Pocster

Well! Been avoiding TAA as the visual artefacts annoy me and I want to get the best before I pull the big performance lever. But after some other smaller optimising this is the route we're going. Most major (if not all AAA ) games do this so it was always inevitable. 60fps is coming back!

Ok! Well Fable really is doing better than Opus 5 on high. Getting right down to the nitty gritty and understanding it. Not quite there yet but big optimisations on the way for trees and grass I think.

stuck. pushed the turbo "fable" button.... here's hoping...

Just a test so 1024 trees. All real meshes. Nearer ones move per branch with wind and have higher fidelity shadows. Take lighting from sun/cloud/terrain/atmosphere (froxel,fog etc etc). Lots to compute!. Been rethinking lots and applying every trick I can to get frame rate sane but still heavy. I'm now onto placement - so got to go and plant 10'000's of trees!

All day at it. Claude's now struggling to find optimisations. So I need to be extra creative on the method rather than the code. 1024 trees on screen kills frame rate. Tricks a plenty needed

Claude's funny. We can go on for literally hours and not resolve issues and then it says "for whom picks this up....". I take that as a hint!. Open a new session same model same level and off we go again. You would assume bigger context would give better reasoning on a problem but sometimes not it would seem. I let Claude now use "auto" so it can roam more files without approval. Literally just churns away... I have to stop it and ask "WTF are you doing?" - frequently its gone down a pointless rabbit hole. I do swear it has its own agenda within mine!

Not 100% correct yet but the detail on a procedural tree can be insane. Snow accumulates and bends individual branches. At a certain point branch will shed the snow and 'flip' back. Intend to do similar with rain i.e. rain accumulates so even when rain has stopped branches/foliage can still drip.

Got lakes to freeze now. Doing this brought up a few long term bugs so was a bit more work than I thought. Avalon 1 did have a basic procedural flora generator. This has since been improved in Avalon 2. Texture wise took basically every free terrain/plant texture from polyhaven!. Alien plants will of course be procedural. Masses of optimisations again. Spent an entire day trying to do bubbling mud for volcanic worlds - still not right. Claude likes to overcomplicate on an epic scale; our "camera" had 127 parameters! - now just 8. Tree lab still has around 80 variables to "grow" the plant (inspired by what Unreal Engine does). Now just main ones on first screen. As for bubbling mud it somehow did an entire advanced physics simulation causing some chronic frame rate issues as you can imagine. It goes off on it's own little adventures I'm sure!

Claude's (expletive deleted)ing thick today!

Sunsets do it for me!. Been working on "real" rainbows. Most engines cheat at this but mine's created correctly (I think!). Now I need the water to freeze over time. GPU burden has been creeping up again as expected. So optimisations required yet again. Lightening thunder storms are extremely impressive and expensive!

Finally got decent looking snow. It accumulates on terrain and thaws.

Tell you what - that's a massive understatement. What I mean is I've created a custom 3d engine that would be commercial grade - yet not 1 line of code. That is the real power of AI. I see plenty of youtube/reddit "wrote this in 5 hours with Fable from 1 prompt". Some are believable some are slop. But most are relatively simple (still impressive that Ai can take "write me a really good minecraft game with dinosaurs in it" and still produce something). But use it properly from experience and build it up layer by layer rather than the final product from 1 prompt and its a phenomenal tool. The accomplishment here is WELL beyond my technical ability. But the "difficult" parts the code and horrific maths aren't my burden 99% of the time. I'm still in awe of what's possible. Must also admit it can take some effort - occasionally days on 1 feature to be correct. Or bastard bugs... we had a few mountain tops going black as you approach. No obvious reason some would and some wouldn't. Check all the render stuff/textures/lighting everything me and Claude could think of. Nothing obvious. Found it out of frustration!. The ray march on water can shoot a ray so far that it can hit a mountain top. Essentially at that point its a NaN so plots black! - what a bast!. Water affecting a mountain maybe 4km away in real space. That kind of bug finding without Claude (as I've done many a time when a software engineer) its pure torture and can literally take weeks to find...

That was an esp32 project and works!. I'll integrate that into Avalon later. One thing at a time! ( well lots of things at lots of times really - so never get bored!)

Theres always ways to optimise. My main lever is of course the shaders (even though Claude wrote them) and pre-baking - 256Gb to burn so we create lots once at runtime to use later and save on some computations. The original premise is still there - just got a bit distracted. But yes the Alex/Chatgpt from Avalon 1 will be imported into this. The robot running over these luscious landscapes will tell you when your chicken is ready and then you're playing Ed Sheeran AGAIN!

Amateur ! Time to analyse shaders ! Optimise 💪