Thursday 23 April 2015

The late character post mortem

This post will be dedicated to writing an extended piece for the character project. This will allow me to evaluate the good and bad decisions made in this project. I will write an action plan regarding my feedback received from the tutors and also how I would like to move forward and develop as an artist in the future.
The character project was the third and last part of the stylised village. My aim was to create a character using myself as reference for the body and face. I chose to create a character that looked like a medieval milkmaid, because I thought that it would be the most suitable option for the village style. I had to follow a set style guide, the same one provided in the previous parts of the project such as the house and tree.
PROJECT DETAILS
I was given 5 weeks to create the character. I tried to evenly distribute work throughout the weeks and do as much as possible before the deadline. The character had to be set to be 180cm-200cm tall in 3DS Max. The texture had to be 1024x1024 32bit targa and a second targa to use for props sized 512x512. The project date and duration was set from the 26th of January and the project conclusion was in the 27th of February. The triangle limit was under 4000, and we were provided with an additional 500 triangles to use for any character props. The brief asked to rig our character and there was a bone limit set to 60 bones and 28 were the minimum.

Before the whole idea generating process began, we had 3 weeks of bootcamps to learn the basics of human anatomy, faces and also clothing, folds and material. Those weeks were spent to thoroughly research and gather as much material as possible. The sources used for gathering imagery were mainly from books when I researched anatomy. Also I had my first attempt at modelling a low poly female figure with references provided by the tutors. During the facial structure research I had to take pictures of my own face and create a head model, which I later used for my characters model. On the last week of research I focused on researching clothing, materials and folds to have a better understanding of how they work and how they change while a person is wearing clothing. 

It was an individual project and I had to evenly distribute work along the 5 weeks given on the project.
Week 1
Because I already had a moodboard ready, I was able to instantly start generating ideas for my character. I chose to photobash my ideas instead of doing silhouettes for this project because it seemed like it would later on be easier to understand how clothing folds and creases if I applied photo images onto my characters body. I researched paintings and clothing that milkmaids wore during the medieval times. I also had a wide array of different images in my moodboard displaying real life and fictional medieval clothing. My concepts were reviewed almost every week and the first feedback wasn't very good as I was told that there was no clear connection of where my ideas came from. Realising the mistake of not annotating my moodboard properly I fixed my mistakes and circled out the images that I thought were interesting and moved on with developing more ideas.

I chose 3 ideas to develop further and test their silhouettes. Doing so allowed me to decide which idea suited the brief most and looked visually interesting.
Week 2
I progressed with work by choosing a final idea and then added some interesting details from other idea itterations to further develop the chosen milkmaid concept. After some more silhouette testing, it was time to do an extensive array of value search and colour tests. This time I had to do 20 different value examples and also 10 colour searches using palettes from medieval paintings. Doing so allowed me to choose the colours for my character that looked good and fit the requirements of the village.
Week 3
Time was already running short and I had to start modelling the character. I started with setting up my own reference planes for the first time. It was quite difficult at first, as most of the images did not line up perfectly from the side and front. Then I modelled a low poly figure to start with using strip modelling and added the previously made head. After attaching I added in more details to the body to make it more smooth and human like. Later on it was time to start extruding out clothing and dress up my character.
Week 4
On week four I unwrapped the model, it was quite difficult because the face had to be unwrapped as one object. But it all became clear and easy once I figured out how to us pelt unwrapping. I continued on with texturing the character. The character needed to be handed in for reviews and personal feedback that I received in week 5.
Week 5
It was the final week of the character project and I was running out of time. During the feedback I was criticised for the character being very flat. The character also did not look feminine, also the topology of the character was wrong in quite a few places. To add my character concepts and ideas did not look very clean or esthetically pleasing. I had to quickly act on the received feedback. I waned to finish up the texture and character to the best level I could. Sadly because of that I did not have time to rig the model.
What worked well?
Although I wasn't very satisfied with the final outcome of the character, I think that there were a  couple of things that worked well. Like the head for example, I think it looked quite well and detailed as I put in a lot of loops and spend a lot of triangles on it. I also think that the way I unwrapped the face made the texture look very good and closely resemble my face. Another thing that worked well was the colour palette I chose for my character, as it was warm and happy similar like the stylised village. 
What went wrong?
There were many things that went wrong on my character. There were multiple errors with the anatomy of my character, like the size of the hands or the shape of the neck and waist. The clothing on the character looked very flat and boring. I made the poor decision of modelling my characters bad as the same object. Because my model wasn't very good my texture did not help in any way to make it look better.Given an opportunity to improve and change things I would completely redo my character and distribute topology evenly on my character to remove any kind of sharp edges. I would also. I  would change my design and chose a different colour palette. And also add hair to my character to make it look more feminine. One most important thing I would do is manage my time better, because my poor time management has taken away a big chunk of things I could have learned and developed.
What changed?
There were a few changes I have made to my character. One of them was that I did not model the braids I was supposed to as I ran out of time and wasn't able to figure out how to.
What would you do differently?
Knowing what I know now I would prioritise managing my time. Doing so would allow me to do better cleaner more detailed concepts. Also I would be able to model a better character and do a better hand painted texture. I would also give more time for researching clothing and folds.
Conclusion
For a meaningful conclusion I would like to say that the character project was very challenging. One of the hardest projects so far. Acting on the feedback I have received I should prioritise learning how to model better characters by creating at least 15 of them in my own time. I would also need to improve of the visual quality of my character concepts. Creating multiple characters would allow me to better understand the human anatomy and also develop my modelling skills. I will start working on my concepts alongside the 3D modelling avoid any unforeseen errors that might occur and this way save more time. 

Week 30 Final week final hand-in

My first year on the Game Art course approached it's and this week has been one of the busiest weeks this year. There was so much to finish, so much to do. Time was growing short and I had to quickly get on with continuing the zombified version of my Toyota truck. 
Modelling the truck wasn't as difficult I thought it would be, so it did not take me very long. Although while modelling the zombified part I did make slight changes to the final design I came up with. The problem was with the smoke launchers as I had them placed at the front of the car in the final idea concept. Modelling I realised that I had too many objects squished into one place and they were overlapping on each other. I would call this an error of not modelling concepting at the same time. If I had modelled my rough idea onto the truck I could have prevented making any changes before deciding my final approach to the zombie version car. 
Overall I am happy and unhappy with my final outcome. I think that the headlights, metal armour and the metal plate at the front of the car look very well. I also like the turn out of my texture. 
I also think that the smoke launchers on the car look out of place, because of the yellow colour. The reason why I chose to make them yellow was because I wanted to stick to the real life reference where smoke launchers are usually attached to dessert coloured tanks. I also think that if I had to do this project again I would develop the idea more and have more colour variations to use as reference before moving onto texturing. I would also change the colour of the metal bars I have applied on the windscreen in front of the car. Perhaps another thing I would change would be the colour of the base Toyota, to a lighter probably grey or white colour to make the metal parts pop out and be more visible. 
My action plan from this project would be to prioritise doing 3D modelling and concepting hand in hand to avoid future unforeseen errors and save valuable time. I will also again try my best to improve on concepts, make them cleaner more appealing to the eye. Add more life and colour to my art, explore more colour iterations and also values. Last but not least once again develop my ability to manage my time to do better in future projects. This year has taught me a lot and I value the experience I have gained. Now it will be up to me to develop my skills over the summer doing personal projects and attempting to redo the set projects this year.

To end my first year, I'd like to mention the lovely last lesson of life drawing. It was probably my favourite, although we did not draw any models. This time we focused on only drapery and how naturally it fell from it's weight. The focus here was the lighting and shadows. And the idea was to use as few lines as possible, rather put all the attention to rendering out values and tones. I was quite satisfied with this drawing as it showed how much I have progressed since day one of the course. I also am pleased with the way I was able to keep a strong contrast between light and darkness. One thing I'd like to add is that I still need to work on my mark making and also my range of values as I tend to only stick to very few.

Week 29 Phase 2 of the Zombie Toyota Pickup

The Easter holidays were over and it was time to get back to work for the last two weeks. This time I started the second phase of the Zombie Toyota Pickup, by creating ideas and concepts of the zombie killing machine version.
I have started with photobashing a couple of ideas together to have an idea of what a potential zombie truck would look like. I was interested in making less of a zombie killing truck, more like a safety escape truck that would protect anyone inside it and possibly help survivors.

Later on I took the quick bashed ideas and started developing the ones I thought looked interesting or somehow unique. In the image on the right you can see that I chose to develop 3 different ideas by adding or taking away parts. I chose car number 3 at first and then took away the parts I didn't like and added smaller bits from other zombified versions.

Having the idea sorted I moved onto bashing on
some more details to have an understanding of 
what the truck will look like from the back, front,
side and what will be going on in the tailgate.
In a way this was my reference image ready to 
start modelling on the simple Toyota Pickup.
I chose this idea as my final because it seemed most appealing to me. I also thought that the
car fit the brief because the front metal part serves as a weapon in a way, also it appealed to me because it had a lot of armour on it and looked a bit like anyone with the right tools could attach those parts to their truck at home.

I continued with searching for colour variations for the truck. There weren't many variations I could generate as the car had a lot of heavy metal on it. I tried to stick with realistic colours metal could have. I wasn't very satisfied with the variations as most of them didn't look very usable so I decided to test colours as I go onto modelling and see how the colours and values really look like rendering images in 3DS Max.
I did try to improve on my concepts by making them cleaner and more pleasing to the eye. I do think that the first bashes I have made look more improved, but the final idea sheet looks quite childish and lacking in detail. Also the colour iterations aren't very nice to look at as well. In the future I should put more effort and time into generating final concepts. The next steps to take will be to start modelling onto the Toyota.

Tuesday 31 March 2015

Week 25 Conclusion of Phase 1 of the Toyota project

This was the last week to finish the Toyota Pickup model and also it meant concluding Phase 1 of the project. The week was special because it also meant once I'll hand in my car on Friday I will be free to enjoy the Easter holidays.
Continuing on with the last part of the car model, it was time to start texturing. I've applied an Ambient Occlusion onto the car model, meaning the 3D modelling software baked in lighting onto my car. The Occlusion effect gave my model a more realistic look. Because I was using it for the first time, texturing in Photoshop required using a few more effects and tools to make the lighting work with images.
I used two copies of the built in lighting render of my models texture sheet, and places them one on top of another. The first one was used for reference where the car parts were places and the second sheet of the render had a multiply effect applied on it so I would be able to use the lighting effects. The rest of the processes were simple: cutting images, adjusting levels and values, applying effects and filters to match the car and give it a realistic look. My goal was to make the Toyota look like it's been driving through muddy roads and had been quite old with the paint peeling off.
I am quite happy with the final outcome and colours I chose for the model. I think the red body and white bonnet work really well together.
If I were to redo this part of the project again I would try and add in more details to the texture and also the car. I would also try and fix the windows to make them, look more reflective because I've tried to do tinted windows with some dirt on it. Also if I unwrapped the front and back windows as one object instead of half I would have been able to add some writing on the dirty glass or imitate the front windscreen was wiped with a hand.  I would also try and use up the remaining tris to add in more detail to the car.

Week 24 Unwrapping the Toyota

Ambient Occlusion applied to the unwrap.
The project continued and I've concluded the modelling process of the truck. I've lost some valuable time last week so I had to make up for it by working extra hours.
The modelling process took a little longer than expected, as I found it a little difficult to model the back of the truck and cover the bottom. Although the model was finished, I did not apply the symmetry modifier yet, because I needed to unwrap the objects first.

The unwrapping process wasn't too difficult, because I followed the example Toyota texture sheet provided by the tutors. The reason I chose to follow the texture sheet  was because it allowed me to learn how to pack the UVW's tightly on the texture sheet.

The project allows us to use two texture sheets - one for the body, windows and the second for the interior, chassis and wheels.
The interior was a stretch goal so I chose to use tinted windows, because I think the modified version of the car will have window protection and the interior won't be visible. That's why after consulting with one of the tutors I've decided to use a smaller second texture sheet for the chassis and wheels.
Now all I have left is to texture the model using photographs next week.

Longer head and face drawings.
In life drawing it was finally time to focus on drawing the head. We started with 10 one minute sketches of different sides of the head. We only had to focus on the shape of the head drawing the quit sketches. When we moved onto longer drawings, I had the time to focus on the facial features. I chose to draw light planes on the face and also covered the shaded areas in planes to make the face look more sculpted. The faces were quite difficult to draw, and I do have to improve on making them look more like the person I am drawing.

Critical studies introduced the requirements for a larger piece of writing that will be needed to be done before the end of the term. The essay will be about the first year on this course and an action plan on how I will move forward.
The tutors were kind enough to give us a very good plan for essay writing, which I am planning to follow.

Week 23 The Toyota Pickup

Toyota Pickup before texturing.
One project ended, another started. This week we started the zombie Toyota Pickup truck brief. The project will be made up of two phases - the simple Toyota model (3 weeks) and then we will need to zombify the car to turn it into a zombie killing vehicle (2weeks).
In this post and the following two weeks I will talk about the process of phase 1 of the project.
After having the brief introduced and a reminder of photobashing basics, we were given supporting reference and images set up in 3DS Max to immediately start the car model. I have made the decision to do the car model first and generate ideas of a zombified version after.
Starting the car model I had to follow a tutorial to create a Toyota that looked as close to the real life car as possible. I used strip modelling and started with the chassis part of the vehicle. Switching between the side, top and front views helped me hasten my work progress.
As the reference images didn't line up perfectly I had to keep checking the perspective views and keep fixing edges. I have modelled the wheels, chassis and the body as separate objects. One of the best things in this project is that only half of the body needs to be modelled, as the car is a symmetrical object. The only thing I will need to do at the end, will be to apply the symmetry modifier and it will save me a lot of valuable time next week.

Warm up one minute drawings.
Longer drawings, focused on more detail.
In life drawing we focused on hands and feet as a lot of students had trouble with the proportions and sizes of them in the previous character project. The aim was to try and draw either two hands or two feet in a short period of time. I found it quite hard to draw foreshortening on the toes as the model kept changing the positions. I also picked a spot in which the feet were in the same position most of the time. It was really fun to draw hands and feet, so I did not find it that hard to sketch and I am quite pleased with the end result.
Post mortems:
The busy week ended with critical studies explaining how to write project post mortems. The lecture was given at a quite good time, because we had just finished the character self assessments. I used to keep my assessments very short, but the lecture gave a plan which it be best follow writing about what happened in the project and how to improve the analytical part of the assessment.

Week 22 The end of the Character Project

The week was focused on feedback and I had to act on it after receiving mine. I do have to say 3D modelling is not my strong area, which was why I had quite a few errors on my characters anatomy and clothing. I can draw a woman, but I can't model a well proportioned female body. During the feedback I've been also told that the character looked quite flat in a lot of areas. To add my concepting and idea generation sheets did not look esthetically good.
At the end of the conversation I was given a task to blog on my feedback and make a plan on how I will improve on my work.


I have finished the texturing part on the character and I admit that it was the easiest part in my opinion. I think that the colours on the texture worked quite well as they are quite calm and harmonious. Following the feedback I also had to fix anatomical errors on the characters model. I had fixed the neck, waist, chest, skirt and sleeves as one of the tutors warned me that such round objects wouldn't deform properly when applying a rig. I did not have enough time at the end to apply a rig or skeleton on my character, which means that I will need to learn rigging on my own soon.
Action plan:
Another action plan for the year of 2015, this time it will be focusing on character modelling and concepts.
To improve my character modelling I will:
1. Start concepting and modelling more characters in my own time to improve my skills in 3DS Max.
2. Also I will try modelling different body types to learn more about anatomy.
3. I will model separate parts of the body to improve topology on fingers, feet and limbs.
4. Practising will help me improve on creating creases and folds in clothing.
5. Modelling characters will help with learning to apply a rig on the characters.
The second part of the my plan will be based on making my work clean and tidy. I will also practise digital painting and cleaning up my photobashes. Also doing colour studies will broaden my knowledge of what colours work well.
The last week of the character project was quite intense and it also opened my eyes that there is always so much to improve on, not only in creativity but also in managing and organising work.