Inside a Sphere Validation File Containing Another Circle

[ericshi03]

Hi, I've found that if I don't normalize my ray's direction in viewing_ray I am able to have the second circle show up when I run Inside a Sphere. That being said, I'm not sure if this is the intended solution as when I asked the prof last week he told me I should be normalizing my ray's direction. I was wondering if anyone else ran into this issue/was able to reproduce the two circles while still keeping the direction normalized.

Also, the JSON file only has one sphere object so I'm confused to why there are two in the validation file. Thanks!

[aakrohra]

Same here, not normalizing is the only way I've found to make inside-a-sphere work.

In Ray.h, "Not necessarily unit-length direction vector. (It is often useful to have non-unit length so that origin+t*direction lands on a special point when t=1.)" So I guess normalizing is not necessary?

[PyroPoro]

The validation file doesnt really make sense though. Theres only one sphere in the json so the normals should be consistent

[aakrohra]

I believe there's only one sphere because it has to do with the position being literally inside the sphere? So the smaller circle is the part of the sphere that is facing towards the camera from inside the sphere. But this doesn't really make sense to me either, clarification would be helpful

[lynnnnnn22]

Thanks aakrohra, yes there's only one sphere.
For the rays that hit the sphere with $t_1$ and $t_2$, $t_1 < t_2$, if $t_1 >$ min_t, we are seeing the outer surface of the sphere and this gives the outer ring we are seeing; if $t_1 <$ min_t $< t_2$, we are seeing the inner surface of the back of the sphere and this gives the inner circle we are seeing.
I don't think you need to normalize the ray direction.