Building Lee

Appologies for the generic pose of Lee- I haven't got a rigged version back yet :) The making of Lees face took two attempts. Apparently, constructing a face is one of the most complex objects to model according to people in the industry, and I defiantly agree. As the characters faces have a lot of screen time, obviously they must be well built. We followed this tutorial when building our characters, as it provides a method that includes good edge flow to your model, which is important for when a face is rigged, to avoid unwanted creasing etc.

My first attempt was dismal unfortunately, as I began cutting in too many lines to add detail to the nose, completely excluding the mouth area which I assumed I would fix later. However, I soon realised when I got to that part, that this would be impossible as there was just too much detail on the mesh. There was no way I was going to be beaten by my own character and so I proceeded to have a second go, making sure to adjust every vert of each new edge that I added. In fact, this advice can be used for any model built. It hadn’t taken long for me to see the original face was wrong, so I still managed to keep to my time schedule. Once the basic face was modelled out, adding detail was not a problem. You can’t fix something if the foundations are wrong.

I then went on to build Lee’s clothes. I started with the body warmer as I thought this would take the most time to construct. I wanted to make sure the puffs weren’t too puffy because as her body warmer is yellow I didn’t want the audience to think she was wearing a life jacket! The collar of the jacket took a while to look the way I wanted it to but it was just a case of adjusting the verts accordingly. I chose to include strips of material that would hide mostly where the zip line would normally be, as I have found on previous characters that making a zip line is a lot of hassle to get looking right, and when its on right people hardly notice, but if its wrong everyone notices. So the less zip on show the better! In the images above you can see that although I have modelled the zip handle, there is no zip line at all, as I am going to model that in the programme Mudbox. This software will allow me to add more detail to my model. In a previous project I had used Zbrush for this same reason but as Mudbox is made by the same company as those who create Maya the controls are very similer and so should in theory be easier than Zbrush. This is a link to some really clear tutorials for anyone interested in learning:

http://www.3dtotal.com/team/Tutorials_3/video_mudbox2009/sculpting01.asp

When Dan redesigned my character, the belt he included was to be made of ribbon. However as I was creating the belt, I felt that if Lee were real, the belt she’d be wearing would be made of plastic, with a shiny plastic bow, the type that I would think young, trendy girls would buy from Claire’s Accessories. I spoke to Dan about it and he didn’t mind me changing the design of the belt :)

I then made her top as I wanted to make sure that I had left enough space for her arm to come through the jacket.

The skirt.

The tights, like the face took two attempts as once again I added too much detail too soon, which made any errors I was making unfixable later on. My first attempt looked like a leg of lamb, was completely flat and so was unusable. My second attempt was a lot better I think.

With the body modelling done, my friend Charlie noticed that Lee didn’t have much of a figure, even for a child. I had been modelling the clothes so closely I had failed to notice that. Once again the lattice deformer saved the day, as I indented her waist a little and gave her more of a bum and curve to the back.

I also created some hair tests for Lee. As a group we had sat down and discussed how we wanted the hair to appear. We had all agreed that the Maya hair tool would be unsuitable for what we were looking for as it was far too realistic. As the modelling style of our characters was very chunky, we needed to find a way to incorporate that look within the hair. Dan R had an image of Poison Ivy and I had some images from the film Monster House, and it was decide to try and make hair that looked similer to these models. The only problem was how. I suggested Mudbox or Zbrush as we had used this before in a previous project but the only problem with this was that it might take too long to get the model to look how we wanted it to and maybe modelling it straight from polygons was easier? Myself, Jake and Dans all carried out tests in Maya, each with dismal results. Mine are shown below. I attempted multiple times, extruding a polygon along a curve. Out of the four of us, Dan C created something that at least covered a characters head, even though it looked like ribbons. It seems that in the end we will have to use Mudbox after all, although that is not our main priority at the moment. First, the environments have to be worked on further.

An example of Zbrushed hair, made on a previous project. Instead of such thin strokes, larger ones would be more appropriate for the look we want to achieve.

Once the characters were built we realised that there was another problem. Kitty. Although she had looked good on her own, next to Lee and Cho she just looked out of place. This wasn’t bad modelling from Jake or a bad character sheet from Dan. Kitty had just not converted well from image to model when next to Lee and Cho. As we didn’t specifically know why she wasn’t fitting in, we again spent a considerable amount of time huddled around a PC with the soft mod Maya tool attempting to squash her face into a more appropriate design. Dan C eventually took the model home and showed me two options. Image 1 or image 2. I chose image 1.

Easter Holidays

Because the dissertation took up so much time, there was no time to lose with the Easter holidays coming up. Apart from the actual Easter bank holiday, we came in and worked every single day. Myself and Dan R’s main priority was to get the characters built. Jake had already made Kitty and so was building the cave, whilst Dan continued animating.

Term 2

This term was mainly taken up with dissertation, especially for myself and Dan R who seem to write slower than Dan C and Jake- curse them :p. Although I thought I was doing well in it, after a tutorial with Kieran, he told me that to get a better grade I needed to research differently, which led to a lot of all night studying sessions with Dan R as he worked on his too. Hopefully it was worth all the effort that we went to, but as we wont know our grades for ages, there’s not much point worrying about it. This is not to say that no work on this project was covered this term, far from it :) (we officially have no lives, hehe).

The first objects that I modelled were the carts that transpost the girls around the ride. The cart with the rounded front is for Lee and Cho, whilst the flatter fronted one is for Kitty.

Lee's cart:

I was aware that the carts would be in around 90% of the film, and in many close shots, so they would have to be modelled well with no shortcuts taken. Although now that the models are complete and I feel I could model them quickly if I were to remake them, when I first began I found it a bit challenging and came up with some hideous results when trying to build the model out of a solid cube. I realised that before I could begin the modelling I should take a moment to think about what I was doing and reassess my design. The design I was working from was a quick pencil sketch that I had made in term one, and showed only the side of the cart. After making some rough sketches, I realised that the best way foreword both aesthetically and logically was to build the cart out of separate objects like it would be in real life as opposed to trying to build it out of one object and extruding sections etc.

Above shows some of the construction proces of the cart. You can notice some changes in the design as I modelled, for example a bumper around the cart. I had always imagined the cart with a bumper, but once it was on the cart, I felt it was out of place and didn’t seem to fit into our Victorian theme, so it had to go. The cart was also appearing too big and needed resizing. Although I had measured the width to fit Cho in comfortably, it was maybe a bit too comfortable as when I added the proxy model of Lee into the cart, it looked ginormous. You can also see the way the cart was modelled out of separate sections on one side then mirrored across to complete the other half. Not wanting the cart to obviously look mirrored with every bump etc symmetrical, I used the lattice deformer to bump and dent the cart unsymmetricaly.

Once Lees cart had been completed, it was now the turn to construct Kittys. Kittys cart had to be more compact from the front as the end of the film has an embracing sequence between the girls over the two carts whilst they are inside them. To construct Kittys cart, I duplicated Lees and then pulled back the front, again using the lattice deformer. This however crushed the previous floor, wheels and front bolts and they had to be reconstructed. The lattice tool was once again used to make sure that the model didn’t share the exact same bumps etc as Lees cart. Two panels on the front also had to be completly re-modelled as this design did not include headlamps.

Dan C also needed environmental proxy models to be built so that he could begin the animation in more detail. A proxy model is a quick rough model, made to the correct proportions so that they can be replaced later in the scene by the real models. Using this method, Dan could blockout the animation which he could use both for the animatic and the final animation whilst we worked on our models. I created proxy models of the actual ride, the helter skelter and the big wheel.

Dan R also created some proxy models, and combined with mine, and Dan C’s proxy models of the characters, were placed into the scenes that we would be using for our films.

As a group we huddled around a PC, as we did when creating the blockout environment, and decided where objects would go in our new proxy environment, taking turns to position objects. As before, it took some time whilst we discussed what would look good where and why.

Director decisions:

Whilst we worked on our individual tasks, I made sure that I was kept updated with all the work the others were producing, and made sure to show them mine for their feedback also. Usually we work together at Uni, but this term due to the dissertation, it meant that it was more convenient to work at our separate homes, where it was less distracting and we could concentrate on writing. I had to make some slight directorial changes to Jakes and Dan C’s work, but nothing major: Jake sent me some screenshots of the character Kitty he was creating where I suggested that her nose round the sides needed to be more defined. I also didn’t feel that her teeth were quite right in style and Dans and Jake agreed. After a few attempts Jake created teeth that much better suited our overall film style :)

The only thing of Dans that I wanted to alter was the shot where Cho gets her bum stuck whilst going through the fence as I felt it needed more exaggeration. It wasn’t a big deal, and as time was short it didn’t need priority. However, I told Dan that if there was time at the end, I wanted that pose exaggerated. Its not even the end of the project and he managed to find time to fix it for me :D

The medieval room also had to be axed. This wasn’t a directorial choice but more of a group realisation as we could tell that with the dissertation taking up so much time there just wasn’t going to be time for this room. The room also wouldn’t affect the narrative at all, so off it was axed!

Christmas Break.

Over the Christmas break, I kept in touch with Dans and Jake socially as well as with work :) During the festive season, Dan R modelled a hand, Dan C updated the animatic and Jake made a miket of Kitty as he felt this would help him later on when he would have to build this character. Apart from dissertation research gathering, I didn’t do much Maya work, except refresh myself with the softwares interface with some speed modelling practice, due to unfortunate family health problems that occurred at this time as well as my nans death and funeral. However, I am determined not to let personal issues interfere with my work once term has begun. On the bright side, we did get our grades back from the previous term, which were all really good so that's encouraging.

Term 1 Summery.

Looking back over my last post, it appears that I may have gotten a bit carried away with the Christmas spirit to be that optimistic! The dissertation took a lot longer than expected- but more about that in due course!

The following is a brief summery of term one:

As a group:

· Discussions on the storyline, editing it to construct a narrative that was better structured. Together we would debate our suggestions until we had created an overall story that we were satisfied with. Although it was important to me to keep the basis of my original story, I felt that editing it as a group instead of individually would make the film be something everyone wanted to work on and feel apart of.

· Thumbnailing a storyboard. Evidence of this can be seen on our group blog. Together we went through shot by shot, of what we imagined would be shown on screen. Everyone had equal input and although people would get outvoted on certain shots in preference to others, the task was successful, in my opinion, in constructing the basis of a film that the whole group would be happy with. Everyone was then assigned different areas of the film to storyboard up more presentably so that individuals outside of the group would know what we were talking about!

· At the same time that the thumbnails were being constructed, a rough blockout environment of the film was being created, with basic shapes. Dan C made the basic shapes and placed them in the scene under all our instructions, so we could get a better understanding if our chosen shots would be appropriate.

· Character moodboard. We gathered images together to decide on a style for our characters. The word to describe the style we are going for is ‘chunky’.

Individually:

Concept art and research into visual references etc as I had a strong idea of what I want this film to look like. I also assigned the group members tasks and organised what we were going to do about sound, as I didn’t want to leave this to the last minute. For this project I only want to work with people I know I can trust with their working skills and reliability. The sound designer that I have chosen to work with is called Jamie Spinks, who like Dans and Jake, I’ve worked with before and know is up to the task. Conveniently, he lives down the road from me and has been dating my cousin for the last few years, so hunting him down if he starts to slack shouldn’t be much of a problem! :p I also created the group timescale and drafted out the group members contracts.

The group worked really well together this term as a team, and we managed to get alot of work done. I sent Dan Ryans character designs back so many times, because even though they were good, there was still something missing. He didn't moan once, and now we've ended up with designs everyone is really happy with. So hopefully we will all continue to work as strongly through the coming months :)

End of term 1.

Wow, term one has finished! We ended the term by presenting our work to the class. Our group went first. It was a little nerve wracking but I think we pulled it off well- we got good class feedback which is positive but you never quite know what your tutors are thinking... until you get your grades back :S

Ok, so now it's Christmas break. I'll probably put this project on hold over the holidays and work on my dissertation instead. Last term, so much time was spent on our essays and presentations that there wasn't much time to spend on our project. I figure that if I at least get 3/4 of a first draft completed of my dissertation that then I can spend the majority of my time next term working on this project :)

Have a great Christmas!

Contract and Timescale.

Because time was slipping away, I wasn't sure if I would have time to completely design and make the contracts for the group. I wrote down some points and gave them to Jake to make into a contract as he could fit it into his time schedule. Below are my points, however Jake added a few of his own, which seeing as they gave me more power I decided to leave in as the group seemed happy!

Original Contract Points:

By the signing of this contract, I agree that:

1. This will be the main project that I work on.


2. I will not abandon the group. In the case of wishing to leave permission must be obtained by the director.


3. I will put my full potential and ability into my work.


4. I will not take credit for work that is not my own.


5. The directors decision is final.

I also created a timescale way back at the beginning of this project, but until now, forgot to upload screenshots of it into my blog. The full version is located in our groups public sharing folder and is updated every time we complete a piece of work.