Lightnews — Scholar-powered news

@jbakergraphics.bsky.social

2.5K followers 1K following 580 posts

Previously Engine Programmer @idSoftware interested in: C++/OpenGL/Voxels/SDFs/Pathtracing/GPU Compute/Photography Co-organizer of Graphics Programming Virtual Meetup Project writeups, blog posts: https://jbaker.graphics/index.html

jbaker.graphics

Posts Media Videos Starter Packs

Pinned

jb @jbakergraphics.bsky.social · Sep 5

$pathtraced fractal in Daedalus, using displacement mapping$

3 3 46

jb @jbakergraphics.bsky.social · 1d

business.kodakmoments.com/sites/defaul... seems kodak publishes some... damn, cool

business.kodakmoments.com

1 1

Reposted by jb

jb @jbakergraphics.bsky.social · 2d

very cool to see meaningful variation in the behavior of the different light LUTs... some experiments with low pressure sodium, fluorescent, halogen sources in a field of glass spheres

abstract image made of light captured by a film plane, pathtraced in Newton

1 1 21

jb @jbakergraphics.bsky.social · 1d

I think it's cool to try this approach, this is what's called forward path tracing, where traditional path tracing is going to start from your eye in search of light... I guess after you've done enough iteration with more traditional path tracers you want to change of pace, maybe lol...

jb @jbakergraphics.bsky.social · 1d

It is like they tell you - very unlikely that any given ray emitted from a light source is going to hit your film... It's more likely when you have perfect mirrors and glass, introduce any diffuse materials and you immediately lose any form of coherence from the random scattering...

1 1

jb @jbakergraphics.bsky.social · 1d

www.ipol.im/pub/art/2017... saw a paper recently that added this layer beyond what I am doing... I do a little jitter, so that each day that hits actually applies energy in a small area, but nothing this elaborate (yet)

www.ipol.im

jb @jbakergraphics.bsky.social · 1d

it is acting a lot like film grain, these bright speckles are like the small silver halide crystals in film stock that get exposed by stray light... here, they are "exposed" by some random, stray ray, contributing to that pixel tally, leaving a firefly

1 3

jb @jbakergraphics.bsky.social · 1d

this rocks - I saw that video and thought it might be interesting to try it with sudoku... but you think for a few minutes and realize, no way, right? just much, much too large, seems like

1 1

jb @jbakergraphics.bsky.social · 1d

It's actually an interesting problem, how you reduce this 32-bit sum to a usable color... this is an insane amount of dynamic range - caustics can be many, many orders of magnitude brighter than background noise. I'm looking at sigmoid curves and implementing some histogram visualization to help

2 5

jb @jbakergraphics.bsky.social · 1d

What's interesting is that this is extremely brute force... rays are emitted from one or more light sources, and refract through multiple interfaces, up to 64 bounces, until they (maybe) hit a film plane. When they do, they make some small contribution to a sum via an atomic increment

1 4

Reposted by jb

Dan Miller @dan0mighte.bsky.social · 2d

@limeblossom.bsky.social gave me the idea to treat each box as unique objects rather than as interchangeable with each other. This creates more states so its a bit harder to analyze. But it also creates a ton of symmetries which I think make it quite pretty!

4 2 8

jb @jbakergraphics.bsky.social · 1d

1 3

jb @jbakergraphics.bsky.social · 1d

1 2

jb @jbakergraphics.bsky.social · 1d

1 2

jb @jbakergraphics.bsky.social · 1d

1 2

jb @jbakergraphics.bsky.social · 1d

1 3

jb @jbakergraphics.bsky.social · 1d

1 3

Reposted by jb

jb @jbakergraphics.bsky.social · 3d

introducing some spectrally varying behavior

1 2 8

jb @jbakergraphics.bsky.social · 2d

very cool to see meaningful variation in the behavior of the different light LUTs... some experiments with low pressure sodium, fluorescent, halogen sources in a field of glass spheres

1 1 21

jb @jbakergraphics.bsky.social · 2d

I'm surprised literally daily by the user facing bugs in windows explorer in w11, idk why software this bad is getting developed, I think it may just have to do with the fact that ms have absolutely no idea what the hell an operating system is for anymore, it now serves a separate business purpose

1 1

jb @jbakergraphics.bsky.social · 2d

interesting pattern I keep hitting... UI element opens an app... which triggers an update... app fails to restore prior hook which opened it

like there's a "focus mode" in the tray, I was curious, hit it the other day... clock app opens and runs an "update" process and then just seems to sit there