I'm having trouble with a parameterized calculation in three forms

((T*P)/2)*pow((clock - index(Start))/index(Duration),2)

(P*T*T)*pow((clock - index(Start))/index(Duration),2)

and

T*P*( (clock - index(Start)) / index(Duration) )

index(T) is calculated by a macro which takes a 3d vector T
_TICK_ is defined in another macro.

#macro index(T)
     #local Seconds = T.x*3600 + T.y*60 * T.z;

     _TICK_*Seconds     // where _TICK_ is (final_clock - initial_clock) /
animLen
#end

I passed various elements of the calculation to a debug statement, and the
problem seems to arise in the (clock - index(I))/index(D) portion, the result
being that my Delta_T is about 10,000 times larger than it should be.

which shouldn't work

When I run the same calculation with set values as below

(5*1080/2)*pow((clock - 0)/index(5),2)

I get exactly what I should get, as confirmed by my pocket calculator with
various different values.

I have three files illustrating this problem in a zip file I'll upload to p.b.a

test.pov illustrates the solution that works while test3.pov illustrates the
problem

I've been rendering with the following command line:

+a0.03 +ki0.0 +kf1.0 +kfi1 +kff600 -D -GA

and encoding with mencoder at 60fps

any insight on this problem would be appreciated

Post a reply to this message