Week 10 and 11

These past two weeks, we were given our last task before we break up for the end of term, the stylised tree project. Working with a similar style guide to the stylised village project, we had to concept and model two trees. Before we engaged in designing our assets our tutor informed us that this is a project, which unfortunately may result in dull looking trees. Therefore, to work around this our tutors showed us how trees and nature itself doesn't have to look like the visual stereotype we establish within our heads when we think of what a tree is. In addition to this, our tutors took us out drawing trees from life to understand how they grow and how gestural they are in reality.

I was quite pleased that we were offered time to do this as I do feel that my time on the course I have been relying on photo reference too much rather than first-hand experience. As a result of this I feel that I am limiting my visual tool kit when it comes to rendering natural objects. This feeling originates from the fact that I feel that personally when I draw from life I really have to look and analyse how the form is made and how it fits in perspective wise, however when drawing from photos I feel as if I'm trying my hardest to copy the form not understand what it means.

When it came to the concepting stage of the tree, I felt as if jumped in too hastily and didn't really think about how a tree looks and grows. This can be reflected in many of my early silhouettes as they read little as a tree.

As you can see above the designs seem like random strokes and convey very little tree form language. One visual characteristic of trees I learnt at the end of this project is that trees may start with a thick base and taper towards their ends creating smaller branches in an asymmetrical pattern. However, this is not evident in my early silhouettes, which is why they read very little like trees. In addition, this is also due to the fact that my approach to producing these silhouettes were to just lay down some simple strokes until some form of image was produced. This however was not the best option to go by as evidenced by this result. This process in the beginning reflects how I need to take my time with my designs and put more research into what I am drawing.

Initially I had already had some concerns on the readability of my silhouettes; however, my concerns were only justified when I posted my early drafts on DMUGA. As the criticism I received were to make my silhouettes simpler and read more as a tree.

As I progressed, with my silhouettes I attempted to make them more tree like to follow my criticism however, I was also trying to balance creating a striking silhouette. This became my main point of error as my silhouettes started to look less like trees and more like people. However, as the image below suggested this was not all bad as I could take this idea of people like trees and push it further.

On DMUGA I was recommended to look at the gesture of dancers and combine that with the shape language of a tree to achieve the result of a stylised tree more easier. This lend me to produce more silhouettes based off of dancers which I felt were both stronger and easier to produce as I wasn't too focused on making something overly stylised as I had reference, instead I had to focus on how do I make this movement more tree like.

My silhouettes following on my crit

After producing my silhouettes, I went on to choose the strongest ones and produce value studies to bring out the volume in the silhouettes as well as add foliage to understand how the silhouettes will translate as an actual 2d concept and an eventual 3d model.

I then produced colour studies of my two trees seeking to choose a harmonious colour palette however if you can see in the image below my two first iterations were scrapped as evidenced by the red cross above them indicating how wrong they were. The reason why they were scrapped was that I felt they were too psychedelic and didn't follow the style guide of natural colours. I wanted to keep these scrapped designs to reflect my progress to the right design as well as to remind myself that it is okay to make mistakes as quite often I feel that I beat myself up a lot when it comes to my own artwork especially when approaching a new project. I tend to put them off sometimes out of fear of failure. 

During the colour study stage I felt as if it was time to improve the synergy of my workflow with 2d and 3d mediums as this was one of my shortcomings with the previous village project. Hence while was producing the colour and value studies I was already modelling a base of my model in 3ds max. This would allow me to iterate on my 3d model constantly making sure I don't end up making a 3d model to the exact look of a 2d design only to find out it doesn't translate so well in 3d like how my blacksmith building did in the previous project.

I modelled the trees using a spline and loft modelling technique I had learnt from game production. Adapting to this method of modelling was quite hard as I would run into many obstacles that caused problems such as random floating verts or deformed geometry.  I wanted to get a base model done for one tree at the end of the week so that I can spend the next week on the 2nd tree. This way I, wouldn't have to worry about concepting more ideas of a tree.

Around the start of week 11 I finished the base model of my first tree which shall be referred to as tree 1 as this post progresses. Around this time as the image below evidences, the base model of the tree didn't read quite well as a tree at first. Hence to resolve this problem firstly I started to analyse more form language of trees by drawing over them and understanding there gesture and movement. From then on, I produced a paintover, which aided my additional modelling.

While working on tree 1 I was also working on my second tree or tree 2 using the loft and spline technique. For tree 2, the loft and spline technique didn't go so well due to the complex nature of my splines shape as seen in the image below.

To resolve this issue I used an alternative method of first creating a plane and using the extrude along spline function with my shape. During this project, I found that I had to constantly find resolutions as there were many problems I would face. This project taught me to be much more independent with my thinking when it comes to 3D and use more of my mental toolkit I have been building over the past term.

At the end of my project I managed to reach the target of producing two trees however there is a massive leap in quality in terms Tree 2 to Tree 1 as I feel that Tree 1 reflects stronger design in both 3d modelling and 2d textures as well as how I handled the use of alphas to create foliage. This was mainly due the fact that when I first started developing the base of Tree 2 I then started to focus on modelling and texturing this tree first as to Tree 1. Therefore all of my mistakes are reflected in Tree 2, as this was my first attempt at painting a bark texture as well as using alphas. By the time I got to modelling Tree 1 I became used to using alphas and I started to understand how to create a convincing bark texture. One of the techniques I used for Tree 1 after learning how to implement alphas properly was to duplicate my alpha plane and rotate this duplicate plane at an angle and repeat this many times until I create a circular or three-dimensional look with these planes. This way I can give the illusion to audiences that the alphas look like a volumetric form rather than being a flat plane.

Tree 1

Tree 2

Bark for tree 1 - made with more understanding on how to render bark texture convincingly

Bark for tree 2 - made via trial and error if I had spare time I would re texture tree 2 as well as apply what I had learn t from using alphas from tree 1 to tree 2

Overall, this project and the projects before have really pushed me to build and understanding of design and 3d modelling and I can't wait for the next projects to come in the new year.