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.