• Dear visitors,

    The email issue has been finally solved.
    Thank you for your patience and happy browsing.

    Team ACM.

W.I.P. Lime Rock

Prototype

Well-Known Member
Every time I see that grass it puts a smile on my face.

Guard rails look good man, looking at these makes me want to redo mine.
Can you post up a wireframe of your model? Curious as to how much geo you're using here.
I've been super conservative, and its showing...
 

Pixelchaser

Well-Known Member
hehe, I have just went with it. the problem with the engine is too much geometry here causes the shadow aliasing problems. but a little bend in the main plane seems ok. some use a flat back to save polys. im just gong for it and will make lower stuff if needed. I built the model around the texture. the back side is a conveniently flipped version of the front which works nicely. with all those panels to choose from uvw xform was my best friend....:lol:


.
 
Last edited:

Prototype

Well-Known Member
Gotcha. Thanks for the reply.

Ok geo wise (top to bottom) mine looks pretty much identical. My texture map looks similar to yours too ... (maybe a touch not as good as yours), so I figure Im falling short is in the shader department... this is my weakest link currently.

Going to pick your guys brains on that in the days to come ...
 

Pixelchaser

Well-Known Member
ive not done any major shader tweaks yet but I plan to do some colour grading against photography. here, this is the most basic shader. ks per pixel, its the most standard shader, alpha channel for specular. I think the tree shader or grass shader turns off self shadowing so ive still to try that actually.

my texture is 4096 long btw. my goal was to cover everything with atleast 50 pixels per metre resolution for all upright objects. road gets 100, and detail channels atleast 1000 if not 2000. and the grass gets 1000 ppm too... ive made everything from scale texturing as a base. id like to think this attention is what brings that energy you speak of... in this engine art gets you far, shaders can actually ruin things because they`re not very good. hence what we have to do with normals for grass and trees. start bouncing too much light of the barrier and make it too dynamic and it just falls apart. I thnk while my barrier looks good there, its probably just a nice screenshot so far.
 
Last edited:

Johnr777

Moderator
I used the ksgrass shader for Mosport’s guardrails, got the idea from LilSKi. The variation layer really helps make it look less repetitive, and the no self shadow was probably a good thing :lol:
 

Johnr777

Moderator
You get one extra slot to load a texture to. Then you can control how much it blends with the diffuse and how much coverage.

Was meant for grass, but we have found use for it in tree lines, guardrail, etc
 

Pixelchaser

Well-Known Member
more rails. heres the start of this hedge kudzo stuff.
morerails.jpg


hehe , to me that looks just nice :), this pseudo 3d stuff which is like the barriers just a bit 3d is actually better. I mean the basic plane floating in and out the fence. I think with just a few leaves scattered in the protruding areas manually will nail its necessity to have volume. but just on the inside. and top.
32kudzo.jpg

and some extra width, theres 2 hedgey fences at lrp. this one doesn't have the kudzo so much.
3dkudz.jpg


not heard from fabz for couple of weeks now. :/

needs some billboard adverts any ideas ?
 
Last edited:

Prototype

Well-Known Member
Makes a big impact that foliage variation. Nice.
Looking forward to seeing the net result of all these detailed aspects when they're put together.
 

Pixelchaser

Well-Known Member
Makes a big impact that foliage variation. Nice.
Looking forward to seeing the net result of all these detailed aspects when they're put together.
cheers. I cant wait to see it too. haha you never know. it might fall apart. you know it may not be liked by everyone. and many wont like it before they have even experienced it too.. but im happy to atleast have some stamp of difference on it where I at least tried to bring new ideas to the table rather do it the usual way with the environments. id like someone to see my track and think that's a pixel chaser track. isnt that an artists job ? i know people hate me for this. but i wont stop. but for that it takes more than 1 track :lol: ! i have another lidar jobby in the works and everything was made with that in mind too. but i really need to get some architectural modelling learned before any more.

another thing is id like fabz to decide on final looks etc too. I might just be missing stuff here im not aware of. but I think im pretty good. certainly its been the best last 4 months ive had in about 5 years personally and creatively. actually just pushing my ideas without cogitating has been just great alongside my crappy shift work job. didnt realise how bad my depression was through starting all this. thankfully it seems its really behind me now. so there aint no stopping me now :lol:.
 
Last edited:

barf

Member
thanks also Pixelchaser for documenting your adventure! it helps new people like me to learn :nerd: will surely be referring to your posts when i get around to installing armco!

and glad to hear you're improving! :cool: i struggle with depression sometimes too, my project helps me and it feels good when others enjoy your handiwork!
 

Pixelchaser

Well-Known Member
Grass looks a bit flat in color over distance. I would suggest ditching the flag shader and using the grass shader with variation.
I agree but only compared to every other track. compared to REAL LIFE that's exactly what happens when its "kept" grass. but yeah its quite clean. and ive been studying that exact thing at lrp in videos where possible as there are a lot of "long" cameras. you`d be surprised, its not that entirely incorrect. the previous grass model whioch was double the stuff going on had the issue even more. the prolem came from the art though back then. but now it comes from the construction but now the effect can be changed with the artwork when I get to it.

, its not the flag shader. but it is the same result actually .. and its using the variation but that is matched to the grass detail texture. . at this stage (not finished) its an intrinsic effect of doing grass a bit differently and that its not finished.. but theres a few more things to introduce to it which will change it out at distance this. (using 3 of 4 atlas variations 2 of which nobody seen yet. and lawn mowed lines might be enough in the distance to break it up) and also there is no lodding on the grass yet.


the scatter method has its intrinsic fake qualities too such as it leaves there being to much variation in the distance due to the production method and the "tool" used in 3dmax. not to fail to mention that real grass compounds more because theres more of it. (Not 1.5 to 2 every metre like the game visualise with scatter method) and that's what my method does actually provide a new more realistic environmental distortion effect to grass over distance. but its not finished and I know exactly what im doing now, got the workflow down to perfection.. different doesn't equate to wrong but you need to try get along with it, and with me. you are the best track maker for Assetto and I have heeded your advice at all levels. I am just doing something different. and yes it still look a bit like Astro Turf technically but I like it. at this stage it may make a better football pitch but itll get better.
 
Last edited:

Pixelchaser

Well-Known Member
this is the everything combined, all that grass, all those trees. detailed barriers. im shocked it was as low as it was.
2polycount.jpg


so grass statistics time.
in comparison to the grass scatter method id like to point out some statistics. the recommended statistic for grass blades per metre according the night eye tutorial is for layer 5 1.5-2.0 blades per metre coverage. my scenery where the grass covers is 81000 metres square. 81000 x 1.5 = 121,500 and at 8 verts per blade that equates to 121,500 x 8 = 972,000 vertices just for layer 5. the lower layer have an exponential decrease but this still equated too 1.8 million vertices in total for scattered grass layers 1-5 . my layer here weighs 1,1 million vertices for grass every 20 cm and every 10 cm covering around the track and for the mowed lines

just the grass model here and statistics. can see the lower height model with texture colour variation for coloured lines. that's the part that's twice as thick at 10 cm. the mowed grass is 5 cm high physically.

3polycount.jpg


as far as im concerned its 60% full and the texture memory is at 42% too. buldings and spectators can get some nice treatment i think. and some custom low poly objects for shadow maps for all objects.

so no kslayers,4,3,2,1 etc... well i have a different method for dealing with that if i need to.
 
Last edited:

Pixelchaser

Well-Known Member
been playing with fabric animations today, to make the plastic covers that cover the tyre walls. will make normal maps from this method alongside some diffuse variants.. hopefully get them shiny in game. heres 100 variants :lol:
bvc-1.gif
 
Last edited:

Prototype

Well-Known Member
Nice update! Your grass stuff is still a little over my pay grade, but watching with expectant curiosity.

Are the plastic covers at this track over the tire walls that thin that it would be affected by wind? Its a cool effect though ...
 
Top