I think I'll operate from Mastodon, at least for a while. This should be a proper link:
Follow
Jacco Bikker
@j_bikker
Ex-game developer, ex-rasterizer, now doing ray tracing, low-level optimization and lecturing about it for fun and profit. Mostly for fun though! C+/"Sane C++".
Houten, The NetherlandsJoined September 2011
Jacco Bikker’s Tweets
for(float i,e,g,s,u;i++<1e2;o+=.01/exp(e*9.+vec4(1,3,4,4)*.1*s)){vec3 p=vec3((FC.xy-.5*r)/r.y*g,g-6.);p.zx*=rotate2D(t*.5);e=s=4.;for(int j;j++<12;p.xy=1.-abs(p.xy+p.z*sin(t+p.z)*.1))s/=u=dot(p,p),p/=-u,p.y=-p.y,p.xy=abs(p.xy+.8),e=min(e,p.y/s+.1/s*sin(t*.5));g+=e;}
#つぶやきGLSL
2
32
247
Show this thread
Hey I work there! Hey I work there too! - Lots of familiar faces and a cool topic. :) Great to have all this together in Breda.
Quote Tweet
Today we start with the second edition of the Rusterizer masterclass @BuasGames where we build a software rasterizer in Rust, with supervision and lectures by colleagues from @TraverseBreda and @EmbarkStudios
1
1
10
I always assumed that this would change the scale, but it doesn't, as simply proven by Tom. Could also be great for voxels perhaps. One thing I'm wondering: it's going to give you a different result than a normal rotation, but why and what's the nature of the difference?
Quote Tweet
Rotation of sprites using three shears.
cohost.org/tomforsyth/pos
Show this thread
3
Proud to serve as paper chair, together with AMD's Christiaan Gribble. Looking forward to seeing you in real life in Delft!
Quote Tweet
High Performance Graphics is coming to Delft, Netherlands, on June 26-28, 2023 (co-located with @EGSympRendering) Virtual attendance details TBD.
Paper abstracts: April 12
Paper submissions: April 17
Poster submissions: May 15
More info: highperformancegraphics.org/2023/call-for-
See you at HPG!
3
3
38
My work-from-home place. So cosy.
8
2
82
By the way, in a little voxel engine this would make for some lovely content. Hmm.. It's almost weekend. :)
1
Show this thread
And this one is for my poor Y1 IGAD students working on procedural stuff for their tank game on the #raspberry pi4. :)
Quote Tweet
1
7
Show this thread
As someone who actually bought computers in the 80s (grandpa's money ofc!), I feel this in my stomach. Pure excitement.
Quote Tweet
8
🧐 I'm honored!
Quote Tweet
Next week, we will read and discuss Jacco Bikker's lecture slides to try to understand the ReSTIR algorithm
meetup.com/graphics-progr
18
Retweeting a small indie procgen game: for just 5 euros you buy some labor of love, with a vast world and a great atmosphere.
Quote Tweet
Replying to @protopop
This looks and runs incredible. The procedural generation is awesome and a big thumbs up for proper (non SSR) reflections :)
6
for(float i,g,e,R,S;i++<1e2;o.rgb+=hsv(.4-.02/R,(e=max(e*R*1e4,.7)),.03/exp(e))){S=1.;vec3 p=vec3((FC.xy/r-.5)*g,g-.3)-i/2e5;p.yz*=rotate2D(.3);for(p=vec3(log(R=length(p))-t,e=asin(-p.z/R)-.1/R,atan(p.x,p.y)*3.);S<1e2;S+=S)e+=pow(abs(dot(sin(p.yxz*S),cos(p*S))),.2)/S;g+=e*R*.1;}
123
1,774
11.6K
Show this thread
Indeed, and after that the machine felt 10x faster but it wasn't but you would still make a fresh image with Norton which you never restored from. Good times.
Quote Tweet
15
Some WIP with a new scene with moving lights. Only direct lighting here. Also no denoiser, only DLSS and RestirDI. I use light tree bvh aconty.com/pdf/many-light and rebuild it every frame on cpu (debug draw lines in the video). It gives pretty decent quality even without Restir
11
44
488
Show this thread
Randomly assign each pixel to an incidence angle, bake it, and lights up the pixel only when it aligns with the view direction...
It actually works but not practical to do🤣
#b3d #procedural
7
19
315
Stream compaction is a useful technique to remove unnecessary work from data processed in a shader, I wrote a post describing how it can be implemented with wave intrinsics using a snippet from FidelityFX's SSSR as an example. Merry Christmas everyone!
4
50
253
Custom realtime compute shader global illumination pathtracer, still a heavy WIP but I need more ideas for stuff to add to it
All up on my github
github.com/Pjbomb2/Realti
#madewithunity #unity3d #pathtracing
2
17
89
Show this thread
Some random pictures from assignment 2 ("BVH") of Advanced Graphics (Utrecht University). I didn't specifically ask for 'pretty pictures', but will do so for the final assignment. :)
3
6
75
Graphics Programming weekly - Issue 267 - December 25th, 2022 jendrikillner.com/post/graphics-
3
58
306
This Sega Saturn homebrew demo is unreal. Seriously, it is Unreal. XL2 , Corvusdeux and Ponut sent the first two levels of this port to the SegaXtreme homebrew contest. A 1998 PC game on the 1994 console, no RAM cart. Custom engine, no Unreal source code.
segaxtreme.net/resources/irre
91
844
2,929
Baking a 3D scene onto a plane using hexagonally packed... panoramas
8192x8192 image texture with ~50k spheres on it to create this 😂
#b3d
44
402
3,437
Show this thread
I smell an excellent excuse for path tracing at a lower resolution. ;)
Quote Tweet
Replying to @aeriflame @evolutionalgd and @ThKasparrr
wait, I wrote a PAL/NTSC filter? :)
(I guess you mean this? github.com/LMP88959/NTSC-, I just posted a link on the 'other' social media site, it's not my work)
1
6
Two students (Idan Grady, Jaimy Kwa, with permission) requested help for their BVH builder, which was producing 100% overlapping AABBs for each split. Of course this happens for two tris in a quad, but they are using 18 triangles... What could be wrong? Then they showed me this:
6
3
33
Call for speakers! If you have anything in the area of #procedural content generation for games to share for our conference on April 21st? Please let us know or use this form: forms.gle/MzbGds6KqCcVsH
#procgen #generativeart #creativecoding #gamedev #indiedevs
13
24
Now why would you use Rust for that. ;)
Quote Tweet
Largely based on jacco.ompf2.com/2022/04/21/how
Show this thread
1
4
Graphics Programming weekly - Issue 266 - December 18th, 2022 jendrikillner.com/post/graphics-
1
61
292
👀 > fastgltf
A superfast SIMD powered glTF 2.0 parser written in C++17 with minimal dependencies
3
28
130
My previous tweet may not have worked out as intended for obvious reasons. Here's another attempt: The boss of this ship needs your opinion on something but certain decisions of the past may make it hard to act. You know what to do.
1
1
8
Musk wants to know if he should resign and promises to honor the outcome of the poll.
Perhaps you blocked his account, "Musk", "Elon Musk", "Musketeers" (sorry famous SF writer) and others; you can temporarily undo this in settings -> "Muted Words".
Poll:
1
1
If you haven't watched it yet, you really should watch 's continuity of splines video. This info is scattered like dust in the wind otherwise and really hard to get such a good holistic view of. Very understandable too.
1
48
235
Show this thread
I've had a few questions about SW rasterization recently, so I'm retweeting this, as it links to the Dr.Dobbs articles by legend Michael Abrash (of Quake 1 optimization fame!).
Quote Tweet
Replying to @pikuma
next step, Michael Abrash rasterizer cs.cmu.edu/afs/cs/academi
1
4
16
A real-time ray tracer based on ReSTIR-DI and ReSTIR-GI as the final project for CIS-565.
github.com/IwakuraRein/CI
3
26
164
#GPUOpen published bits of my work on real-time denoising : bit.ly/3PwGTvV. If you have any question, feel free to ask me!
5
35
162