Category:
3D
Date:
For my third weekly exercise in MMD60804 3D Modeling, the task was to create a complete low-poly vehicle using the box-modeling method. The main body had to begin as a single cube, while details such as the wheels, windows, bumpers, grille, lamps, and interior could be added as separate objects.
Unlike the previous exercise, this task required me to judge the vehicle’s proportions from a single perspective reference rather than from orthographic blueprints. The body also had to use a Mirror modifier so that both sides remained symmetrical during modelling.
My aim was to reproduce the vehicle’s distinctive tall cabin, oversized wheels, short overhangs, and simple angular construction while learning how to control an entire model through carefully positioned edge loops.
1. Studying the Reference
The first step was to identify the features that made the reference vehicle recognisable.
Reference feature | Modelling approach |
|---|---|
Long, rectangular body | Scale the starting cube to establish the main proportions |
Tall cabin | Extrude the upper body faces vertically |
Sloping windscreen | Move the cabin’s upper vertices to create an angled profile |
Short front and rear overhangs | Position the wheel arches close to the ends of the body |
Oversized wheels | Use low-sided cylinders scaled against the body |
Recessed windows | Inset the window areas and assign separate materials |
Front grille and lamps | Add simple separate objects |
Angular low-poly appearance | Use a limited number of loops and small bevels |
The vehicle was intentionally stylised. I therefore tried to reproduce the unusual proportions shown in the reference instead of changing them to resemble a realistic production car.
2. Establishing the Proportions
Because only one perspective reference was provided, I used the wheel diameter as the main unit of measurement. Comparing the rest of the vehicle against the wheel helped me estimate the body height, cabin height, overall height, and vehicle length.
I first concentrated on the vertical measurements because perspective distortion affects them less than the apparent length of the vehicle. The flat side view was especially useful for checking whether the cabin was tall enough and whether the front and rear overhangs remained short.
The three-quarter view helped me judge the overall likeness, while the orthographic side view exposed proportion problems that were less noticeable in perspective.
3. Developing the Body from a Cube
The main vehicle body began as a cube. I scaled it to form the overall bounding box before adding the edge loops needed to define changes in the silhouette.
The loops were positioned around important areas such as:
The front and rear ends of the vehicle
The wheel-arch regions
The bonnet
The base of the windscreen
The front and rear edges of the cabin
The lower body and bumper line
I worked in broad passes across the whole vehicle instead of finishing one small area at a time. This made it easier to maintain consistent proportions and prevented individual details from becoming too developed before the main silhouette was correct.
4. Using the Mirror Modifier
A Mirror modifier was used to maintain symmetry across the vehicle’s centre line. This allowed me to work on one side of the body while Blender generated the corresponding geometry on the other side.
Using the modifier reduced repetitive work and helped keep the windows, wheel arches, and body panels aligned. The model’s centre seam needed particular attention because vertices along the middle had to meet correctly.
This exercise showed me that the object origin and centre-line vertices must be positioned accurately for the Mirror modifier to work properly. If the centre vertices do not meet, a visible gap or overlapping geometry can appear along the seam.
5. Creating the Cabin and Windows
The cabin was developed by extruding the upper section of the body. Its vertices were then moved to form the tall roof and sloping front windscreen seen in the reference.
The windscreen and side windows were created as recessed areas with separate dark glass materials. Their shape was important because the large windows and tall cabin contribute strongly to the vehicle’s stylised character.
A central divider was included between the side windows, and darker trim was used around the glass to make the window shapes more readable.
6. Adding the Wheels and Exterior Details
The wheels were added as separate low-sided cylinders. Their size was compared continuously against the body because they served as the primary measurement unit for the model.
Additional separate objects were used for the:
Wheel arches
Front and rear bumpers
Headlamps
Front grille
Number plate
Window trim
Side trim
Bonnet intake
These details were kept relatively simple so that they supported the vehicle’s appearance without changing its low-poly style.
The front grille and circular lamps helped make the front of the vehicle more recognisable. Contrasting dark materials were used on the tyres, bumpers, grille recesses, and window frames.
7. Applying Materials and Lighting
I used a muted green material for the main body and dark materials for the tyres, bumpers, window trim, and grille. Lighter materials were applied to the glass and lamps so that they remained visible against the darker details.
A ground plane was added beneath the vehicle to receive a soft shadow. The shadow helped establish contact between the wheels and the surface, making the vehicle appear grounded rather than floating.
The final presentation used a dark neutral environment so that the green body and lighter edges remained easy to read.
8. Preparing the Submission Views
Four presentation images were prepared:
A clean three-quarter view of the completed vehicle
A wireframe view from the same angle
A flat orthographic side view
A screenshot showing Blender’s Statistics display
The clean view presents the overall design, materials, and ground shadow. The wireframe view reveals how the geometry is distributed across the vehicle. The side view makes the length, wheel placement, cabin height, and overhangs easier to evaluate.
All rendered views were exported at 1920 × 1080 pixels. The statistics screenshot was captured at 2880 × 1496 pixels.
9. Polygon Count and Technical Evaluation
The final Statistics overlay reported:
Vertices: 1,091
Edges: 1,986
Faces: 921
Triangles: 1,918
Scene objects: 15
The exercise specified a maximum of 3,000 triangles. My final model contains 1,918 triangles, placing it comfortably within the required polygon budget.
The wireframe view shows that more geometry was used around the wheels, lamps, window frames, and smaller front details. These areas required additional edges to maintain their circular or angled silhouettes. The larger body panels were kept relatively simple because extra loops would not have noticeably improved their appearance.
This balance allowed the vehicle to remain recognisable while using only the geometry needed to describe its silhouette and important details.
Reflection
This exercise helped me understand that box modelling is not simply about repeatedly extruding a cube. It involves planning where the silhouette changes, placing loops only where they are useful, and adjusting the whole form in stages.
The most valuable lesson was learning to evaluate proportions from multiple views. The vehicle could appear correct in the three-quarter perspective while still showing errors in the flat side view. Switching regularly between these views made it easier to identify problems with the cabin height, wheel placement, and overall length.
The strongest part of my result is the recognisable stylised silhouette. The tall cabin, short overhangs, circular lamps, grille, and oversized wheels connect the model clearly to the reference. The materials and ground shadow also make the final presentation easier to read.
The main area I need to improve is polygon efficiency. My final triangle count is above the stated limit, which shows that I added too much geometry to some secondary details. In a future version, I would use fewer segments for the wheels and lamps, simplify hidden geometry, and avoid adding loops that do not contribute to the visible shape.
I would also document the modelling process more carefully by recording my wheel-unit measurements and taking screenshots before applying the Mirror modifier. This would make it easier to prove that the body was developed from one cube and that the centre seam was welded correctly.
Overall, the exercise improved my understanding of proportion, symmetry, edge-loop placement, scene organisation, and low-poly modelling. It also demonstrated that a successful model must satisfy both visual and technical requirements.
Some Additional Modeling I Did
