I now its been said a million times ... but SOL and Shader Patch has literally given us AC 2.0 Pretty rewarding seeing your work with these two 'supermods' installed huh? ; - )
I think its great. however it lacks any dynamic range in shadows and lit areas and just removing light is bit naff all round. shame its not transitioned into a 2nd light source like a moon. and a new sky colour. I think 3 pp modes would work nicely for twilight, day and night.
hmmm. actually, after bumping the magic multipliers up. and re tweaking everything im getting interested again. perseverance is key . just put up with my tantrums and we`ll get there just fine . to me this looks nice regards a re balance.
no, was using BOH values, I think sol + CM disappeared them. same happened with trees before this test. and well I just went with it. I just love my trees now for some reason. wasn't doing them properly before (not that they are proper now-no idea)
yeah that boh figure is very destructive now. midday light no lens dof. THE GRASS RETURNS ! and now I feel completely different about the whole thing. /schitzo off
It's looking great @Pixelchaser ! I might have orthophotos, and this LiDAR and DEM/DSM stuff figured out, but still have a long way to go to achieve this type of result with attention to such small details. This track and Kyalami inspire me and shows what can be done You got me thinking about burnouts. And not the good type with tyres. I'm trying to avoid it by using GitLab to manage tasks/issues and assign those to milestones in a project. That way I can have a 'definition of done'. I think I need to otherwise I'll never finish my tracks...
cheers , its gona work out now I think. this basis I have now will stick and just tweak some art and colours. the photography and painting (and pixel chasing) ive done lately alongside studying old dead masters art has allowed me to understand something more than just re creating a reality. there is this thing with colour science in cameras more notably with my canon type, and after studying these colours I can instantly set the science in the scene in the game. ive also worked out how to get that convincing sunlit neon hyper green coloured grass that comes through. just not sure whether representing a world through a lens is correct. but its not like im anywhere near Lime rock park to really know its look. . in painting we use a colour wheel a lot to understand bounced light and what tone shadows and highlights goto without greyscale shading. so ive been applying that in terms of objects further and closer away through whats called atmospheric perspective. im actually going to take the trees even further to represent atmospheric banding where possible in the scene. in the last screenshot we have 4 groups of trees. dark, light dark light. the two light sets are actually at different depths to the camera, but its not apparent. coz the fog shader is shit. but the point is we can make that happen if we treat it like a painting would be treated or like the camera would capture. I may even building this in manually to all the mip mapping. \o/ motivated again I think.... until the next fall. haha. edit. /schizo on ..............now im thinking if its possible to build in a lens bokeh effect to mip maps for the shallow depths of field. in the forested screenshot we have the car in focus, but theres not much range behind it... I wonder...….
the wonder continues...… but no more I think. this is a simple test to see if I can manage the alpha transmission as the object goes out of focus. and by golly gosh I think ive done it. this is the alpha in action on a de focused mip map. (exaggerated effect) ill manually build these mip maps now and introduce a little of the atmosphere perspective too. if I increase the mip map counts too I can hopefully simulate a high f-stop figure into it.
I have a deep quandry regards mip maps and per pixel per metre visuals. I cant decide where the original texture should start in scene. turns out the majority of the scene uses about mip map 2 and 3 (of 4) unless you get very close of course. so that mip map 2 is half the size of the original which means I can actually half the size of my texture. and see more mip map range. its a good idea to bake a block colour texture change for every material to test all mip mapping at the same time. also tells you when your texture sizes and pixel per metre scales are appropriate or not and how all the objects gel together visually and how their detail diminishes in unison or not... im sure ill be deemed as crazy yet again with this type of stuff
the mip maps. mip map view when textures are too large. 4096 half the size of resolution you see more red which is the best quality mip (original texture). 2048 proof bigger resolution textures are not always better.
my own mip map filter. including brightening for atmosphere and a very little focal blur. and too much sharpening from mip 2 onwards. oops! shit works although subtle enough ! might need more focus I think especially the last mip….. not exact same, but old shot and new shot with the goodies. and half size texture than before.
but isn't the "original" in the 2048 the seconds mip stage in the 4092 texture? So red in high res is the same as yellow in the lower res?
I think that's right acc. im seeing it as that having too massive res means you just need to get closer to see it... when you move around in the editor getting the red to appear is actually very hard unless you are stupidly close to it. might need to test less and more mips too. the res probably isn't the key issue, maybe its the amount of mips??
been practicing using the brenizer method of photography to see if I can make the ultimate texture . way way way beyond normal levels of detail of course. the plan is to make a texture pack of 4k textures that could work with pbr shaders or be adapted elsewhere.... anywhere. but yeah this camera is the superstar (bought another 3 lenses ) https://en.wikipedia.org/wiki/Brenizer_Method
Yes, I believe this project has stopped. Pixelchaser said in a post a page back he’s moving on to other things.
I'm always sadly amazed and puzzled in equal measures when projects that are deep into development just suddenly get abandoned. Can they not be passed on to others for completion? It seems such a shame to waste all of that work.
