"Anthony D. Baye" <Sha### [at] spamnomorehotmailcom> wrote:
> Christian Froeschlin <chr### [at] chrfrde> wrote:
> > On 22.01.2015 6:17, Anthony D. Baye wrote:
> >
> > > A lot of that is because of the way I'm storing the point data between frames.
> > > I'm currently working on some improvements that could speed things up
> > > significantly.
> >
> > You might benefit from persistent variables in Uberpov for this
>
> no worries.  I fixed it.  It's WAY faster now.
>
> I simply changed the way I saved the data -- from an array of points to a union
> for each frame, with an index file that gets rewritten every frame and forms a
> union of the unions.  My test render only took about 8 hours.
>
> Though I suppose persistent variables would save quite a bit of time, I really
> want this to be useful in the official version.
>
> I also fixed a problem with the random distribution of particles being the same
> for each frame.
>
> Regards,
> A.D.B

I should have mentioned that the previous animation was actually a shorter test,
only 15 seconds as compared to 30. just to test the random fix.

I have more ideas for this. I'm thinking about support for multiple files --
different files for each spraycan -- and I'm considering changing the way I
handle the setup.  Currently there are quite a few parameters, and I have some
ideas...

Here's the second test. It looked really bad with my first attempt at fixing the
rand bug.

Regards,
A.D.B.

Post a reply to this message

Attachments:
Download 'spray2.avi.dat' (163 KB)