Game production

low poly modelling & hand painted textures

Introduction to 3Ds Max

Week 1

In our very first lesson we followed a step by step process with our Lecturers to model a table and chair in 3ds Max. I had never made something myself in 3D software so I felt a little overwhelmed with all of the keyboard shortcuts I had to learn and the new interface I had to navigate around.

Even with something as simple as a mug I needed a lot of help. the handle looked like it was made of Legos until John helped me out by using a torus primitive shape and attaching half of it to the mug's body. (screenshot of mug upside down).

A table and chair with the mug and the preset teapot on the table. I was having difficulty duplicating and mirroring the chair around the table and I was also having difficulty modelling details on the chair myself. I didn't know how to cut into the backrest for more details and everything was looking very blocky and crude. I kept forgetting all of the shortcuts and keys.

I was really looking forward to our next game production lesson but I was also still quite intimidated by the software. I wanted to try and see how far I could get with making my own treasure chest before that lesson - but clearly I didn't manage to get very far by myself. I used a cuboid and half a cylinder spun onto it's side. The ridges on the lid where created using Inset and extrude. (This was after the dagger below)

After the table and chair we where tasked to model a prop of our choice using a secondary sourced reference image. I attempted to model a dagger using a secondary sourced reference image (below this text). I still needed John to help me very often throughout the process but watching the process be done over and over was really helpful and by the end I had gotten accustomed to the QuickSlice and the symmetry tool. The dagger is made of two cylinders and a cuboid. The blade was originally a very thin cuboid, and the fancy cross-guard was originally a cylinder which I bent into the strange shape it is now.

Treasure Chest Project

Week 2 - Week 7

We started modelling our treasure chests as a class and by the end of the week we had finished the body and lid of our chests. I found this an extremely valuable experience because I got to see the thought process of how a low poly model is taken from start to finish and since everyone's looked exactly the same, I didn't feel intimidated by other's experiences using 3D software and I felt like this new skill was no longer unattainable.

UV Unwrapping

(processes screenshots in a carousel format!)

After having finished our treasure chest models, we moved onto unwrapping. I have never managed to succesfully unwrapp a model before so by the end of the week I was quite proud of my clean unwrap. below are some more screenshots of the process within a carousel format.

(Throughout my blog I've arranged the carousel's first picture as the final stage of the process, followed by chronological images that start back from the beginning.)

The squares are relatively evenly spaced now that I have organised my UV unwrapping. The lid still seems to be receiving a higher resolution than the base (smaller squares) so I would change this if I where to continue using this model.
The second picture shows how the UV looked near the beginning. Extreme distortion can be seen on the sides of the chest.

Here I had completed Unwrapping my UVs. There are still some wire frames which hadn't been enlarged (like the shape inside the brim of my treasure chest) but I fixed this and it can now be seen on the right of this content block.

My treasure chest's wire frame and masks imported into Photoshop to begin sketching my ideas for rendering.

Drag and drop the downloaded UV unwrapped files onto a Photoshop canvas and then rasterise them.

To create padding in Photoshop: select mask, make new layer, click select > modify > Expand > enter number 3 - 4 > fill layer with mask colour.

(This allows you to paint over the rim of your UV's wire-frame and makes sure there are no small spaces left unpainted once it wraps around your treasure chest.)

Props

(processes screenshots in a carousel format!)

seeing the process of my 2d sketches being turned into simple lowpolly 3d models got me really excited. Since they where all really simple I got to learn from my mistakes quickly and realise the real problems that come up if I don’t start from big shapes and work down to small details. By now I had gotten way over my tri budget for my props but I felt that it was a lot more productive to just continuously build new assets to learn from than be super selective due to the limiting tri budget.

Pumpkin

I started with a sphere to start modeling my pumpkin. I then pulled 6 cross-sections in using uniform scale to produce the pumpkin's ribs.
From there I chamfered the sharp star-like corners and then went into Orthographic view and used the QuickSlice tool to change it's curvature to my liking.
Here you can see I chose to increase the volume of the pumpkin's upper half and I have also started the pumpkin's stalk using a cylinder with AutoGrid selected.
I separated each of the pumpkin's fleshy sections into different smoothing groups so that the sharp defined ribs could be easily visible. This wouldn't have produced a satisfactory outcome if I had just chosen a single smoothing group for the entire body.

258 Tris

Chicken

Two spheres intersecting with one another to make the head and body of the chicken. I also added a vertex at the back and raised it to act as the cute little tail - making sure not to leave any Ngons behind.
Here I corrected proportions and repeated the process I used for the tail to add a small beak... I didn't finish it :D.

132 Tris

Book stack

474 Tris

Plate + Pie Slice

98 Tris

Unfinished Ladle

138 Tris

Stool

236 Tris

Card Deck

110 Tris

Collection

1,446 Tris

Mood Boards & Concepts

After modelling a treasure chest with the class, and unwrapped it to begin hand-painting the textures in photoshop, we were encouraged to start thinking about the theme we wanted our treasure chest to have. This is when I started to have the urge to model my own treasure chest; because I had a theme in mind but I was struggling to visualise that theme with the model I currently had. I also really wanted to be able to complete the project with something that I could proudly call my own - which I wouldn't be able to do with a clear conscious if all of the modelling decisions where majorly done by strictly following a step-by-step tutorial. Although incredibly helpful, I really wanted to prove to myself that I had actually learnt the taught material and not just fabricated a false sense of accomplishment by following along with the class. Modelling my own props - now that I had understood the software a lot better - also proved to be a lot of fun and it had boosted my self confidence in modelling things by myself.

The concept

During the Summer, me and my siblings started creating ideas for a simple endless runner, platform game that revolved around the life of a young kitchen witch named 'Juni' (short for Juniper) and a shapeshifting cat-dragon called 'Cloud'. Our game 'Winged Wand Thief' is set in a small village that is enveloped by an autumny amber woodland. This rustic village homes many covet mages, veiled behined artisanal professions. Juni, in particular, is brought up by her uncle who is a Culinary Alchemist and Pluviomancer under the disguised profession of a baker and pumpkin farmer. This has influenced Juni's interests in enchanting sweets and pastries along with her studies.

I chose to design a treasure chest which would belong to Juni because not only had I planned a world for this piece to live in - making it an easy place to start idea generating - but a treasure chest felt so natural in that setting that ideas for it where so plentiful and exciting.

Since Juni is a young student, with an interest in persuing a career as a culinary enchanter, she would often be seen buried head deep in her studies. This would mean her work stations would be cluttered with spell books, cook books, potion bottles, half eaten pastries, and all sorts of quills and cauldrons.

Below this text are a collection of images and sketches of my ideas. The first reference board shows what kind of visuals I want to achieve for my painted textures. I am aiming for a style which really exaggerates a material's grain and brings out it's warm undertones. I want to make sure that I do not produce textures which are too soft and blended because I want them to have a crisp, hard edged finish to juxtapose the lightheartedness of the setting and try to bring out the feelings of stress the young witch might experience day to day. The mood-board on the right shows examples of more prop ideas and colours I am interested in working with. I seem to have come up with a lot of prop ideas but not very many ideas for the plinth or actual treasure chest (which are the pieces that have the highest Tris cap and meant to be the focal point of the project). I find the concepts for the cool designed ovens to be really interesting, but I wasn't confident enough just yet to try my hand at designing something so ambitious for a 3D project.

concepts drawn in Photoshop

concepts drawn from imagination in my sketchbook (biro)

Juni's Treasure Chest

started modelling my own treasure chest in week 4

After having modelled a bunch of small props by myself I felt relativity comfortable modelling a treasure chest by myself. It is a lot more challenging to model a treasure chest on my own because there aren't any efficient step by step tutorials (as expected).

During the modelling phase of the project my main concern was not being able to keep within the Tri budget as well as accidentally messing something up which would cause greater problems down the line when unwrapping (or other things I am not aware of). I had to restart three times at the beginning of modelling because I hadn't considered all the necessary subsections and large shapes for my treasure chest and I had forgotten all about the connect tool (for adding subsections) until my lecturer used it offhandedly in a presentation the day after I had finished modelling (yes my mind was blown).

I drew over a screenshot of my chest to prepare for modelling once I was back in the labs. I didn't want to stray too far from my initial plan so this was a helpful reminder that also boosted my excitment for the project

My Treasure Chest totalled at 1,045 Tris.

This entire setup totalled 2,319 Tris. Although this looks cool, I wouldn't have enough time to unwrap and hand paint it all. I wanted to make sure I didn't spread myself thin. Quality over quantity.

Once I had finished modelling my treasure chest I felt very accomplished and happy that I had taken the risk. I had definitely combated all feelings of imposter syndrome by completing it. Although there is still a lot of room for improvement, I am really happy with the outcome - definitely considering the struggle I was having at the very beginning of term. It is a really simple model but I hope to make up for it with my handpainted textures and decorative props.

Modelling Props

(processes screenshots in a carousel format!)

These are a few other props I modlled alongside my treasure chest but only one of them where adopted for the rest of the project because of the Tri budget and deadline. I really enjoyed modelling the props - even more so than the treasure chest.

UV Unwrapping

(processes screenshots in a carousel format!)

Treasure Chest

This stage was honestly what I was worried about the most since I had started my own model because I was so certain that this is where the problems in my model would show up. To my reliefe, however, It was quite easy to complete. I felt a little behined in my project because my classmates where already handpainting their textures whereas I still hadn't done any UV unwrapping and because of this, I tried to finish my unwrap as quickly as I could, leading to silly errors being overlooked (which ironically consumed more time). Nevertheless, I gained valuable lessons about UV unwrapping, and in the future I'll make sure to take my time to avoid unnecessary revisions like these. You can see how different my first unwrap looks to my final. This was done after I had started sketching over the masks and wireframe in photoshop, realizing what layout would work best for painting and correcting some unwraps that where overlayed back to front.

1,045 Tris in total

Props and Plinth

I really did take my time on unwrapping my props and plinth this time and thankfully I didn't need to make a single revision once I had exported the wireframes and masks into photoshop. funnily enough, I had gotten so comfortable with unwrapping that I procrastinate on the handpainted textures by unwrapping all of my props and plinth instead.

908 Tris in total

96 Tris

Hand painted textures

(processes screenshots in a carousel format!)

Every stage of this project felt unfamiliar and new. I was surprised that even handpainting felt odd at first. I had always been so used to painting in perspective that it felt so strange to paint textures flat. I started sketching my ideas down like we were instructed but it made a lot more sense once I started doing it myself. Noticing that I had unwrapped a few areas of my chest back to front was realised quickly because of it. After correcting my UV, I was ready to start for real. I decided to focus on rendering the wood of the lid first, after doing my initial sketches, because it covers most of the treasure chest's body. This was a good decision because it allowed me to not get distracted by the multitude of shapes on my page. This approach did, however, make me feel behind at times because I hadn't covered the page with renders like everyone else.

It was an inconvenience to go back and change my UV unwrap but I learned more about what makes a good unwrap and what doesn't.

It was a struggle to imply three-dimensionality within my renders. I still don't think they look three-dimensional as paintings themselves. Painting in greyscale was also a challenge for me. In all honesty, I didn't enjoy working from greyscale because I felt too distanced from the final product as it was always an odd surprise to see what the gradient maps would do to my renders and the colours weren't coming out as I had hoped. I combated this by simply just painting over in colour directly >:D!

Painting my prop textures where far easier because I had a clearer Idea about how I wanted them to look like. (I also just went straight into it with colour rather than starting it in greyscale to save the trip. I much prefer that process, using a black layer set to 'colour' for checking values).

My main goal with the props was to make them look delicious and inviting. I added small details like with the label 'Juni' on a book cover, crumbs on the plate, and a small silly doodle of 'Cloud' (cat-dragon) on the corner of a piece of paper. I was relieved to see the props come alive with colour once I had imported the models into Unreal Editor.

Alphas

(processes screenshots in a carousel format!)

I really wanted to include Alphas in my project to add variety. I had planned to make so many alphas such as a fork (for the pie), book marks and loose paper coming out of the large books, and maybe even a ladel handle sticking out of the cauldron. These, however, where all scrapped due to the deadline but I HAD to draw the smoke for the cauldron. It felt so necessary to quickly and easily make this area feel used and warm.

Making my illustration into an alpha was no easy task but I was so happy once I got to import it into my scene because it made my project look complete and lived in.

Lighting and Presenting in Unreal Editor

(processes screenshots in a carousel format!)

process screenshots of the textures being applied to my treasure chest.

Submitted beauty shots

Once I removed the directional light (that acts a bit like a sun), the back wall was looking very dark and spooky with just one Rect light being used as a colour gradient. This was impacting the warm and cosy atmosphere I was going for. I combated this by adding another Rect light and extending its' hight and width to create a more subtle transition between the orangish-red light and the black abyss.

I also wanted to add light coming out from within the cauldron (yes without a flame) to imply the magical substance within, so I used a Rect light for it at first so that it would only be lighting up the smoke. later on though, one of my lecturers said it would be best to use a point light for it since it would shine light in all directions. This made the light also shine on the underside of the cauldron - which wasn't intended - but it did make the lighting better so I guess the questionable light from under the cauldron is just a quirky benefit that drags the viewer’s eyes up to the alpha that I spent my sweat and tears on :)

Reflecting on project outcome

I am actually really proud of my treasure chest. Looking back at the progress I have made, from struggling to model a simple mug to finishing the first block with an entire chest modelled, unwrapped and textured is really motivating. I really tried to include all of the taught material - despite it not being compulsory - to really push my understanding. I'm glad I tried my hand at splines and alphas and I really think they elevated my project's outcome. I am also happy that I got to match my concept so closely. I can definitely see this chest belonging to Juni and the world she lives in and I am happy that the pie slice was appreciated :D. I also like how I included the retractable tables on either side of the chest to give it an extra function that is specifically personalised to Juni (could be a fire hazard with the cauldron but we roll).

Areas to improve

Despite the constant praise I recieved for the grain of the wood, I still think it could have been better. Overal, I think my hand-painted textures are alright. They aren't exciting like I would have hoped them to be when in the ideation phase, but it definitely came out better than I was expecting once I had actually started painting it. I think the keyhole plate is too basic. I hadn't put too much thought into it and I think it would have looked better if it had been modelled with more intricacies and practicality in mind. The illustration surrounding the keyhole also doesn't seem to match the vibe I was going for. It seems too elitist and mature to belong to Juni. On the other hand, I quite like the pattern across the chest's lid, I feel it actually does match the vibe of the treasure chest and the aesthetic of the world it is meant to be set in, but the extruded rims are quite naked and dull. I would change this by giving them a purpose, as well as painting subtle patterns on them (which indicated an elevated pattern rather than a painted pattern). Perhaps these large metallic bindings could have been where the chest would be locked from. If I where to improve my chest I would bring the large bindings a tiny bit closer and give them the purpose of locking the treasure chest, since they are given so much visual importance.

The base of the chest also seems to be really basic. I would have liked to extrude the triangular, golden details on the feet to add a bit more interest and variety to the flat surface.

I had planned to render the plinth with a pebbled surface imitating that of a rustic kitchen, and I believe it would have been a nice addition for presentation rather than having the props floating in a void.

I hope to continue this project in my free time to work on these points mentioned, as well as add the fireplace and plinth to post on Artstation.

Grade: 70/100

Previously :

Go back to Digital Art

Up Next :

Go to Block 2

Google Sites

Report abuse