From: mscir@yahoo.com   
      
   On 6/5/2017 7:13 AM, JTS wrote:   
   > On 2017-06-05 01:00, JTS wrote:   
   >   
   >   
   >>   
   >> By the way I have written up this afternoon the Newton method for   
   >> ray-curve intersection (one dimensional and without any refinement) in   
   >> Matlab and I hope to get an insight soon of where is that the ray   
   >> tracing of the CPC fails (above linked image).   
   >>   
   >   
   > I got a hint for what is wrong between Goptical and the CPC, and   
   > following the program execution I know now what is happening: at the   
   > start value the CPC sagitta function evaluates to NaN (it has to happen   
   > because of the way the CPC profile is defined). So I extended the   
   > sagitta definition outside the aperture of the CPC by reflection (so to   
   > have a smooth derivative at the aperture radius) and the program is   
   > behaving much better:   
   >   
   > https://www.flickr.com/photos/21334234@N07/34304612313   
   >   
   > There are still a few rays for which the intersections are not found, I   
   > need to figure out whether I made a mistake somewhere with my extended   
   > function definition or it is the Newton method which is - in this case   
   > really - not working for this surface.   
      
   Interesting. Any chance you could post both functions here, if that's   
   not difficult, or asking for proprietary code?   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)