/img/ - Aesthetics & Images

posting this here but i wonder if someone has the source for this clip?

>>26987
extremely based

Same

>>25558
>>25559
He's so hot. I've been wanting to play this game just for him for the longest time.

probably just Agent Sunglasses™ but he's such a megane I think he fits ITT as well. I'd do him

>>26658
don't have a good example unfortunately, but the tree you highlighted is actually on the right side in reality. The glass marble mirrors the image in both axis as well as adding a fisheye effect, that's why its not in the frame on the picture.

I guess I've learned this in physics class? Look into how lenses work, but there are various optical effects that are hard to model with lenses like caustics and that weird thing water does when you look up from below.

>>26676
there is no easy way to map the contour on an irregular surface like a glass of water accurately. every ray of light you see takes an independent path between your retina and the object, and gets refracted on each material boundary based on the density of the lense material.

To accurately calculate the outline you would have to model every ray of light in it, and know rhe differences of density of the air, glass and water. A difficult task for a raytracing computer, a very difficult task for a human.

I think the best you can do is watch some youtube videos on the basics of lenses and try to make a believable guess when you are drawing, or use a similar object in real life to see how it would look. (if this is for drawing)

>>26677 aw man. i lost this thread
well i have seen the lens theory and by assuming n1.sin@1 = n2.sin@2 , one can then assume than sin@2 is just n1/n2 * sin@1, thus the size of @2 is as big as @1 but with n1 defining the "forward" and n2 multiplying the "hypothenuse", since they are the same side of the ratio, i guess, so the bigger the n2, the longer the hypothenus, such and such

but this all is my assumption only, and honestly, while it is expensive for computer,
i am surprised it is not done manually first to affirm the guessing, unlike other things like reflections which are more general in drawing? very strange
expectantly i need a stronger assurance than that of a 3d software or photo, which also means youtube tutorials are only enough descriptive than completely

bump more info i guess
just looking for some sort of… finality to this understanding
https://4-ch.net/img/res/4686.html#i4686

i guess it's about, "countouring" and that's no science

Glass

>>26788
>>26787
I thought these were a repost. How original, heh?

>>23804
https://novelai.net/
Stable Diffusion locally

i like the bing one.. my pc can run its own but i'm too stupid to set it up