Designed completely in Blender 2.9 and rendered in part using cycles. Retouching done in Adobe Lightroom and Photoshop.
For those who are interested, I’ll be creating a follow-up reply to this post going into a bit more detail into some of the steps and phases I took to finally end up with the design you see here. This project ended up landing me a design gig with a private car builder for a new design, so I hope to elaborate in the future on how Blender has become a valuable tool in the design process. Thank you for the inspirational work many of you post here.
For those that are interested, this is The H6 Guerriera Story. This is a little bit of what I do in blender and my mentality in setting myself up for design and execution. This is extremely general and condensed:
Research
After I finished my last design, it had been a long time since i worked on a car. I knew I wanted to do something more realistic, and more in-tune with something you would see against some of the more famous brands in the automotive industry. Although I enjoy coming up with weird and wild designs, I wanted to challenge myself to do something closer to something out of Motor Trend.
I wanted something familiar, yet unique. Something a little wild, but mostly familiar. Something inspired by all my favorite car designs. Something people could easily remember and pick out from a lineup of cars. However, this is alot more difficult than it sounds. I spent about 6 months just studying different cars and what makes them really work. I wanted to find new inspiration.
I finally decided that I wanted to use the Ferrari F8 Tributo platform as my general inspiration for laying out my proportions and general dimensions for the car. I really fell in love with the design.
I found every specification I could get on the F8 to layout the dimensions I wanted to start from on my car.
I knew that if I wanted to achieve the level of realism and really sell potential builders on the idea of a project that doesn’t exist yet, I had to really focus on designing every possible detail you would find on existing cars. Down to the rubber lining and plastic seals you find around windows, headlights, door handles. As far as I could possibly go.
I don’t have any formal automotive design training, only the hands on experience I’m learning with the car builder I work with and my experience working at an engineering firm.
Design Philosophy
Organization and detail before anything. For me, the most important thing above anything else in a very complex job is having an extremely detailed sense of organization. I like to lay out the entire plan by phases from the very beginning of the project and setting clear goals and checkpoints for each phase. This helps me control and organize my creative urges. To me, design has a hierarchy and as much as I want to start with the extreme details from the very beginning, it’s almost impossible to do that without a solid foundation, that builds into the next and into the next. To me, every single step and detail, no matter how small, needs your entire focus and planning. I feel that being a good designer is about solving problems and making really good decisions hundreds, if not thousands of times within a single project.
For this project, I ended up with about 7 major design phases. I like to keep my phases within a single blender file creating a new copy of my scene so I can quickly go back and forth to see the transition from where I started to my current phase. This helps me see and fix (or change) design features pretty quickly based on the direction of the design.
Ideation and Sketching
I generally like to start sketch designing and layout with the side profile of the car. This tends to be the most recognizable and most impactful aspect of the design of the car. Every famous car to me starts with an incredibly strong and iconic silhouette. It’s in this body shape that the eye can immediately “trace” the lines of the car before we inspect into closer detail. To me, a powerful 2D side projection of the car makes or breaks the direction where the design will go.
To stay realistic to true and familiar proportions, I started with the Ferrari F8 dimensions. That means every possible dimension: wheel sizes, wheelbase, ground clearance dimensions, general wheel gap clearances, height, width, length, length of the rear bumper to center of the rear wheel, front wheel to the front bumper, line of sight, etc. I created basic dimension boxes to guide me in the sketching of the lines within those boxes.
I spent about 2 months sketching and deciding on a final shape for the silouette. Im a terrible hand sketcher, so I sketch strictly manipulating and combining rough beizer lines. Constantly rotating the car, visualizing the shape of the car and surfaces (even without having to create surfaces yet). I dont worry about getting to detailed in the beginning. i just want to outline general shapes and lines that work in harmony from the major projections. Side, Top, Front, Back and Isometric views. Once that feels perfect, I start my next phase.
Creating the Curve Network using Beziers and Blender Snap features
Once I’ve laid out the general shapes and curves, I start going deeper and “snapping” more bezier curves together to fill in the body shape. I think of it as building the skeleton of the car. I’m constantly refining the shape of the car as I see more of the lines emerge together. These are the lines that I will eventually convert to mesh to create the skin later on.
In this project, I developed a method that works for me in where I snap beizer lines either by vertices or by edge. This lets me slide the ends of the bezier curves up and down another curve until I fine-tune the specific shape I’m working on. Later on I “weld” the crossing points of the curves that become the surface. This prepares me for the next phase, general surfacing.
General Surfacing dnd Complex Surfacing of the Base Mesh
As I have explained in other posts and articles, when I use Blender for surfacing a vehicle, I rely heavily on the base mesh/shrink wrap method. I learned how to use this effectively in CG Masters Car Creation Course. https://cgmasters.net/training-courses/master-car-creation-in-blender/
This is the part I start stitching and building the surface of the car by connecting and creating either triangles or 4 sided polygons with a light subsurface modifier applied.
Since Blender is not designed to really facilitate Class-A surfacing, I have to rely on an extremely calculated layout of my vertices and meshes, using as few as possible to create the base mesh. The cleaner and stronger the base mesh looks and flows, the stronger and cleaner the surfacing on the car ends up being when applying the copy to shrink wrap over the base mesh. In the beginning, the car surface tends to resemble more of an ugly, soft clay appearance. A lot like if your sculpting in real clay. But I know that once i get my primary shapes down, I will edge out and “harden” the body using a combination of careful edge loop slides (sometimes just removing certain vertices along certain shapes and transition to make the curve work).
This is one of the most time-consuming parts. Im constantly checking and rechecking the surfaces using every possible shadow, unflattering light, unflattering (but revealing Matcaps) to really expose the flaws in the surface. I know a lot of designers work against zebra stripes or stripe lighting to really flesh out the imperfections in the surfaces. The primary surfaces tend to be the easiest while the secondary and transitional surfaces tend to be the hardest to control.
It usually takes a lot of practice moves to properly layout what edge loops need to be added to harden some edges and keep others soft and transitional. The beauty and complexity of any vehicle is that there are a large number of different edge curvatures and shapes. Some lines are hard accent lines while others tend to be soft gradual lines. That was the hardest part to layout, especially working around the intake channels of the doors and the rear end design details.
Surface Shrinkwrap and Detailing
Once I created the base mesh, I start my next phase and create the shrinkwrap copy of my base mesh to start detailing the car and cutting out panels, doors, sketching my headlights, etc. If I don’t like a certain feature of the car by this point, I can easily go back to a previous phase, rework the body to a different or better shape, and then import that new mesh into my next phase.
The idea is to have the best possible single solid body to work from. Ideally, you can actually print the the base mesh in real life and use as your buck model or plug to surface the car.
Shrinkwrapping, when done correctly, really enhances, smooths and finalizes the final shape of the car before cutting. There’s a lot of detail that goes into that which I will save for the design book I am writing.
This is is the last step of the design process where all the final details come into play towards really making the car more believable and more exciting to watch. Things like grills, vents, intakes, cutting out the panels, applying emblems, designing the details of the interior, etc. That all happens here.
I generally like to start another file specifically focused on creating the interior design assets. Im still learning how to create detailed interiors, so this is a very new thing for me.
My philosophy when it comes to rendering is to eliminate as much of the noise reduction need as possible. Im finding that selecting good quality HDRIs and boosting the right amount of HDRI lighting helps me achieve those goals. Sharper lines as much as possible outside of the box before editing. I dont use any sharpening or edge enhancement in post. Going too sharp tends to bring out a CGI quality and less of the realism quality. I avoid sharpening images in post, but I do find myself boosting exposure a lot more (learned from reading articles about District 9) to add realism and compensate for low lighting if the HDRI looks too washed out before post. Lightroom to me is a very simple yet powerful tool in helping me boost lighting.
Every HDRI that I select and bring is saved in 4 rotations (0 degrees, 90, 180 and 270 degrees). This lets me quickly cycle through the angles of light to find the best look for the car color and angle. I find dawn or sunset lighting angles work best. Noon to direct above lighting generally was too harsh and unflattering for most automotive setups.
In lightroom, I have different settings depending on the scene and lighting situation. Darker “sunset” lighting has more of a boosted dark setting with boosted highlighting and slight presence. If I wanted a more contrast and bold look, I switch to another setting that leaves the image mostly alone and only boosts presence and exposure. In photoshop, I generally boost the exposure a bit more with the “Curve > Lighter” setting which setup the images just right for me.
Conclusion
The most important takeaways from this experience is that it is extremely worth taking your time and not rushing any detail. The whole process feels like a highly calculated performance. No aspect was slapped together or rushed. I greatly learned the value of not releasing a design until it passes every stage within yourself. And of course that multiples once a client and engineering is involved.
I approached this project like I was an entire design studio in one. I took the design through various design reviews, I really looked at it and I wasn’t afraid to throw away the design and start over or move a few steps back. So this was strictly a research project for me to get me ready to move into more serious design roles with certain clients.
I struggled a lot with the interior and I put it off for a long time. But then I decided that you will only be able to overcome that fear by practicing and doing. Although Im very far away from creating the kind of interior I wanted, I believe this is a good start. I hope this inspires some to create and not be afraid to create their own designs.
If you have any questions, I’ll be happy to answer as best as I can.
Amazing work… and big thanks for breakdown. I like that design is … believable, something which you can see on street… well not every street I see to many super - car concepts where artist go wild… and result is unrealistic and often awkward. This car look like something which you can actually buy ( if you sell kidneys, lungs, liver and couple other unnecessary parts ).
Thank you for showing your workflow.The result is looking really great.
One thing,if you only want to boost your highlights,then you could do it in the Compositing,with the colorbalance node and the ASC-CDL formula selected.(the ASC-CDL is HDR correction save vs the other mode was clamping IIRC).
I can see the ammount of work you put in, but there are many drawbacks unfortunatelly - in the model itself and rendering part.
About the model - why so big panel gaps? Like you can put fingers in them. Some do not fit together very well (trunk to rear bumper for example) - looks like Tesla early production
About the rendering:
Flakes look like noise and not like flakes - it’s really not good.
Ground shadow is pitch black in every shot, where any other shadow is not.It breaks the composition. Tyres are quite bright and shadow is so dark that it feels like the car is not on the ground. The fact that you can see fireflies/noise in it makes it even worse.
Shot 865 - why are the front wheels turned - does he wants to go offroad?
Shot 866 - the same as before, but worse cause you can see a turn up ahead but in opposite direction
Well I guess since the project is a total failure or “really not good” as you say, I don’t even know where to begin answering this kind of critique. I guess I have a lot to think about.
I suppose I had no business posting it in finished projects and should have posted somewhere else. Once I get past the burn, I’ll figure out a better design that hopefully meets your standards.
I’ve never said that it’s a “total failure” nor “really not good” - only some parts are like that (ground shadow, paint flakes).
As somebody who for the past 4+ years in automotive I’m rather sensitive about cars
Car design is a very hard job and there is very few examples of cars that are universaly liked.
I understand. Just a time and place for that level of critique. I don’t usually critique work an artist calls final, no matter how bad or good it is. I usually ask “would you like some input?” Because I don’t know the history or what that artist went through to get to this point.
Like nobody knows the hundreds of hours it took to get to this result and I just finally burnt out and decided to finish it because it was eating up all my time and I like to keep moving forward.
But when i see words like “really bad”, I mean then I guess it’s really bad.
For example, In the engineering office I work at, we pull someone aside in private and then rip their work apart but we encourage. Something that took me years to learn managing teams.
I wouldn’t do that in the conference room with all the engineers watching. It just comes off more like “you’re wrong I’m right, and I want everyone to know” and dosnt really help the learning process. It just leaves the person deflated. In private it feels like we’re actually working towards something.
Honestly I don’t post these to seek approval, which lets me choose what information I keep and what I don’t. I just enjoy sharing my work and learn at my pace behind the scenes. For whatever reason this car has had the most praise and critics of anything I’ve ever designed.
But I’ll keep in mind for the next project. I appreciate your feedback. Yeah obviously you’re very talented and hopefully someday I’ll be that good.
I see to many super - car concepts where artist go wild… and result is unrealistic and often awkward. This car look like something which you can actually buy ( if you sell kidneys, lungs, liver and couple other unnecessary parts ).