Kenny Paulnocchio

The work I'm doing right now isn't really anywhere close to an interesting visual state but here are some of my favorite reduction fired bowls from the last year. #ceramics #pottery

Cool to see more of your videos popping up! They were probably some of the most influential content that got me into shader programming back in the day. My first big project at my last job was actually implementing better compression options for SDF based Glyph textures.

My dad's been driving his '99 ranger for 16 years now and every time he asks his mechanic what he should look for as an upgrade they tell him it's irreplaceable and he couldn't possibly do better.

smoky mountain finds

#minolta X-500
MD W.Rokkor 35mm f2.8 with extension tube
on Vision3 250D

#believeinfilm #filmphotography

I need to have harsh words with whichever Bloodborne fan got their hands on an exotic, wish granting relic about both specificity and the trappings of irony.

Multi room masking now (I only have the one demo room I made last week 🥴). I also tweaked the projection so the same wall always faces the back and the texture only uses 3 walls.

Reworked my UV math to get texture projection working properly and finally working with demo textures. Next up masking for randomly assigning room textures. Also maybe a new system for rendering out interior textures for more convincing projections.

With the Lies of P DLC on the horizon I'd like to see more Korean and Italian collaborations. This is the only one I'm aware of and obviously the countries had some synergy that puts them on a whole new level.

Honestly as fun as it is to build out flashy effects as custom nodes, I need to fill out my library with more simple point to planes or custom blends. I get way more mileage out of a quick and dirty operation that saves me 3 nodes than some whitepaper I copied into node form.

I have all the math in place for changing room scale and I can source separate textures for floor/walls/ceiling or tile them into a single texture file. It's tempting to add fine grained controls for everything but part of the appeal of using prebuilt systems is that they give you a workflow.

Still writing godot shader node addons. Workshopping a parallax interior mapping setup now. I have to figure out how much I want to make it extensible vs how much behavior I want to enforce which is an interesting tradeoff.

simplex noise mask versus height mask. I wrote my first height based mask shader almost half a decade ago but it's still so satisfying it'll never get old.

Ever since I started working on projects with others, if a shader function gets messy I just write custom node for it. Not everything can be a uniform and if someone else has to hop in my nodes it's way nicer if it's a self explanatory chain of node functions rather than a spaghetti nest.

Playing with the controls on a custom height blend node in #Godot visual shaders. This one is a simple heightblend that generates a blend mask from 2 height textures and some fine tuning controls.

Are you aware of any projects that use the index drivers as standalone headphones? I found a reddit post where someone pulled off a pretty good looking build but audio related stuff is uncharted waters for me so I don't know much of that space.

Fun to see the Jeff awards cycle up this year since I'm in the credits for a nominee. A little bittersweet when they laid me off this summer but now I can follow the results as a dispassionate observer.

Hmmm every time I try to show the specifics of something I'm working on low frame rates or compression kind of obscure the details. I suppose posting small visual details in engine is an art form.

Rendering 3D objects with a pixelart shader. I could write a custom script using a looping timer to update the render target every iteration or I could set an auto looping animation player to set the render_target_update_mode to once at the beginning and set the duration to the desired FPS.

If I wanted to really illustrate this I'd have uploaded video but last time I tried that compression crunched everything way past the noticeable point for scaling artifacts so people will have to trust me that it looks way better lol.

I'll have to move some depth sorting code and rewrite a couple things but my biggest grip with bilinear texture upscaling on post processed volumetrics is the obvious artifacts on boundaries. FSR gives much cleaner edges and seems well suited towards the cloud usecase but more testing is required.

I've been thinking about optimizations for my #godot clouds for a while. Currently experimenting with FSR upscaling for a cloud viewport I write transmittance and depth data to, and then do actual depth sorting and coloring when I pull that viewport texture. Below Native, Bilinear Filtering and FSR.

I've made a couple. I'm not the fanciest material artist but I just used shape and transform nodes to block out masks and then was off to the races.

Honestly I've used material maker so much designer grates on me now. I know there are artist pumping out textures I couldn't dream of with workflows beyond my comprehension in substance but MM is just so elegant and powerful in the ways I want it to be I can't help but gas it at every opportunity.

Still chipping away at #Nodevember2024 I'm doing like 3 other challenges this month with friends but I came back for the dwarf forge prompt because I love dwarves and I love #materialmaker. I just need to find out how people make their graphs not a horrible tangle.

I always go through the rear wheel and seatpost for my ulock. In my head I could see getting enough play to pull the wheel out the dropouts and contort things through the stays but I've never tried and I could be fooling myself.

Simondev's video on clouds and the guerilla games whitepaper were super helpful for this project. I wanted very stylized puffy clouds but I started from a base of realism with modelled transmittance, scattering, and absorbtion and then fine tuned it for performance and a more stylized look.

Finally dug out an animated screencap of some volumetric clouds I built in #godot This was my first attempt at raymarching anything other than distance fields. It's still not 100% where I want it for performance but I'm happy with the visuals

My first foray from purely visual shader nodes into glsl was building a custom gerstner wave node for godot 3.2. Gerstner waves are complicated enough and require enough iterations that without a subgraph system I didn't want to think about the rat's nest doing it with only nodes would incur.

no more raymarching in expression nodes going forward. I finally reread the custom node docs.

Nodevember exploit. "Why yes this was all built in a visual shader editor."