> I'm confused. Is this all written in POV SDL?

Yup, all in POV-SDL.

> If so, how are you rendering multiple frames of an animation?

Well, the command-line "+KFI1 +KFF150", for example would give me 150 frames
with numbers from 1 to 150. Look it up in the docs. ;-)

> Doesn't an animation need a POV file for each frame, basically?

No, it doesn't, that's what the clock-variable is for: if you put a sphere
at <0,clock,0>, then during those 150 frames, it will move from <0,0,0> to
<0,1,0> (as clock is "0" in first and moves to "1" in last frame by
default).

> Couldn't you output these all first, then distribute the render?

I'm not sure if you know what Theo was talking about: he wasn't referring to
having different machines render different frames (like the IMP does it),
but instead, ALL machines render parts of the SAME frame. This is especially
useful for images which would otherwise require days and days of rendering
time. But for this to work, every machine must have the exact same data for
the scene...
Thinking about it some more: if every machine renders part of every frame,
then actually the distributed rendering should work. It just doesn't work if
the System has to backtrace the particles.

> I haven't used SDL very intensively, and certainly nothing as
> sophisticated as doing animation calculations in it beyond just "clock"
> based stuff, so I'm curious how you're using SDL to invoke the renderer
> yourself or something?

No, nothing of the sort. What I do is simple: if a certain file is present,
I load the data from there and use that for calculations. If it's not, I
generate new data.

Depending on some other variables, I thus either set the data to the initial
state, or just base it upon former data. In this manner, the same scene-file
with render differently every time it is run. Now, with an animation, that
is exactly the case: the same scene-file is run several times, and only the
clock-variable changes. If I additionally base my calculations on data that
is stored externally, and also change that data during the runtime, every
consecutive run will yield different results.

So, for this particle system I just put some extra parameters in these files
to track their "age", and let the data get "older" for every frame.
Understood, or should I clarify some more?

Regards,
Tim

-- 
"Tim Nikias v2.0"
Homepage: <http://www.nolights.de>
Email: tim.nikias (@) nolights.de

Post a reply to this message