Hi! I’m Vladislav Barzo (Barzo is, of course, my nickname; my real surname is different 🙂 ) I was born and live in Ukraine, Kirovogradskaya reg. (it’s in the middle part of the country). Since my childhood I have loved to draw. My informal education is as a “designer artist-constructor”, but I don’t call myself an artist, I can just draw. My formal education was directed towards interior design, but we were taught almost everything regarding design: photography, industrial design, painting, sculpting, etc. After graduation, I got a graphics designer job at a local company and kept working there for the next 3 years. Then I got a new job (again graphics designer), but the second company was working in the sphere of advertisements (billboards, visit cards, lightboxes, banners, etc.). At the moment I’m freelancing as a 2D-3D artist. At the same time, and since the beginning of 2014, I am a member of an indie team, VR-Gamers.
I’m making concepts, 3D models, and textures. I’m also drawing a comic book based on our game ideas, but it’s in Russian, so you probably won’t understand it. 🙂 Not long ago I started learning 3D animation.
I got myself into 2D-3D in about 2009-10 because of a childhood dream to make my own game (which is still alive inside me, despite being a grown-up dude). At that time I was working in Milkshape 3D; I read tutorials and taught myself low poly modeling.
At that time I was working for nothing. My models weren’t used anywhere until 2014 when I joined the VR-Gamers team, and we started working on a third prototype of the game, HASTE (Time-bending action), for which all the models were still made in Milkshape 3D.
When our work on the prototype wrapped, we published it for free on GameJolt and decided to make the next prototype and expand it to a full-fledged game. We had to improve graphics fidelity, which forced me to switch to Blender 3D (the switch was rather painful for me since I was working in 3ds Max at my job, but I quickly got used to Blender), which I’m using to this day.
The thing is that I still want to realize my childhood dream. At the moment I’m writing a design document for the game I envisioned and discussing it with the team. I have described the characters, weapons, etc., visualizing everything, and making game-ready 3D models (some of them you can see in my profile on Sketchfab), which will be used. The project will be comic book stylized and the Bagel Van will be an in-game lobby for players, from which they will be able to travel across the game map.
Some time ago I stumbled upon Sheng Lam’s works on the Internet and loved them because he created volume without shading. At the time I didn’t know whether his works were 3D-models or just drawings. Later I couldn’t resist the urge to try and make something similar. My first model in that style was a revolver, which turned out really well and I continued working in that style.
For this comic book style I used just two programs:
- Blender (where I make the models, unwrap them, paint, etc.)
- Photoshop (cosmetic texture improvements, adding decals and grunge). And yes, I drew all the textures with mouse 🙂
So, model creation starts with a concept, almost all of my models are concepts. I think concepts may be divided into 2 schemes: clear and unclear. Clear is when you have a 100% thoroughly thought-out idea, and you just need to realize it like a robot. Unclear is when there’s an idea, drafts, but nothing is set in stone, and the whole model is developed on the go. Most of the time I work by the second scheme.
Bagel Van is a usual soviet van UAZ-452, which is commonly called “a loaf”. I wrote “Bagel” on the windshield in Russian to personalize the vehicle. The main idea was that this van is a lobby, a Soviet van with a hardened chassis and a satellite dish:
But this is not a clear visualization of my idea, it’s just a draft on a piece of paper, just an image. Then I googled common real-life modifications for that van: some add cargo carriers on the roof, some protective tubes on the windshield, some add more plastic details near the wheels, etc. Then I downloaded actual blueprints of the van and started modeling using them:
Modeling + UVs
The overall modeling process is as everywhere else: I made one side of the van, then mirrored it. While modeling, I realized that I didn’t need a side door like on the original van, thus I skipped it and saved some space on the texture. After the modeling process was complete, I started the process of applying UV seams for unwrapping. But one must keep in mind that the model is mirrored, so if there were a text decal, it would be mirrored as well. There are a few ways to bypass this problem. The first and the easiest way is to just tear out a group of polygons from the decal area and place it on a different spot on the texture. I did something similar to the rear doors of the van (from the inside you can see that they are mirrored, but from the outside, they are unique and beautiful).
The second way is to create a second UV layer and draw all the decals there, thus creating 2 textures, in this case, a clear green van texture and an overlaid decal texture. Also one should save a place on the main texture for the overlay color (this may be read from another texture, but it’s better to keep texture count as small as possible).
Shading and Texturing
After modeling and unwrapping processes were completed, it was time for coloring. There’s nothing too complicated—I just selected 3 working colors (more colors will decrease the overall quality of the image). I used green as the primary color, orange as the secondary color (I use orange on most of the models for this project to establish a distinctive visual feature), and grey. I also used light blue to paint bare steel. The more colors you use, the more attention needs to be paid to their distribution on the model. Then I just filled the van using that palette. The model looks flat and lifeless at first, but it’s just because of a lack of outlines.
The next step was to prepare the texture for the outline painting. To do this, we opened the saved texture in Photoshop, removed the black background spaces from the resulting texture, and filled it with the colors of the texture itself. It is possible to select the black color with the selection tool and click delete, after which we copied the resulting image into 3 layers, turned off the 3rd layer, blurred the 2nd layer through the Gaussian filter (without cranking it too high), duplicated the resulting layer 10 times, combined them all into one and turned off its visibility. Then we went to the first layer and blurred it harder to fill all the black spaces, then duplicated it 10 times, combined, then turned on all the layers, and saved. This way we created texture padding, so there wouldn’t be any visual artifacts while observing the model from a distance.
Next came the long and tedious process of outlining the contour of the texture and shape of the model. I opened Blender3D, went to Texture Paint, used a Brush Tool, chose a black color and traced over the geometry to emphasize its shape. It’s not so convenient to paint in Texture display mode, so I switched to Solid display mode and back to see the results. The lines did not need to be uninterrupted lines—I could break them, add small stitches, add fracture details, etc….
When I reached the desired result, I saved the texture and fine-tuned it in Photoshop. I added new layers with dirt, grunge, and text decals if needed. That’s how we get from the original flat texture to the final one:
Also, as I said before, one should remember about having an outline not just on the texture, but on the whole model. I have an unused solid black space on the texture for an outline model.
For the windshield texture everything is the same, except for adding a transparency layer to the texture.
For the outline model, I duplicated the original model, increased its size by a few millimeters using a Solidify modifier, cranked up the Thickness value and disabled the Fill Rim checkbox (this prevents the new model from merging with the original one). Then I deleted all the inside polygons, scaled the outline model’s texture coordinates and put them inside the solid black space on the texture. After that, I flipped the outline polygons. That’s it, the model gets exported and the job is done!
Setup on Sketchfab
While uploading the model to Sketchfab we first opened the Materials window, opened the Faces Rendering tab, and pressed Single-Sided to disable backface culling. Then in the Scene window we opened the General Render tab and toggled PBR (you may use Classic, but the model gets somewhat desaturated), and in Shading selected Shadeless and set up a white background. On the Lighting tab we disabled the light source; we don’t need one, but the thing we need is the Ground Shadows so we pressed Baked AO. We didn’t need most of the post-processing filters, so I used only a little SSAO and Sharpness. Everything is straightforward. And that’s it 🙂