New! Pixel8r v1.5
https://gum.co/pixel8r
New! Advanced Normal to Roughness Node
https://gum.co/n2r
Dog Library Generators - Substance Designer
https://gum.co/DogLibrary
3D Wood Generator - Substance Designer
https://gumroad.com/l/3dwood
If you find these valuable, please consider purchasing in order to help support my efforts towards providing more things like this.
Thank you for looking! All feedback is welcome!
Replies
I'll post some more tomorrow (I've got 10 others ones from the library to upload), and perhaps some of the continuing work I've done on the textures in my first post. Destroy my butt w/ crits pls!
Edited to fix an obvious repetition. Note that this is tiling 2x.
Here's some more things I've been working on. The first are rendered in SD.
Rocky dirt:
Old hammered copper:
The following are rendered in Unity with my custom shader:
Candle wax! The subsurface shader variant has parameters for internal color, pulling the transmission color from the albedo, and can take a thickness map or simply just using a float. It was a darker earlier which looked better imo but I couldn't be arsed to change it.
Painted walls:
Concrete:
The quality is varying on these mainly due to a lot of experimenting and improvement along the way, as well as all this being on a quick turnaround. I'll continually improve them! As always, criticism is very welcome. Tell me I suck!
In the meantime I've done a couple other textures. I'm pretty satisfied with this one but couldn't get the color to my liking. This is my third go at making bark. This pic is also my first time using Toolbag!
And here's a generic rust. I tried to take forms and colors from various reference and incorporate the most visually interesting aspects
C&C is welcome as always!
Edit:
I was dissatisfied with the gaps between the bark. It looked good at a distance and was somewhat like some types of bark out there, but I didn't feel like it was as cool. I also brought the color I used to have back in, which wasn't posted before.
Finally there were some peeps discussing marble in the Substance section, and while I'm by no means even close to making a great marble I thought my approach could benefit some because allows you to simply and continuously add complexity. Linkage: http://www.polycount.com/forum/showpost.php?p=2288796&postcount=1635
Tiling 2x. This is just a quick proof of concept for the method I came up with. The goal was to create a mask with even spacing so it can be controlled very finely, which I found was much easier than expected. I added some quick cracks and chipping to make it less boring for presentation, but that will get thrown out. Hope you all like it!
As a bonus here's the mask that was generated:
And another bonus, a SD screenie showing how easy it is to modify! The latter image is a two minute slider adjustment job, nothing more. I'm very pleased with the ease of use and how this experiment came out overall!
This was posted in the Substance thread but I figured it's a good update to go along with the above:
There's some great discussion about the technique in the Substance Designer thread on this page and the next:
http://www.polycount.com/forum/showthread.php?p=2291158#post2291158
Everyone is using the splatter circular node to place a shape around and then combining the results of doing that repeatedly together. My original continually applied the fan mask shape but I realized afterwards it only needed to be applied at the last step, so my graph became a bit simpler than the one I posted.
My result is different because I use a blurry shape to create stones that fit the areas rather than bricks, which I like because you don't have the result of bricks looking like they're cut off prematurely.
I started on my own take on the Defiant from Deep Space 9. My goal is to merge some of the more modern sci-fi style elements from the new films, while keeping it true to the spirit of the original. Right now I'm just experimenting and blocking things out while I do as much research into Star Trek design philosophy as possible.
The end result will be designed to make sure it accommodates a properly measured 8 deck layout that I'll design. Happily, I somehow nailed the initial measurements by eye. The hull so far is measured to fit the decks (and jefferies tubes) and sits at 170m long, making it a good deal larger overall than the original Defiant. Ignore the random greeble textures, I don't like looking at grey.
Here's my current progress, experimented with rendering in Corona which I already love:
I thought I'd show a bit of my previous progress renders over the days. As you can see I ditched the saucer top in favor of something more like the original, but am thinking of ways to merge the two forms in a more unique way.
I'll be continuing this slowly until the end of time~ Any feedback or ideas for exploring new directions is extremely welcome, especially from any fans out there!
Great job so far
I've only used Substance for these. It's very fun once you get into the workflow of it! The methodology lies in analyzing and breaking down a surface's structure, then combining different procedural patterns that approximate the components of a material, layering in complexity as you go. Observation is really key, something which I'm still learning a lot
Here's an example of a complete graph, this one for those wood planks
Just to not seem like a dangus, the unrealistic cracking was intentionally there in yesterday's post just to show my experiments on branching crack patterns. As always, tell me what I can do better!
Edit: Updated Substance, textures, and showcase images to reflect fixing minor fuck ups on my part.
One thing I'm particularly happy about is getting random values for shapes in a black and white mask. It's definitely not perfect and I'd love to find a better method, but tweaking around with a couple of the nodes gets really great results!
Now here's the Substance file for whatever your pleasure is~ I hope it's valuable to some!
https://drive.google.com/file/d/0B4vo_o3utgw6ektEdlpLMGpVa1E/view?usp=sharing
A zip of the textures can be found here here:
https://drive.google.com/file/d/0B4vo_o3utgw6MERrUzBDNWt6TTA/view?usp=sharing
Feel free to do what you want with these. Attribution is not necessary but much appreciated. Link back to me if you like my work! Thanks for viewing and all feedback is welcome!
I'm a complete noob still learning the substance package and its workflow. when you created the wood substance, did you start with the initial 4 nodes (albedo, normal, roughness, metal) and work backwards, or did you start with the far left nodes and progress from there?
From there I start to create finer and finer levels of whatever the materials structure is. I do that in sets of masks that I start to blend together in different ways for each texture. I'll make a simple base pass for everything, sometimes just filling some texture slots with solid values/colors to get a feel for the overall material.
Then I focus on building a good heightmap so I can distinguish what lies where, and convert it to a normal map later down the line. The heightmap usually has a big influence on some of the other textures so it's the first thing to nail for me. All the while I'm still working in passes to add in the various effects I want. If you check out my cobble Substance, I did it exactly in a left to right order. Each time I add a new pass in (each pass is marked in the file and the Albedo is probably the most clear to understand initially) I hook up new versions of each map, then tweak or redo things until I'm happy with not only the result, but how I achieved the result.
The most important thing to me is making each approach to a pass work as understandably and predictably as I'm able. I hope this helps! And good luck
Small update too, I'm working on a beat up metal for something!
Don't care how this sounds but I'm loving the wood bro!
I was gonna post this when the west coast polycounters were getting settled at home for the evening, but couldn't resist!
I overhauled the wood a great deal and am releasing it for free. It's much more efficient, and despite the pattern taking some doing it totally paid off! There's some ideas in my head for how to make knots in wood, but I didn't bother with that this go around since it's not the most important visual detail to worry about.
Grab the Substance file!
Grab the texture maps!
Attribution is appreciated if anyone uses these
Edit: Here's a two minute edit showing how it's very easy to modify. This is a blend between a worn and clean version. The graph was unchanged, I only modified things like opacity values on masks.
Now to get back to the Defiant!
~Download both sets of textures and Substances~
I've also started on a shingles substance
This is a Pre-Integrated Skin shader I created, running in Unity 4. Models aren't mine, they're just for testing purposes
There's 2 shader variants: one computing curvature in the shader, and one utilizing a curvature map that gets combined with the mesh curvature. It also can use an optional transmission map. There are controls for blurring the normal map just for the transmission (making things appear bumpier or waxier) as well as a curvature multiplier to control falloff. Also, I darkened the skin color of the middle ones in order to evaluate different skin complexions.
Credits for the models go to:
Zombie - 3D Molier Studio, Tatiana Seguin - EISKO, Male - Ten24
Working on some bricks
You, my friend have just made a sale haha
This is my second test with MD. The polygon shapes are caused by the character the garment was simulated on not being very high res. For a while I had a lot of problems getting the waistband to stick (via sewing) without causing clipping in the sim, so I didn't bother reseating the pants after I solved that problem. Just imagine Theodore Twombly owns these jeans instead :P
And in case you were wondering, yes. Those are real ass pockets
This was a fun experiment and I want to branch out into non-clothing stuff soon!
more MD stuff
Today I did a warmup of a nylon military strap in MD but I don't have a render of that. Instead I'll show what that warmup helped me do afterwards
Quick and simple render. All that was done outside of MD was add the buttons, pole, and links. The pole and links also contributed to the sim. If only it were possible to simulate rigid bodies as well I'd do it all in one go! I do love the fact that the topology inherently contains the stitch lines, so adding stitching in a 3d app would be as simple as converting them to splines and displacing a stitch mesh. I'll probably toy with another piece and take it to a full finish this weekend to do just that!
WIP Snowy Rocks, for a game.
It'll have adjustable parameters for rock size, snow amount, etc once I create the functions
It's very simple at the moment. The albedo and roughness are flat values and this is how the height and normals are constructed.
Hi all! I've been working on a series of articles to help people better understand how rendering works. I've always wanted to provide people with the knowledge to make the best art they can, and have seen firsthand just how hard the learning process can be. Check out the first article from the link below:
Please if you have an feedback or suggestions for topics to explore, leave a comment or contact me through PM or the email provided in the article!
I'm sorry if this is a somewhat harsh critique of your article, I just wanted to give you an honest opinion of it, from somebody who has to communicate with artists about technical subjects all the time. As soon as I mention anything slightly mathematical people tend to fall asleep! I look forward to reading your future articles and hope you can take some of my points on board. I'd love to see an artist-friendly breakdown of rasterisation to show to people!
Your critique is perfect
The first article presents the reader with the bare-bones rendering equation to describe a framework that the rest of the articles will be based upon. Next I'm going to be writing about physical values with the core subjects being gamma and tone-mapping. After that I plan to get into basic rendering implementations, then actually get into modern PBR setups.
I know that the math tends to turn heads, but I don't want to shy away from any of it because in its absence I've seen artists go very awry. The actual amount I get into is heavily going to depend on the subject at hand. But rest assured I will provide a solid, approachable foundation for understanding exactly what you're doing when you make any texture and then what happens behind the scenes. It's a very difficult balancing act, one that I'm relying on the community to help me solve through continuous feedback. In this thread and in my articles I will make as many explanations as needed to help anyone understand a concept they're having a hard time with. Honestly though, while my primary target audience is nontechnical artists it's ultimately up to them to want to learn something that is likely going to be a bit hard. This work is a process of me collecting all the correct information in one place and making their learning as easy as possible. I want different types of people to gain from this depending on just how far they want to go, so with that goal in mind I've got to cover all the bases.
To answer your specifics: Starting from that framework I'll be talking about just how you take the theory and turn it into both a path tracer and a rasteriser while drawing comparisons between the two all along the way. Articles 1 and 2 are prerequisites for that process since I need to teach people about light and energy conservation.
Article 3 will include a historical account of major rendering milestones, so you can expect things like NdotL and various lighting functions to be covered here without calculus.
I am imagining these first works will become more enlightening retrospectively, so I hope people can be patient with my line of thought. Who knows though, I may do a total rewrite after this point! Something I also want to cover is how various engines put everything into practice and the strengths/shortcomings of various techniques. Would people be interested in this as its own article down the line?
Anyways I hope this gave some insight into my thought process. Please don't let all this discourage anyone from telling me my approach is flawed though! If I can form better ways for structuring these articles then I will gladly do the work
Currently finalizing the first (and hardest) of part of it. It's split into discussing the concepts behind capturing light values then processing with white balance, gamma, and tonemapping to produce a final image!
In the meantime have a few oldies showing off my custom Unity shader. IIRC the high poly models for the three phones are to an accuracy of about 1.5mm besides having to fudge some lens elements and the tiny holes over the speakers! Those features are all there though. Even the fresnel light on the iPhones was modeled.
I'm nervous to post this here because I suck at drawing, but I'm designing an "Overseer's PipBoy" with a wrist mounted addon. It is heavily inspired from concept art from Fallout 3! For those of you who have seen those you'll surely recognize the telephone style cord. Any criticism or feedback I can get is much appreciated, I'll be taking this lil thing from concept to finished!
Nothing is final right now other than wanting to keep the main design elements: There's a 16:9 screen on the PipBoy itself, not really seen in the lore afaik but it looks cool. For the addon it's got a broadcast mic and dial, 14 character keypad + shift + space, extra holotape slot, and the cord will be using the same connector type as the one featured on the back of the Fallout 4 PipBoy.
Thanks all
Oh yeah and I just put shit over the pipboy because I haven't really worked on it. I've been spending days just working on the design of the addon first.
^ A bit old, it's not so fat now.
Been pulling assets from Fallout 4 to help with scale. I've also got the PipBoy Edition which is handy. This goes in hand with the materials, which I've begun on as well. Got a lot of nice closeups photos of various surface types the other day.
This gives me a 60s-90s cgi reel vibe for some reason haha
The scale appears larger than shown in your concept as well, it seems to overlap the hand in the blockout, while being much more compact in the drawing, but I guess with the size of the buttons you'd need it to be that scale in order to use comfortably. The concept reminded me of blackberry keyboards, but the dials look about the right size. Looking forward to see what you do with the pipboy itself
Thanks for the feedback man! My first big goal was nailing the 50s-60s styled ergonomics of Fallout 4's assets. I don't have a design in mind for the actual base, but have considered that it should be more form fitting eventually (most likely I'll design around a ref of a splayed hand to not restrict movement much). I'll be attaching it to a glove so I have spent time on solving how it would attach/detach from the glove.
I agree with the scaling issue. Seems pretty weighty at the moment. The addon got scaled up in 3D in proportion to the hand once I ripped the Holotape model from the game and fit it to that. I'll probably also take measurements from keyboard keys soon and make some design tradeoffs to slim everything back down.
Here's some of the references I've gathered, I can upload my highest res versions if anyone wants them!
Oh yeah and I got this lil guy which of course you can see I immediately broke the damn top button on.
Taking my mind off the PipBoy for a while to work on a simple comfy scene.