[Plugin] Raytracer
-
@panga said:
I need to make a raytrace from on point in every direction
@unknownuser said:
I know that could generates a lot of rays and calculation, but It would be nice to have this function with maybe an option where we can choose the number of rays like : 1 to 10 rays / 10 o 100 rays / 100 to 1 000 rays.
I say, yes, it'd be a whole lot of rays. So, would it be like sub-dividing a sphere and shoot a ray for every segment?
@panga said:
If it's possible, the rays should be drawn as guidelines, and a guide point could be automatically drawn on each intersection with the rest of the model.
YEs, can draw them a guidelines, yes. But how many points of origins are we talking about? There'd a huge amount of CLines. Would there be a limit to how far they reached? Would it be ok to draw only the rays that hit something?
Do the rays need to ignore other geometry other than the target group/component? -
@unknownuser said:
YEs, can draw them a guidelines, yes. But how many points of origins are we talking about?
For my specific usage, there's two possibilities and the first one of them could be already fanstastic:
- From one point at time
- From one complex object (component or group)
@unknownuser said:
There'd a huge amount of CLines. Would there be a limit to how far they reached? Would it be ok to draw only the rays that hit something?
Do the rays need to ignore other geometry other than the target group/component?Yes, Yes and Yes. In fact I only need to evaluate "covisibility". So I could select a point on my model (or a complex object) and the plugin draw a certain number of rays only where there's intersections with the model, or with a specific component/group if it's easier to implement/calculate.
@unknownuser said:
So, would it be like sub-dividing a sphere and shoot a ray for every segment?
This could be a good solution with a sphere that surrounds the whole model.
Thanks for your answer.
Regards.
-
@panga said:
@unknownuser said:
So, would it be like sub-dividing a sphere and shoot a ray for every segment?
This could be a good solution with a sphere that surrounds the whole model.
I didn't mean to put an actual sphere in the model, just how one would determine which of the infinite possible directions the rays should be shot. What I meant by the sphere was they the ray's be evenly distributed based on how many rays one shoot.
-
@unknownuser said:
I didn't mean to put an actual sphere in the model, just how one would determine which of the infinite possible directions the rays should be shot. What I meant by the sphere was they the ray's be evenly distributed based on how many rays one shoot.
Yes I understood you. And controlling the number of segments/faces of the "virtual sphere" could help to have more or less rays. Whatever you can do for that, thanks in advance.
EDIT : In fact, what could be really nice is a "sun" or a omnilight that I could place wherever I want in my SU model, and with shadows activated I could see the illuminated parts of my models and the dark ones...
-
I'm writing this down in the todo list. Will see what I can do.
-
hmm.... I'm not quite sure how to calculate the rays to shoot. Been all too long since I've done math like this...
-
My knowledge in maths are very limited...so I'll not be able to help you on that. Anyway, thanks for giving it a try !!
-
Found a solution!
-
I had no doubt about that !!
-
When I say I I mean, someone smarter than me helped me out.
-
Whatever if someone helped you or not, you were the one looking for the solution...so you found it !
-
sir Thomthom,
thank you very much for this plugin.
what a coincidence! This day I've been thinking of raytracing and what is its concept. I think I'll understand this once I use your plugin and of course read some notes about it. -
grrrr.... SU8 has even more troublesome bug with the
raytest
method. I'm struggling to find a workaround.
I got the Spray Ray working in SU7 - but SU8 is being a PITA. -
Version 1.1.0
- Requires TT_Lib2 (v2.3.0)
- Trace Spray Ray.
- Bugfixes.
Note: SU8 does not process rays properly and the results are not reliable!
I've been unable to work out a successful workaround. -
Kind-of like a cob-web... if the CLines was replaced with catenary curves...
Wiki uses cob-webs as an example for catenary curves http://en.wikipedia.org/wiki/Catenary
I have no idea what the usage would be, other than for Halloween... but... cool...
-
Thanks Thomthom, I'll try it. Hope you can find a way to make it work under SU8 later.
EDIT :
Hi Thomthom,
I made a Test under SU8, and it Works pretty well. Until now, I test it with a max of 10 000 rays, and it takes about 15 seconds to calculate, so It's pretty fast. The only thing is it seems that some directions are ignored by rays, see on the picture below :
Whatever, It's already a really good tool for me. I'm thinking of other options, but I'll do some tests before talking about...See u and thanks again.
-
Damned I have missed this hidden collection under a falacious title
-
@panga said:
EDIT :
Hi Thomthom,
I made a Test under SU8, and it Works pretty well. Until now, I test it with a max of 10 000 rays, and it takes about 15 seconds to calculate, so It's pretty fast. The only thing is it seems that some directions are ignored by rays, see on the picture below :
Sorry for the last reply - but I never noticed this edit. (Don't think edits marks the topic as unread.)
As mentioned in the first post, SU8 has a nasty bug in the raytracer. Only thing one can do is wait for Google to fix it and release a new version. -
Hi Thom
I am getting a few weird results with the drop component option...
I thought I would try an test with a series of cars on a slope to simulate a parking lot scenrio.??? There are some non conformers
Test file attached.
Cheers
Dave
-
Drop by Bounds, right?
I see the car crashes as well...
Advertisement