Vitaliy Ishkov talked about the way he approaches weapon concept design, rendering, materials, and post-process.
Hello, my name is Vitaliy Ishkov. I am a freelance hard-surface artist living in Saint-Petersburg, Russia, at the moment. I have been doing 3D for not so long, starting in 2015 at the university. One of the first works I posted on the Internet was a vehicle project called Buggy:
After that, I received a job offer from our local outsource studio and this was my first professional experience in the CG industry. In the studio, I worked on a variety of projects, for example, Warthunder. At the moment, I work remotely on several unannounced projects doing weapons, sci-fi armor and stuff like this.
In my spare time, I focus on conceptual design and try to build an iterative approach to it, develop my own visual language and learn different aspects of design to achieve efficient and high-quality results. Also, I experiment a lot with tools and different pipelines.
As I said before, I have been doing 3D since 2015, so I’ve been in the industry for 4+ years now. My way into CG began at the university where I studied Automotive Engineering and did a lot of CAD modeling. Little by little, I went into modeling more and more, started to practice in my spare time, learn polygonal modeling and rendering. I think my learning process was pretty standard: I didn’t have a mentor and used any source of information I could find like Gumroad, Gnomon, Learn Squared, etc.
After I got my first job in the game dev industry I became very interested in concept design. I think it happened because of the fact that I was fond of all sorts of engineering since childhood and especially, I was crazy about automotive design. Eventually, I found the opportunity and tools to express this passion by doing hard-surface concept art.
Speaking of self-education, to my mind, discipline is the best «tool» for me in reaching goals and becoming a better artist. However, it’s also the biggest challenge for me.
SP 300 Pistol: Start
This sci-fi pistol was made for practicing and learning design. At this period, I try to focus on weapons and props and I had a plan to design something like that a long time ago.
As usual, I started with the R&D process and references. Here is the ref board I collected:
Designing a Weapon
I have quite a standard design algorithm, and the next step after collecting references is thumbnails. During this stage, I get a better idea about the silhouette, proportions, etc. Here are the thumbnails I’ve made:
As you can see, the chosen silhouette is not that similar to the final design, but it doesn’t matter. This stage is more «psychological» for me, I just try to get rid of the blank canvas and get some base for further work. Also, when I work on the silhouettes, I go deeper into references, play with gun photos in Photoshop, do some photobashing and get a more complete view of the final design.
After I created several silhouettes, I moved onto modeling and made a very simple blockout using polygonal modeling in 3ds Max. The main task of this stage is to transfer a 2D silhouette into 3D and start working on the main (biggest) forms. I don’t think about the small details, topology and precise forms yet, just overall proportions and silhouette. The only technical thing that has to be solved at the blockout stage is to determinate which parts would be more convenient for me to do with subd modeling and which parts would be easy to do in Fusion 360. In the image, the green element was decided to do with subd modeling as I think it’s easier to make major forms using polygons and then add small details, cuts, etc. in Fusion 360.
Before exporting the green element into Fusion, I polished shapes a bit to bring them closer to the final idea and added support loops for correct smoothing. Here you can see this element prepared for exporting:
Then, I brought the whole blockout model into Fusion 360. You can find a lot of information about the process of importing a polygonal model into Fusion on the Internet.
The rest of the work was done in Fusion. After I had brought the model there, I started to work on large and medium forms.
Here you can see the blockout after a bit of work in Fusion:
To describe the main idea of the pistol: I tried to create a light personal weapon, quite classical but equipped with a stock to make it more interesting. Moreover, the gun shoots with electrical energy from the batteries which are located in the body of the gun and additionally in the stock.
As for the functional aspect of the design, at the moment functionality is on the second place for me, and I’m trying to concentrate more on the visual part. So I didn’t worry a lot about the technical details, and there is no science behind the pistol. But at the same time, if we are talking about weapon design, I always try to pay more attention to the grip and the elements users interact with (buttons, switches, etc.) and make those elements more ergonomic, give them the right scale and proportions. In my opinion, grip often draws the most attention and if it looks clumsy, super uncomfortable (with way too sharp corners, wrong proportions, strange forms, etc.), the design can seem broken.
I know good design is a form + function, but as I said, right now, I focus on the visual part more. I believe that the visual part is more important in concept art for video games and cinematics. Once I achieve a certain level in the visuals I will work on the functionality.
Visual Aspects of Design
Now, let’s talk about the visual aspect of design. Generally, it’s the trickiest part for me, and I find it hard to put together all my thoughts on this topic. Below, I listed a few materials that helped me to figure out some rules for this:
- Alex Senechal, «Visual Design Basics»
- Anthony Jones, «Design Core Principles»
- Mike Hill, «Intrinsic Vehicle Design»
- Joe Peterson, «Tech Tips Series»
- A Russian book called «Композиция в технике» by Somov Y. S.
Next, let’s talk about some rules I keep in mind during the design process:
- Avoid evenness:
I think dividing an object into primary, secondary and tertiary forms or main and subforms is a good way to create a hierarchy: something that bigger is more important than something smaller. Also, this rule helps you to create size contrast and variation that feels more natural than the same sized objects.
Also, I try to create a hierarchy on different levels, for example, the first level is the whole weapon (dividing big forms):
Then I divide medium forms, for example:
And the last level is small details:
- The next rule is «Dynamic Balance» which I also call«Zig-zag» detail distribution:
This means that you need to choose the main axis (or 2 axes) of an object and distribute details asymmetrically relative to this axis (axes). I found this asymmetric weight distribution visually more interesting than symmetric one and also this method helps you to balance composition between main axes.
- Echoing (details/angles/shapes/colors):
It’s a pretty common rule. You can take a detail and copy it several times, plus you can copy angles and colors. This rule helps you to make your design simpler and more holistic.
- Grouping (details):
By grouping we create natural rest areas and also it helps to fortify existing areas of details.
- Plane changes/depth:
Big flat surfaces are boring! I often try to make forms look more interesting by using plane changes and depth.
That’s it. You can find more rules and greater explanation in tutorials that are listed above.
Let’s talk about rendering and materials a bit. In this work, I used KeyShot for rendering and Photoshop for post-production. Unfortunately, I don’t have any super-secret effective techniques that I could share with you, so I’ll just describe my process.
First things first, I distribute standard Keyshot materials and choose a color scheme. During this process, I am using one of the standard Keyshot HDR map and simple black background; I don’t worry about lighting, background or environment at this step, I just distribute various types of materials according to my taste. For me, Brian Sum‘s works are one of the best examples of color distribution. I like the way Brian uses contrasting colors to create hierarchy and determine primary, secondary and tertiary forms.
Here you can see the pistol with basic materials on it:
After that, I usually start to work on the lighting setup. As a base, I use a standard 3 point light scheme. According to this scheme I set up 3 emitters:
- Key light;
- Fill light;
- Rim light.
The Key Light: This is your main source of light and is the brightest light in your scene. The key light is what will give you your scene and subject its overall exposure. Usually, it’s placed in front of the subject and is often off to one side to create some dimension and depth.
The Fill Light: This is your second source of light. It’s usually much dimmer than the key light and is used to fill in any shadows created by the key light. The reason you may want to use a fill light is to retain some detail in the shadow areas and to reduce the overall contrast of the scene.
The Rim Light: This is your third and final source of light in a three-point lighting setup. It’s usually placed behind the subject, sometimes off to one side, directly behind or overhead but still behind the subject. The backlight is used to create separation between the subject and the background so that they don’t fade into each other. It’s achieved by creating a “highlight” around the outline of the subject.
After a basic 3-point lighting setup, I start to think about the environment. The environment can strongly affect the lighting because of the reflections, color bleeding and stuff like this, so I think you need to figure out the environment on the early stage of lightning. As you can see, I decided to use just a plane, but what is interesting, for this plane, I used the material called Thin Film in Keyshot.
I experimented a lot with the plane and when I accidentally applied this material I realized that this is what I was looking for. Thin Film has non-uniform reflections of different colors, and for me, it looks very interesting. I also added a standard normal bump map from the Keyshot library and a black and white roughness map to make reflections even more interesting.
Next, I proceeded to in-depth work on lighting. I started adding more light sources in order to achieve the following goals:
- Well-readable forms: it’s necessary to achieve different tones in the areas where the plane changes and nice gradients on cylindrical surfaces, but I try to prevent super-contrasting transitions between the planes to make sure that image doesn’t look messy and overloaded.
- Depth/overlaps: I always try to make sure that the deepest areas (negative spaces, deep cuts etc.) are darkened. Also, if I have overlapped objects, I try to make objects that are closer (to the camera) brighter than those that are farther away. It helps to create a feeling of depth.
Then, I started to work on the materials. I think it’s convenient to work on them in parallel with the lighting setup, but I highlighted this stage as a separate chapter after the lighting setup, just to make the article consistent.
When I work with materials, I always keep in mind that forms should be well-readable (materials + lighting = good form reading) and I don’t really worry about material parameters. I think when you are doing concept art it’s not so important to keep all parameters of the materials physically corrected, compared to production, for example. Also, I keep materials fairly simple and clean, as to my mind, the forms read better this way.
Most of the shaders just standard Keyshot materials with roughness and normal maps. For example, for the grip, I used a standard Keyshot material called Black Paint and just added roughness and normal bump maps to it. Almost all of the maps that are used in this project (normal and roughness) are bought on Quixel Megascans. I use the same process for all materials in this work.
As you can see, everything is quite simple. I feel it’s all about the rendering process and the lighting setup, so I spend most of the time on those.
This is my raw render:
And here’s the result after post-production:
Let me describe post-process on the example of this image below.
The raw render looks like that:
First of all, I’ve added this white pattern:
The model hasn’t UVs so I just rendered another shot with white grip and added those white spots in Photoshop using a mask.
I wasn’t really happy with the lightning in several areas (marked with red in the image):
So, I’ve added a few additional emitters in the Keyshot scene, rendered another shot and overlayed it in Photoshop using a mask, I did this because I didn’t want to change the whole lighting of the frame but only in these areas. Yes, it’s not physically accurate but it doesn’t matter in this specific case.
And I’ve got this:
Then, I’ve added smart sharpness in Photoshop and just masked out areas that are not in focus. It helps to pop-up details a bit. I also changed the color of the grip (If I remember right, I used Clown Pass in Keyshot to select grip only and change the hue for it). Also, I played a bit with color grading, saturation, and contrast using standard Photoshop plugin called Filter Camera Raw. I made the whole image a bit warmer, increased exposure a bit, and decreased the floor darkness.
And here you can see what I’ve got:
In the end, I added a really slight flare effect on the glass using flares library from the Knoll Light Factory.
Then, I put a photo texture (found on the internet) on top of the whole image with Screen mode and really low opacity (something like 5 or 10%).
Here is the final result: