[DiscordArchive] Sure your ray detects the triangle, but how is the value calculated once you have found the triangle
[DiscordArchive] Sure your ray detects the triangle, but how is the value calculated once you have found the triangle
Original source
<:think:822486128329883719>
rektbyfaith
07-02-2023, 11:29 AM #11Archived author: Nix • Posted: 2023-07-02T11:29:58.519000+00:00
Original source
<:think:822486128329883719>
Original source
also for what I'm writing about, any regular method will work for majority of all triangles until they get elongated. This is the particular place where typical traditional checks fails.
rektbyfaith
07-02-2023, 11:31 AM #12Archived author: Nix • Posted: 2023-07-02T11:31:12.147000+00:00
Original source
also for what I'm writing about, any regular method will work for majority of all triangles until they get elongated. This is the particular place where typical traditional checks fails.
Original source
With the obvious exception that if you do an actual ray and simulate it, that should fix it too
rektbyfaith
07-02-2023, 11:31 AM #13Archived author: Nix • Posted: 2023-07-02T11:31:29.468000+00:00
Original source
With the obvious exception that if you do an actual ray and simulate it, that should fix it too
Original source
because I used shoot incorrectly, it's not like the ray move or anything, it's just to check if a triangle is intersected with some fancy math, admittedly idk how ray tracing works but I doubt it checks each triangle of the scene for each ray it shoots, that sounds way too expensive in that case
rektbyfaith
07-02-2023, 11:33 AM #14Archived author: Adspartan • Posted: 2023-07-02T11:33:39.590000+00:00
Original source
because I used shoot incorrectly, it's not like the ray move or anything, it's just to check if a triangle is intersected with some fancy math, admittedly idk how ray tracing works but I doubt it checks each triangle of the scene for each ray it shoots, that sounds way too expensive in that case
Original source
Then that sounds like a continous ray
rektbyfaith
07-02-2023, 11:33 AM #15Archived author: Nix • Posted: 2023-07-02T11:33:52.432000+00:00
Original source
Then that sounds like a continous ray
Original source
You find the earliest point of intersection which typically gives you T back, or time to collision
rektbyfaith
07-02-2023, 11:34 AM #16Archived author: Nix • Posted: 2023-07-02T11:34:09.922000+00:00
Original source
You find the earliest point of intersection which typically gives you T back, or time to collision
Original source
You can then as you say, use your movement vector and forward it with the speed and time to find the exact point of collision
rektbyfaith
07-02-2023, 11:34 AM #17Archived author: Nix • Posted: 2023-07-02T11:34:51.872000+00:00
Original source
You can then as you say, use your movement vector and forward it with the speed and time to find the exact point of collision
Original source
Typically done using SAT (Single Axis Theorem)
rektbyfaith
07-02-2023, 11:35 AM #18Archived author: Nix • Posted: 2023-07-02T11:35:20.195000+00:00
Original source
Typically done using SAT (Single Axis Theorem)
Original source
Anyway, my point wasn't to say Noggit does it wrong, just to highlight that if any such issues occur with elongated triangles, BTI is an option
rektbyfaith
07-02-2023, 11:35 AM #19Archived author: Nix • Posted: 2023-07-02T11:35:55.007000+00:00
Original source
Anyway, my point wasn't to say Noggit does it wrong, just to highlight that if any such issues occur with elongated triangles, BTI is an option
Original source
![[Image: elon.png?ex=690c882d&is=690b36ad&hm=722d...81090a0bf&]](https://cdn.discordapp.com/attachments/408013572430626817/1125027145761161246/elon.png?ex=690c882d&is=690b36ad&hm=722dc0261aa536230dd9fd8af7cd57db8072df11168788f0c20cfe281090a0bf&)
rektbyfaith
07-02-2023, 11:36 AM #20Archived author: Adspartan • Posted: 2023-07-02T11:36:13.623000+00:00
Original source
1 Guest(s)
1 Guest(s)