Go to bottom

Random "work in progress" shots

category: general [glöplog]
An early area lighting test, which evolved into the "barcode" scene in my 16k.

BB Image

both up on shadertoy now: proto, partyver :)
added on the 2018-04-03 18:26:31 by yx yx
@yx: your intro was super cool, please keep up the good work!
added on the 2018-04-05 09:58:49 by deepr deepr
BB Image
added on the 2018-04-05 23:34:53 by Blueberry Blueberry
first invertebrates sketches, made during the holidays after pondering about the ants crawling our balcony

BB Image
added on the 2018-04-06 14:24:01 by Steltek Steltek
So here are some wip shots of lightfields when I tried to get the rendering good and fast enough. In the end I ofc ended up cutting out some of the things for the sake of space. :P

BB Image
BB Image
added on the 2018-04-06 23:00:38 by rimina rimina
A little demo framework I'm working on.
Its uses the code base of my assignment for the computer graphics course (that used the fixed pipeline).
It supports catmul rom curves (that can be used as paths for objects and cameras), and the scene is described on a XML file.

However going from a fixed pipeline to a programmable one is a bit harder than I'm expecting.
(And I'm really resisting the urge to re-write everything - no wonder I cant get any demo done).
Gotta pay a visit to the wanted.scene.org to get a hand......

BB Image
added on the 2018-04-07 14:27:01 by Jae686 Jae686
This year I couldn't be in Revision so anyway I wanted to contribute a little more to this lovely thread by posting some WIP videos of discarded ideas for our only Architect's release ,"Where the dead things dwell" (http://www.pouet.net/prod.php?which=69691).

I am so sorry about the quality of the videos, they are terrible. This is due to the fact of I had a very crappy GPU at the time and I wanted them to be captured in RT as sending these videos were our first way of tech debate at the time.

First attemp on consistent UVs with rotating raymarched objects.

Testing some structure patterns. Were finally discarded. I dunno whether it was a good idea or not :\

Daylight tests with some wood and concrete materials.

First attempt on emissive materials with real-time indirect lighting.

First attemp on first intro scene, fake neon lights, DOF and emissive materials with real-time indirect lighting]. Finally discarded, tubes were used only in the last scene (the woody path), emissive materials on the scene before the last one (the hexagon field), and the tunnel anyway was somewhat used at the begining.

Test on real-time indirect lighting. Just used to refine the overall indirect lighting algorithm used in the intro. Was intented to be a scene but I didn't manage to make it visually appealing at all.

Testing the deferred lighting pipeline. The final intro had 22 actual lights simultaneously rendering, 2 of them being shadow casters.

Hope you enjoy them!
added on the 2018-04-09 01:37:30 by merry merry
Thanks, COOL! ;)
just some more old records /for the fun/
Not my balls!
Don't mess with SubZero!
More serious stuff maybe later. I usually don't record that much... like the next generations. :P
added on the 2018-04-11 19:20:35 by aqu aqu
added on the 2018-04-11 19:22:20 by ferris ferris
fizzer: that would make the intro more scary (and cursed).
added on the 2018-04-15 13:47:20 by Flashira Flashira
early subsurface radiosity tests for http://www.pouet.net/prod.php?which=75806 (Autogenic by Nuance)

BB Image
BB Image
added on the 2018-04-17 14:08:03 by pro pro
BB Image
BB Image
added on the 2018-04-17 14:09:12 by pro pro
added on the 2018-04-17 14:42:33 by pohar pohar
pro: looks great! whats the technique?
added on the 2018-04-17 21:08:13 by smash smash
Smash: Thanks! It actually turned out to be a quick trick in the end. Just derive the object thickness by either baking it (i think substance painter and 3d coat can do that) or during runtime in screen space and use that to 'modulate' your emissive map for the object. Then drive this emission through your favourite radiosity thingy and make sure you allow self-illumination. i used a screenspace approach for that.

what's also nice when modulating the emissive map is taking the fresnel incident into account. this almost eliminates the need for thickness in round objects.

The dog has thickness prebaked, the pyramid was calculated during runtime.

we wanted to show this effect a bit more in the demo, but since the main guy dropped out for personal reasons cosmic and i had to put something together in a few days. you can see the effect in a limited way e.g. in the sphere, the drone and the pyramids.

here's another example from 2 years ago when i used the tech without thickness:
BB Image
(But the demo itself sucks since i had to do all the modelling and dancing on my own and i'm not particulary an expert in either of that.)

added on the 2018-04-18 10:17:48 by pro pro
Is it just me or the reflections look wrong here?

BB Image
added on the 2018-04-18 10:44:45 by Gargaj Gargaj
it is.
added on the 2018-04-18 10:56:01 by pro pro
wrong :D
added on the 2018-04-18 11:02:34 by pro pro
Turns out metaballs on an FPGA are harder than I thought! :P

BB Image
added on the 2018-04-18 15:42:22 by oni_ oni_

"i think substance painter and 3d coat "

What is this ?
added on the 2018-04-18 16:08:00 by Navis Navis
Substance Painter is a semi-procedural non-linear texturing tool, super useful; I highly recommend it: https://www.youtube.com/watch?v=Qk_YRkDsJrQ
3D-Coat is more of a sculpting/topology management software, also quite useful.
added on the 2018-04-18 16:14:29 by Gargaj Gargaj
wow good thanks. I'll check it out
added on the 2018-04-18 16:33:30 by Navis Navis
BB Image

BB Image
added on the 2018-04-18 17:15:18 by superplek superplek
BB Image
added on the 2018-04-22 22:56:45 by rudi rudi


Go to top