@Tristan: I studied computergrafics for 5 years. I'm making 3D art now since about half a year fulltime, but I had some experience before that. Its hard to focus on one thing, it took me half a year to understand most of the vegetation creation pipelines. For speeding up your workflow maybe spend a bit time with the megascans library. Making 3D vegetation starts from going outside for photoscanns to profiling your assets. Start with one thing and master this. @Maxime: The difference between my technique and Z-passing on distant objects is quiet the same. (- the higher vertex count) I would start using this at about 10-15m+. In this inner radius you are using (mostly high) cascaded shadows, the less the shader complexety in this areas, the less the shader instructions. When I started this project, the polycount was a bit to high. Now I found the best balance between a "lowpoly" mesh and the less possible overdraw. The conclusion of this technique is easily using a slightly higher vertex count on the mesh for reducing the quad overdraw and shader complexity. In matters visual quality a "high poly" plant will allways look better than a blade of grass on a plane.
Is this not like gear VR or anything else
Stanislav Mikhailov did a breakdown of his mind-blowing procedural architectural facade made 99% in Substance Designer and talked about the advantages of the procedural materials like this.
Hi everyone. My name is Stanislav Mikhailov or just Stas. I am 29, I live in Minsk, the capital of Belarus. I have been working in gamedev for a little more than a year. Let me tell you some details about how I got into it.
I was keen on video games, movie VFX and drawing ever since school days. If my remember it correctly, it was my cousin who showed me one of the first versions of 3ds Max. The idea of making CG sparked my interest in the blink of an eye. These years I had no idea about the amount of educational materials we would have today. You know, once I had to make a few exactly the same flying ships from primitives because I didn’t know that I could just copy them. Now I feel like it was ages ago. Long story short, after school I enrolled at a university to study architecture. I had been learning a composition, coloristics, history of arts, academic drawing, classical sculpture and so on. All this knowledge really help me today. After the university, I worked as an architect for 4 years and devoted more and more time for an Arch Viz. Also, I was looking for a job in gamedev but my portfolio was not strong enough. Only after 4 years of practice I was lucky enough to get through the job interview at DoorZ where I have been working up to now.
Right now I am working on a project Naked Sun, it is a VR shooter available on Steam. When I was appointed to that project, I had to upgrade my skills really quickly because my duties were not only making an environment but also characters and weapons. I was involved in the creation of weapon modifications (you can find these artworks in my ArtStation portfolio). Among the project’s features were open spaces, quick progress and fast-paced action scenes. VR required a lot of restrictions and we had to invent a lot of tricks for a performance optimization. One more thing I want to add is that being surrounded by colleagues and game development atmosphere, I felt that my level had been growing really quickly after I found my way into the industry.
My story has a lot of twists but the main idea I want to share with you is that the way into gamedev is not always straight and easy. If now you are working at a bank or a shop, you are over 30, and not in the industry yet – don’t give up. Provided with all that available courses and tutorials, if you try hard you can get a desirable job pretty quickly.
Architectural Facades: Why Substance Designer?
For me creation of complicated architectural facades in Substance Designer is a kind of a study. I have never used facades made in Substance Designer only. Usually, facades in games are a combination of geometry, seamless materials, decals, and floaters. The last ones add a variety and hide noticeable tiling. I believe that complicated facades made completely in Substance Designer demonstrate a high level of your skills and will look great in your portfolio. Most viewers (among them could be your potential employers) will be more impressed by a facade made with Substance Designer then the same model made in 3ds Max, Blender, Modo or Maya. A Substance Designer sign on the preview of your artwork always attracts the attention of ArtStation visitors.
Using facades made in Substance Designer has strong technical advantages. Substance files allow you to control necessary parameters directly in a game engine. Bind parameters to the code and get impressive results at generativity and randomness. If you need to create 50 different houses with a variety of tiles (e.g. with graffitis in random places), you can use Substance to create an advanced facades generator and do it with the code. If we also involve in the process a Houdini specialist to create a geometry generator and merge it with ours (for facades, hardwood, roof tiles or something else) we will get a completely generative solution for architectural objects in games. The trend towards facades creation in Substance Designer appeared when new tools met with the old demand. Nowadays this approach is a must-have for large projects. If you do everything manually, an urgent task to change windows’ color, roof tiles or make a house two floors lower will utterly ruin your deadlines. This is unacceptable!
The most important feature of Substance Designer is not even the ability to create patterns, cloth materials, believable scratches, and dust. For me, it is a wide range of tools for height maps creation.
You can literally sculpt volumetric structures. Look at the artworks by Daniel Thiger, he makes both photorealistic stones and organic surfaces and generators for car wheels, wrist watches, swords and medals. The possibilities of the procedural materials in Substance are endless.
I forgot to mention that I have been seriously learning Substance Designer for about a month. I began with free YouTube lessons, got a general understanding and became familiar with the interface. Next step was Daniel Thiger’s lessons. In his short videos he shares really cool tricks and techniques, most of them are exactly on the creation of complicated height maps: how to transform a gradient to the editable false profile, how to randomly distribute blocks of columns, grow grass around stones, create realistic cracks and damaged edges of plates. All of that and many other things I picked up from his lessons.
I did not make any blockouts for this project. Once I met an artwork by Enrico Tammekänd on Facebook I felt as if I needed a similar presentation to demonstrate the skills obtained at Daniel Thiger’s lessons. I drew the main blocks and architectural elements on paper, the first version didn’t have upper windows and many current details. Then I decided to add to the project an Ionic column. Frankly speaking, that was the object I began with. When I was making it, I figured out how the full facade would look like.
Then I made a part of the graph responsible for the facade’s structure. It consisted of placeholders that showed how deeply each element will be embossed and where it will be placed. After that I had to make lower windows (I sketched them on the reverse side of a shop receipt), upper windows, roof tiles and attach these elements as placeholders. It may be said that it was an iterative process: many elements were changed during the work.
When I proceeded to the presentation in Marmoset, I was initially going to set the facade on a platform but after a few experiments, I assigned the material to a two-storied cylinder. The result impressed me, so I decided to opt for it.
Now about the nodes I used. For the cap of the column, I downloaded Spiral Generator.
Also, I frequently used nodes Gradient and Curve to make all profiles (for the cornice and so on).
For the columns, I used Gradient to create an effect of an attached cylinder, 2D Transformation to mold a shape for a thin column, and Levels to get a proper bend. Directional Wrap was used to get a natural random displacement of the cornice and columns’ blocks.
The roof tiles were made with gradients. In fact, it is the same attached cylinder. I just used Tile Sampler to distribute tiles around the surface. Also, a very important role was played by Slope Blur: it helped to make ragged edges and chuckholes.
These are the basic nodes I used. I think it’s right not to elaborate on them – you can find all the secrets and tricks in Daniel’s lessons.
As for the structure, all the elements of the graph are placed from specific to general. The first level (the deepest), is a separate cap of the column, its base, etc. The second level, the middle one, is the column’s assembly with some effects. The third (upper) level of the graph contains the elements of the middle level on their places with 2D Transform, Tile Sampler, and Mirror nodes.
Here I will start with the curtains. I’ve mentioned everywhere that this project is 99% Substance Designer. By the last percent I meant the curtains. They were cut from the stock photos in Photoshop and added to the Emissive Map. I also had a node for a glass mask. I exported it from Substance to Photoshop as .png and made an Emissive Map from a few pictures of curtains. Initially, I set up the glow of the central part of the lower windows in Marmoset by increasing the glow intensity of the material. But it resulted in overexposed curtains, so I made them very dark in the Emissive Map to make them distinguishable.
I could easily do it in Substance but the graph I built was too heavy for my computer, so I didn’t want to add new elements.
As for the glass, I had a mask for it, so I used it to color the glass in blue inside the Base Map and to decrease grain. The only thing that needs to be mentioned is flares on windows. In Marmoset, if you render your projects without an environment and parallax on windows, the only way to show a glass material is a glossy flare from a light object. You can compare my images by yourself, some of them don’t have this effect.
The Level of Modularity
The structure of my project is fairly modular and to a certain extent generative. I can change profiles of all the elements, a height of the cornice, a number of its elements, scale roof tiles and change its slope, change a number of spiral turns on the cap, all colors, shape of big and small damages, and many other features. But as much as I’d like to, I can’t turn it into something like the Early Gothic facade.
I have learned some interesting approaches to the facades creation and I can use them to reach the final result on a proper level. But, as I have already mentioned before, I have been using Substance Designer not for a long time and in my graphs I had to use some tricks and a “duct tape” . I am sure that elegance, reusability, and optimization will come to me with practice.
Can the Material Be Used in the Game Production?
It is also possible to use this material in games. For example, if it is an interior with a window facing the street, we can use this facade for a nearby house. If the windows are narrow and we can’t see facade’s borders, it is possible to use a parallax material to imitate the depth.
The second option is to use height maps to emboss geometry in 3D software and then make some simple geometry above the embossed one or above the retopology. In that case, we will need to assign our material to a planar UV projection without a height map.
If you are new to Substance Designer, I can recommend you to watch any free courses for beginners covering interface and philosophy of substance materials. There is no point to go into the details, it should be an introductory tutorial. Opt for a well-structured course than for some separate lessons on different topics. The course below was the first I watched.
Don’t get down to the large projects, they require too much time and effort. If you are going to bridge the skill gaps along the way, there is a big chance that you will not find the necessary information at once and won’t bring your project to an end. The worst thing is that you will waste your time and there will be nothing to add to your portfolio.
Don’t polish some large ideal project, as you will level up faster on a few small projects instead. I want to pay a special attention to that idea. I believe that the right solution is to work on the projects which will show your skills obtained on some course. Be realistic and do something within your strength, something you will definitely be able to finish. If you don’t finish projects and just do some experiments with nodes, you will run out of gas pretty quickly. You must create many small projects on the way to the big ones. If you do nothing except learning, your brain gets bored. Where are the likes, attention, subscribers, and friends’ compliments? Why on earth am I doing all of that? Support your zest for learning by showing your finished artworks to the audience.
When you are familiar with the basics, I would advise you to finish a more specific course. Don’t consider this interview as Daniel Thiger’s courses advertising, but I want to recommend him again. He gives a lot of comprehensive information within a very short time. Recently he’s finished the fourth part of the Substance Designer fundamentals. It will also be helpful to read threads on the Allegorithmic forum. Another method is an in-depth study of the graphs made by other authors who share them on ArtStation.
There’s one more step in studying SD that I’d call focusing on a specific topic. When you feel confident and learn general techniques to be able to make a passable hardwood floor, architecture, stones, etc., it will be the right time to go deeper into more narrow topics and concentrate on the lessons about, for example, sci-fi or metal (even narrower). Learn only the chosen topic for a few weeks to push the boundaries of your next project.
So, this it it. I hope someone will find this article helpful. This is my first interview with 80.lv. Thanks to the team for giving me this opportunity.