Celestia Forums

Chris, I have updated my code from the CVS and saw that you have added the Cubic interpolation...

I have tried it but it doesn't seem to work !

For example for Galileo instead of a straight line I have a sort of Cycloid !

The probe rotate around a virtual vertical axis with a radius of ~500000 km

normal :
_____________________>

cubic :

_ __ __>
X X
| | | |
\/ \/

Kendrix wrote:Chris, I have updated my code from the CVS and saw that you have added the Cubic interpolation...

I have tried it but it doesn't seem to work !

For example for Galileo instead of a straight line I have a sort of Cycloid !

The probe rotate around a virtual vertical axis with a radius of ~500000 km

normal :
_____________________>

cubic :

_ __ __>
X X
| | | |
\/ \/

When did you update? There was a bug in the original code that I fixed with a subsequent checkin. Also, it's disabled by default; my plan is to modify the .xyz format to allow metadata specifying the interpolation type and precision of the data.

--Chris

I updated 15 min ago !

And I have modified the constructor to have cubic by default...

I used Galileo and Cassini to test the code, and they seemed to move smoothly. Can you give me a specific date/time where you see Galileo behaving weirdly with cubic interpolation on?

--Chris

Ok for example with Galileo :

on the 6 Nov 2002 if you set the timerate to 1000x you will clearly see it !!!

If you are away of the planet (5,000,000 km) the bug seems to disapear !

There was indeed a bug . . . I checked in a fix for it and made cubic interpolation the default (though I may switch it back to linear and require cubic interpolation to be specified in the .xyz file.)

--Chris

I'll try that tonight...

Thanks !