The BRAWL² Tournament Challenge has been announced!
It starts May 12, and ends Sept 12. Let's see what you got!
https://polycount.com/discussion/237047/the-brawl²-tournament
It starts May 12, and ends Sept 12. Let's see what you got!
https://polycount.com/discussion/237047/the-brawl²-tournament
Best Of
Learn to Draw in 100 Hours - Padlet Order List
Hi
Martina from "Nerdforge" created
a resource list of tutorials for learning to draw in 100 hours. They made a
video of her husband, Hansi, going through the tutorials in order and
documenting his progress.
I've made a Padlet if
anyone fancies giving it a go, your welcome 👍
https://padlet.com/littleclaude/learning-to-draw-8zhtxehblhgc6ort
I had a student many years ago who did 1-2 hours drawing per day and he become very good by the end of the year, here is his results - https://www.artstation.com/artwork/zAWVo2
littleclaude
3 ·
Re: People with Autism + ADHD innatentive who aren't medicated do you have any advice to be productive?
"Does it matter if I draw digitally or it's better with pencil and paper?"
IMHO you'll progress about 10x faster with paper, as digital is nowehere near as accurate and responsive. And you can always have a tiny dollar store sketchbook/notebook on you, as opposed to having to carry an absurdly expensive and fragile digital tablet.
"I know it sounds weird but I've never liked drawing. I think it has something to do with the fact that I know I am not good at it and so having to relearn to be good at something again I don't find it particularly fun. I will make an effort if it is required to be a great modeler."
Well, it all depends if you're in for the fun of instant gratification or if you're in in order to learn something. Learning something is not quick, requires effort, and it can suck at first. But you can't hope to become a designer without learning ... how to observe and design !
If anything I'd say that things start to click once one accepts that it will suck indeed, and for a long time - perhaps forever even. But the silver lining is that even after years of trying to improve, things do eventually improve ; and the little Eureka moments coming from finally understanding something new are always delightful. Like how this very experienced artist is thrilled to talk about the new approach he took in order to draw fishes from life.
https://www.youtube.com/watch?v=Uhx5HHoWCmQ
Now if you are mostly interested in precisely reproducing existing objects (like for instance making CAD recreations of firearms or vehicules) then that's a whole different story. But I don't tink that's the case. And going back to the video above (about the sketches of fishes), I am sure you can tell that this artist would be able to sculpt a badass dynamic-looking fish in no time (even without much prior knowledge of digital sculpting), whereas someone attempting to model a 3D fish without any intricate knowledge of the subject and working from whatever ref has been pinned on Pinterest would get much less interesting results.
On the flipside if you *do* want some more instant/fun gratification through modeling and progressing as a designer at the same time, then you might want to adopt a way more restrictive visual style. For instance voxel modeling, or retro-spec modeling under extremely tight constraints (PS1/Saturn-era). The very low triangle count, barebones texturing and absence of shading will act as strong limiters. Any lack of art fundamentals will still be apparent and hinder your progress ... but perhaps less so, and you'll be able to move on to the next project faster hence improve faster. The risk though is to end up stuck in "retro fandom hell", like the modelers who can only model celshaded furries or VRchat avatars and nothing else.
Hitting up a regular local sculpting class certainly wouldn't hurt either. Overall I'd say you simply have to be okay with the fact that knowing Blender for X amount of years really has no relevance to your progress as an artist. It simply means that you have a solid (and certainly valuable) technical knowledge about a specific medium.
IMHO you'll progress about 10x faster with paper, as digital is nowehere near as accurate and responsive. And you can always have a tiny dollar store sketchbook/notebook on you, as opposed to having to carry an absurdly expensive and fragile digital tablet.
"I know it sounds weird but I've never liked drawing. I think it has something to do with the fact that I know I am not good at it and so having to relearn to be good at something again I don't find it particularly fun. I will make an effort if it is required to be a great modeler."
Well, it all depends if you're in for the fun of instant gratification or if you're in in order to learn something. Learning something is not quick, requires effort, and it can suck at first. But you can't hope to become a designer without learning ... how to observe and design !
If anything I'd say that things start to click once one accepts that it will suck indeed, and for a long time - perhaps forever even. But the silver lining is that even after years of trying to improve, things do eventually improve ; and the little Eureka moments coming from finally understanding something new are always delightful. Like how this very experienced artist is thrilled to talk about the new approach he took in order to draw fishes from life.
https://www.youtube.com/watch?v=Uhx5HHoWCmQNow if you are mostly interested in precisely reproducing existing objects (like for instance making CAD recreations of firearms or vehicules) then that's a whole different story. But I don't tink that's the case. And going back to the video above (about the sketches of fishes), I am sure you can tell that this artist would be able to sculpt a badass dynamic-looking fish in no time (even without much prior knowledge of digital sculpting), whereas someone attempting to model a 3D fish without any intricate knowledge of the subject and working from whatever ref has been pinned on Pinterest would get much less interesting results.
On the flipside if you *do* want some more instant/fun gratification through modeling and progressing as a designer at the same time, then you might want to adopt a way more restrictive visual style. For instance voxel modeling, or retro-spec modeling under extremely tight constraints (PS1/Saturn-era). The very low triangle count, barebones texturing and absence of shading will act as strong limiters. Any lack of art fundamentals will still be apparent and hinder your progress ... but perhaps less so, and you'll be able to move on to the next project faster hence improve faster. The risk though is to end up stuck in "retro fandom hell", like the modelers who can only model celshaded furries or VRchat avatars and nothing else.
Hitting up a regular local sculpting class certainly wouldn't hurt either. Overall I'd say you simply have to be okay with the fact that knowing Blender for X amount of years really has no relevance to your progress as an artist. It simply means that you have a solid (and certainly valuable) technical knowledge about a specific medium.
pior
4 ·
Re: [WIP] Two-Story House - Environment
Back again!
Most of the work over the last 4 days happened on paper, not in-engine. I put together a rough prop list to populate the scene — about 50 core items, but many of them will have variations (like different tables and chairs). So yeah, I’ve basically signed myself up for a fun little adventure of modeling 150+ unique assets. Thankfully, most of them are pretty simple and small in shape and material.
As for in-engine progress — I identified some boring/empty areas and added points of interest. For example, there's now a wardrobe space right near the entrance
Some marked spots are still empty though — haven’t found a nice and logical idea for them yet.
I also started filling in the last room on the first floor — the bedroom
And as you can see, I threw some materials on the meshes to get a rough idea of the scene’s color palette.
Next up: time to dive into prop modeling. To speed things up, I’ll try to find schematics or solid reference images for each item — the less guessing, the better
adddistress
3 ·
Re: The Bi-Monthly Environment Art Challenge | July - August (97)
Very late to the party but decided I wanted to give it a try. 1000 tris and 1 texture set at 2k
Azlios
4 ·
Re: Early-Late 2000s CGI/Game textures
haha back in my day
This is the book I used for character art modelling.
This is the book I used for character art modelling.
HAWK12HT
5 ·
DeathStranding2 Sniper Rifle-- Need your critique to Learn
This is Tranq Sniper Rifle from deathstranding2. Basically I got lots of fusion360 modeling experience but this is my first attempt of rendering with Blender to get a clean pic. I've tried my best but I guess I will need others critique to improve. So I hope you can give me some advice, truly appreciate.
HOLLOVIA
3 ·
Re: 2D chain texture
@5rettski Observation and speculation can be informative but sometimes the best way to learn is to actually recreate the effect by starting from first principles. It may be helpful to break this problem down into it's constituent parts then work through solving each of them individually to build up an understanding of the entire process from start to finish.
Analysis of the original image: both the pendant and silver necklace chain float above the character's clothing and appear to be non-dimensional or "flat" when compared to the rest of the model. The way the structure of the necklace appears to twist as it comes up over the shoulder and the strong mipping on the texture there suggests that this is a poly strip (probably one sided) that has a basic tiling texture applied to it.
Since the textures on the necklace and pendant are see through and have a higher textel density than the rest of the clothing it's also quite likely that this part of the model has it's own texture sheet or larger portion of some other clothing / accessory sheet with an alpha channel.
These assumptions could be confirmed or proven by observing the model under different conditions in game, looking at how similar models were made in that era, or by viewing wire-frames of the model. If that's not possible or practical then the next best thing is to ask other artists how they think it was done and try to replicate the results based on whatever information can be found.
What's already been suggested: several experienced artists have already offered concise explanations of how things were likely done and provided links to similar examples and or tutorial content. Supporting content that's been shared points back to creating a trim sheet (Which only needs to tile in one direction and given how narrow the chain is that leaves plenty of room on a square texture sheet for the pendant's texture.) and creating a tiled poly strip to apply the chain texture to.
What's been found: so far there's at least two examples of low poly necklace models and both of the models and their corresponding texture sheets confirm a lot of what's already been suggested. (Both of these example models use poly strips or low poly quad tube geometry with tiling textures to represent the necklace chain.)
At this point there's enough information to attempt replicating the effect.
Recreating the chain texture: there's several different options and which makes the most sense depends on the type of chain and the visual style being replicated, as well as the available tool set.
Photosourcing could be as simple as doing an image search, finding a suitable photo, clipping out a section of the necklace chain, and inserting it into the texture sheet. If transparency is required between the links then there's the additional step of isolating or "clipping" the background out of the image. Depending on the quality of the photo and the consistency of the lighting it could be as simple as using the magic wand tool to select either the background or the chain itself or it may be necessary to manually select each link segment using the lasso tool, quick select too, or the pen / path tool.
Another option for photo sourcing is to find a pre-clipped stock image or stock texture that already has image transparency. When it comes to searching for images, start with a broad search to learn the specific terms related to the topic. In this case it's important to figure out what type of chain is used for each of the three necklaces.
A good place to start would be a general search for something like: "Jewelry chain styles" (https://en.wikipedia.org/wiki/Jewellery_chain#Styles) then compare the types of chain in the list to the chain in the model and narrow the image search down to a specific type. The links in both the first and second example models appear to have a flat locking overlap which visually matches the way Curb chain / Cuban chain lays.
(The argument could be made that the links in first example model might be a type of trace chain but the links in trace tend to sit perpendicular to each other. Since the chain the second example is definitely a type of Cuban chain it makes more sense to go with that for this exercise.)
Drawing the links in a 2D application might sound like a lot of work but it's actually pretty easy. Just trace out or free hand a single link as a transparent layer then tile it by copying that single drawing several times. This could be done in something like Photoshop or Illustrator or even Substance Designer. One of the benefits of this approach is the texture already has transparency information and it's possible to quickly add consistent lighting effects with things like gradients, shadows, emboss, and other non-destructive layer effects.
Baking the links in a 3D application provides a lot more control over the texture and it can be as simple or as complex as required. Though, depending on the complexity of the bake and just needing 3D skills, this approach probably takes a bit more effort or time to setup than the 2D approach.
If all that's needed is the basic shape, unlit shading, and image transparency then it's as simple as modeling a single link, duplicating it with an array, then grab an isometric viewport render of the chain section. If more complex surface information like normals, AO, curves, etc. are required then just place an unwrapped poly strip above the high poly chain model and bake it the same as any other model.
Putting it all together: here's a simple example of what this all could look like. Start by modeling a basic Cuban chain link then copy it with an array modifier, switch to a neutral unlit matcap, and render the viewport. Save the resulting image to a format that has transparency and do whatever image editing needs to be done to get into the texture sheet.
Here's a link to a write-up that covers modeling Cuban chain links: https://polycount.com/discussion/comment/2710581#Comment_2710581
Create a single rectangular poly strip, unwrap it, and place the UV island over the tiling chain trim section in the texture sheet, then subdivide or add loop cuts to the poly strip to create additional segments. Insert a curve (either open ended or closed loop) and edit the vertices to form the path around the character's neck, shoulders, and torso. Select the poly strip and use a deform along curve / path modifier to conform it along the shape of the curve object then use an array modifier to copy the poly strip along the length of the curve. Adjust tilt of the curve to turn the direction of the outer faces of the poly strip and adjust the length of the poly strip to match the length of the curve.
This same basic process can be used with more complex low poly strips and trim textures. It's not uncommon to see L or diamond shape poly strips and trim textures with edge details that bleed off onto the sides to create the illusion of depth or thickness. Also, depending on the technical limitations or resource constraints, it wasn't all that uncommon for transparency on tiling textures to be substituted with a darker solid background color on older or lower spec game models.
When it comes down to it, there's a lot of ways to approach each step of the process and sometimes the best way to figure out the strengths and weakness of each is to try them and compare the results. Actually working through a sample project like this using active learning and first principles techniques can really help build a ground up understanding of the how and why behind the way things are done.
This is just something of an aside, not directed at anyone in particular, but thinking tools like Socratic questioning and rubber duck debugging can also be useful for self directed discovery. It's often helpful to structure the line of inquiry towards a tangible goal like a practical demonstration of knowledge or skills, so there's a measurable outcome and a way to tell if things are getting closer or further away from the desired answers.
Search really seems to be geared for instant gratification now and while that has it's place, some of these technical topics have a sort of depth to them that requires looking at things both top down and bottom up. To understand the why you must understand the how and to understand the how you must understand the why. Something that tends to require active learning through the practical application of knowledge to work on a sample project.
There's also this sort of elegant simplicity to a lot of older technical solutions to modeling, texturing, and rendering problems. At the same time, certain older workflows were either poorly documented, obtuse, or otherwise confusing for many artists back then and probably seem even more so now. Which is why it often requires a slightly different thought process when compared to now when it's possible to brute force stuff by just throwing polygons, complex shaders, and compute power at things.
That doesn't make it right or wrong, better or worse, simply different.
Additional links to write-ups and discussions about similar models with poly strips and tiling trim sheets:
FrankPolygon
6 ·
Re: AI and future of 3d artists
We're currently 5 years into AI taking your coding job in six months..
It's all bullshit - all of it.
Also - there's no quick buck to be found here. making 3d models does not pay well
It's all bullshit - all of it.
Also - there's no quick buck to be found here. making 3d models does not pay well
poopipe
5 ·