Loading video...

Video Failed to Load

Go Home

Testing a silly (or genius?) idea - accumulate custom shadow maps in a buffer to make some sort of Global AO. The implementation or the sample pattern could be better of course, this is just a rough sketch. It's pretty fast too, since you can do just one scene...

12,418 views • 1 year ago •via X (Twitter)

10 Comments

Dmitry Karpukhin's profile picture
Dmitry Karpukhin1 year ago

The inspiration came from here:

Vladimir Cazacu | 𝕧𝕠𝕝𝕠𝕕▿'s profile picture
Vladimir Cazacu | 𝕧𝕠𝕝𝕠𝕕▿1 year ago

+1 vote from me. SSAO with Lumen is too unstable.

Jean-Baptiste | Dev's profile picture
Jean-Baptiste | Dev1 year ago

After mega lights, mega shadows

Vini Cortez's profile picture
Vini Cortez1 year ago

that's very interesting! with some artistic controls can be very usefull

𝗧𝘂𝗳𝗮𝗻 𝗛𝗮𝗻 🇹🇷's profile picture
𝗧𝘂𝗳𝗮𝗻 𝗛𝗮𝗻 🇹🇷1 year ago

I think it is brilliant because simple. This AO is basically shadow from multiple angles and smooth out that, ta da basic GTAO :D

Gustav Sterbrant's profile picture
Gustav Sterbrant1 year ago

Rendering a top down shadow map for sky occlusion isn’t a bad idea .

stamisme's profile picture
stamisme1 year ago

not really that good for indoor areas since this isnt actual AO just use DFAO ‼️‼️‼️‼️

Dmitry Karpukhin's profile picture
Dmitry Karpukhin1 year ago

With native hard coded sample count (9 cone traces, I believe) It's not giving good results if you need occlusion in buildings, where the roof is higher than 10-20 meters :)

Nicholas Chapman's profile picture
Nicholas Chapman1 year ago

Should work for AO. The problem is AO by itself is not that useful. You want indirect lighting, and to know what part of the hemisphere is occluded, ideally.

Dmitry Karpukhin's profile picture
Dmitry Karpukhin1 year ago

Ideally, yes, but suppose you have a game where dynamic global AO artistically matters more than bounce lighting, and if this is that cheap, I thought it could be worth to try :)

Related Videos

A closer look on the GSAP powered animations in Webflow. Just from my perspective. In short, it’s way more powerful than I first thought. The more I use it, the more I like it. Of course, there are some parts missing. But it's the first version. There's still a lot to come. For example, custom ease would already make a huge difference. Breakpoints. Maybe a way to add custom functions as sometimes you need to calculate stuff in JavaScript to get complex animations. The biggest part for me is that you can just build the animation inside a timeline. This makes it so much easier than writing it as code. You can fine tune it so much more. When there would be a way to export the tween as code to add some more functionality would be awesome. You can also add custom classes with the timeline, which can then be used in custom CSS to add more functionality. With a bit of custom code you can also use breakpoints for your animations. By using the custom event the animation can be used in custom code which can also be pretty powerful. Didn’t test it so much yet, but would be cool if you could just take the event and use for example the onComplete() from GSAP or other stuff. For example. let mm = gsap.matchMedia(); const module = Webflow.require("ix3"); module.ready().then(() => { mm.add( { isMobile: "(max-width: 992px)", }, (context) => { let { isMobile } = context.conditions; if (isMobile) { module.emit("Mobile Animation"); } } ); }); "Mobile Animation" is the name of your custom event. I hope that part will get more powerful. Something I miss on the timeline is that when you select more actions you can just drag all actions. It would be cool if you could move it frame by frame with the arrow keys, like in Jitter for example. Or just write a specific point like when you have one selected. For now I would further write the code myself, but I’m excited to see where it’s going. Maybe there will be a time where you could animate most parts inside Webflow 🙂 Can’t wait for the next GSAP update!

Eduard Bodak

12,590 views • 10 months ago