Best Of
Re: The Bi-Monthly Environment Art Challenge | January - February (100)
Greetings 
First Polycount post & challenge! I started on the 18th of Feburary and have gotten this far. I'm also looking to get some hand painting practise in with this project.
Not sure how much more I should push the geometry as I'd like to see what I can get away with via textures.
Not sure how much more I should push the geometry as I'd like to see what I can get away with via textures.
haid
3 ·
Re: What Are You Working On? (3D) 2026
Alright modeled most of what I wanted to include, I've also applied tri-planar pbr materials too 'previs' the object's silhouette since there's no need for UVs at this point. Anyway aim is to run it through substance painter using UE4 showcasing final renders so we'll see how things turnout once results are assessed.
sacboi
3 ·
Re: What Are You Working On? (3D) 2026
Extremely new to modeling.
Working with a Star Wars mod team to make models that go in game.
I wanted to make a blaster that fits in universe but not exactly one that would be in the shows or movies. I used a Kotor Blaster Pistol as the reference for the shape. It has to be mid tier poly for First Person views. Currently 8.5k tris and 4.5k vertices.
Not even close to some of the amazing work in this forum but I will keep trying!
Working with a Star Wars mod team to make models that go in game.
I wanted to make a blaster that fits in universe but not exactly one that would be in the shows or movies. I used a Kotor Blaster Pistol as the reference for the shape. It has to be mid tier poly for First Person views. Currently 8.5k tris and 4.5k vertices.
Not even close to some of the amazing work in this forum but I will keep trying!
Jerec
3 ·
Re: What Are You Working On? (3D) 2026
Currently chipping away a bit at this Dugong sculpture along with this Orangutan.
Still a heavy WIP , both of them.
Still a heavy WIP , both of them.
Francois_K
3 ·
Re: CAHDD™ (Computer Aided Human Designed & Developed) — voluntary workflow transparency + watermark too
" [...] it’s more that it’s already in the mix whether people like it or not"
Well, not only do creative people not like that it is in the mix but it also didn't take long at all for the tech to get a very bad rep in the broader public as well. And since anyone with a smartphone (from Linkedin middle-managers and AI-bros ... to any old grandma) can prompt img2img slop instantly, pretty much everyone can tell that "something looks AI" by now without the need for a watermark. Regular CG also gets mistakenly lumped in, so the challenge for people making a living from beauty renders isn't going to be about indicating which % of their images was "humanly art-directed" with a watermark or some footnotes, but rather to convince their clients and the public that they did any work at all ...
On the topic of archviz in particular, I was just chatting with someone whose previous job involved interfacing between architecture firms and archiv firms and she mentioned that there is a trend towards a return to renderings that are much less advanced and polished since realistic but overly polished and glamourous images end up being not quite representative of the actual final project and clients are fed up with this. Therefore, a renewed interest in a more simple or even sketched look.
TwoListen hit the nail on the head really : people who don't use AI can simply say they don't use it. IMHO It's up to the AI-bros so in love with their already tired yellow backlit images to not get left behind and adapt to the fact that their ship has sailed !
Lastly, the type of responses you are getting here is because this is a real-time gameart forum. That is to say : a line of CG that is about an order of magnitude more complex than archviz (no offense). People here know very well how tech works and aren't easily impressed.
Lastly, the type of responses you are getting here is because this is a real-time gameart forum. That is to say : a line of CG that is about an order of magnitude more complex than archviz (no offense). People here know very well how tech works and aren't easily impressed.
pior
3 ·
Re: Sketchbook: Frank Polygon
Subdivision sketch: Basic tubular frame intersections.
This is a visual quick reference for segment matching when joining offset cylinders. The cylinders in this example have different diameters but it still works when the diameters are the same. Note the subtle changes in segment count and loop routing as the cylinders get closer to overlapping.
For more detailed write-ups on the basics of segment matching on curved shape intersections, check out the following posts:
https://polycount.com/discussion/comment/2803910/#Comment_2803910For more detailed write-ups on the basics of segment matching on curved shape intersections, check out the following posts:
≈50-55% overlap. Larger cylinder has 28 segments and the smaller cylinder has 24 segments. The loop routing carries the edges across each intersecting shape in the same balanced pattern.
≈75-85% overlap. Larger cylinder has 30 segments and the smaller cylinder has 24 segments. Count on the larger cylinder increases to match the increased overlap and maintains the balanced loop routing pattern across the intersecting shape.
≈95-99% overlap. Larger cylinder has 28 segments and the smaller cylinder has 22 segments. Count decreases on the smaller cylinder to maintain segment alignment. Loop routing changes on one side so the edges turn up and run along the intersecting shape to prevent disrupting the segment spacing with additional geometry that would otherwise cause pinching artifacts.
This produces the same balanced topology layout when the smaller intersecting cylinder is truncated where it joins the larger shape. The overlap and segment counts here are the same as the previous example. Large cylinder 28 segments, small cylinder 22 segments.
View showing the other side of the shape intersection with the same balanced topology layout and clean loop path around the shape intersection.
A gentle reminder that the same basic topology used to join shapes also tends to work when subtracting shapes... The only real difference is you just have to imagine the shape goes inwards instead of outwards when searching for similar examples.
Recap: Keep things simple, no need to try and develop complex rules or formulas. Simply offset the intersecting shapes, match the intersecting segments by pairing them while counting, adjust the number of segments until the edges more or less line up, then join the shapes and use the space between the inner and outer support loops around the shape intersection to hide any differences between the surfaces of the intersecting curves.
FrankPolygon
5 ·
Re: LOW-POLY ART
Hey guys, nice seeing stuff still getting posted, i've been in this thread for 20 years XD
redmond_david
4 ·
480tris 256map