the fanciful stage and I wasn't sure how it would be received. In
retrospect I kinda wish I'd suggested it then.
On a personal note, I did not enter the contest myself for a complex of
reasons, but one huge obstacle for me was the high resolution required
in the specification. I know that had I made the proposal, and had you
accepted, questions about the pitfalls of producing high res images
would have been salient for me. And still are. It seems to me that
that is one of the milestones that separates you and the other
organizers, who have greater experience technically, from many of us who
are less far along. Just wanted to throw that into the mix as you are
putting together your tuts.
Gilles Tran wrote:
> The following links are examples of recent 3D work taken outside the POV-Ray
> community. It's just a very small sample, but it covers a lot of ground.
> These works are for most of them created by professional graphic artists
> using high-end expensive tools, and able to spend a lot of time on their 3D
Wow, thanks. Those are some eye-openers and no mistake. I was kind of
stuck in the "how can I make the most realistic and detailed flowers and
trees" mode. That was competing with the "to heck with the flowers and
trees, maybe I can model some stuff on a table in a room" approach.
Time to do some serious re-thinking (and now we do have time.)