Pabs wrote:

> David Fontaine wrote:
>
> > Comments?
>
> No but an observation - From the second mpg it is not obvious but from both
> it seems that either the math is a bit out or none of your objects are aligned
>
> in the plane of no change in projection area - the plane in which, no matter
> where the camera is in such a sequence an object will occupy the same
> proportion of the image.
> If the lego pirate was in this plane it should jump out at you without getting
> bigger AFAIK

Oh, is that how it's usually done? Well that's easy enough too :)

Here's what I was doing:

---------
|   a   /|
|      / |c
|-----/  |
| b  /|
|   / |
|  /  |d
| /   |
|/    |
C

C=camera, d varies with a and b, c is constant, clock = 0..1
I want a and b to change linearly, so
a = a0 + (a1-a0)*clock
b = b0 + (b1-b0)*clock
a/b = (d+c)/d, d = bc/(a-b)
angle = 2*atan(b/d) =
2*atan((b0+(b1-b0)*clock)/((b0+(b1-b0)*clock)*c/((a0+(a1-a0)*clock)-(b0+(b1-b0)*clock))))

Now you can do it for any rate of change in size :)

--
David Fontaine  <dav### [at] faricynet>  ICQ 55354965
My raytracing gallery:  http://davidf.faricy.net/

Post a reply to this message