One of the biggest challenges that I have been presented with this semester has been the task of creating the boy model for the animation.
I returned to university this semester to increase my abilities in environmental design and modelling, so admittedly being given the responsibility was not ideally my first choice. However, as Rachel was the person on the team with the most 3D experience during her experience at Donkey Crew for placement, it was thought to be of benefit to the entire team if she were to create the environment.
Previous attempts at character modelling have never been successful for me, and as a result, I felt the additional pressure of creating something so vital to the core of the project. The boy is the protagonist of the piece – the story is a tale of events which happen to him, and he features in both settings within the animation. To summarise, nearly every single shot would be likely to feature the boy in some form. No problem, right?
(E. 1 – Shown above, the last character modelling attempts)
To include context at just how poor my prior experiences with character modelling have been, I have included two examples which are shown above. The image on the left was supposed to be the image of a fly, whilst the image on the right was supposed to be a ladybird insect. Despite the fact that these models appear to have been created within a short interval, what you are viewing was the outcome from hours of struggling and frustration. From the second semester of our first year, I have never again attempted to model a character as my confidence had really taken a knock during that period.
Understandably, I was experiencing a lot of fear at the thought of being unable to produce something of decent standard yet again. The team was relying on me to achieve this task, as we had such a substantial amount of work to do between now and the deadline in January. Whilst it was still late October, I was aware that time had to be permitted for me to understand how to create a character model that was both true to the concept, and most importantly, functional.
At first, I was eased into the task by first creating a block shape model that could be used as a size reference, both against the bear and as a marker within the environments. In a sad admission, I do have to explain that the fact that the block out of the boy resembling a somewhat human form felt like a major achievement at the time.
(E. 2 – Shown above, is block out model of the boy. Also featured is the bear block out that Katie created)
Once the block out had been created, it was time to proceed towards creating the real model. My first attempt was unpromising, and as you can see from the image featured below, my workflow was self-destructive. I was focusing on creating shape too early, rather than prioritising anatomy and shape of the boy character – this resulted in a very poor attempt at creating the torso. Apparently, my subconscious discovered this fact too as the file seems to have been deleted from my computer.
(E. 3 – Shown above, the first attempt at modelling the boy character)
Admittedly, it was around this time that university work skidded towards a sudden stop, as my sister gave birth an entire month premature. Due to the commitments at home, I was honest with my team about how I would be unable to proceed much further for the next few days as my time required gathering items for my nephew, and factoring in daily hospital visits.
In retrospect, I feel that the time apart from the project brought some clarity regarding my workflow to create the boy model. Whilst I had been diligently reviewing tutorials (featured here), trying to imitate their workflow was not successful for me. Therefore, I adjusted my focus to examine their steps rather than repeating them so that I could understand the purpose behind the decisions being made, and make that applicable within my own process.
(E. 4 – Shown above, from left to right – the process that I made creating the base model in Autodesk Maya)
The process ran a lot more smoothly during my second attempt modelling the boy. Shown above is the stages that the boy went through during the blocking out period. Admittedly, my lack of skill in character modelling made the intention of replicating the original concept of the boy within a three-dimensional form to be very challenging. As the concept of the boy had been posed, I created front-facing and side body silhouettes to use as a guide during the blocking stages.
(E.5 – Shown above, the body silhouettes that were used during the blocking stages)
However, upon reviewing my process using the silhouette guides, I discovered that I had strayed away from the original intention of the team – the discussion was that the boy and the bear would be polar opposites in terms of design. The boy is small, skinny with an oversized head, whilst the bear is large, chunky and with a more realistically proportioned anatomy. The third model that is shown in example four (E.4) illustrates how reliant I had become upon my reference during the modelling process. If the original design had been conceptualised as a larger child than this would have been fine. However, this outcome was not true to the character design, and therefore, I sought a new method to correct that issue.
(E.6 – Shown above, the boy (drawn by myself) shown with the bear (drawn by Katie) for a scale and proportion reference)
The solution was accompanied by a new challenge – being thrown in at the deep end by learning entirely new software. Whilst Autodesk Mudbox had been suggested to me because as a relative to Maya, one would expect that it would be easier to navigate, I elected to explore Z-Brush Core. During the summer, Katie and I decided to split the money between us to purchase a single use license so that we could experiment with 3D sculpting in preparation for returning to
During the summer, Katie and I decided to split the money between us to purchase a single use license so that we could experiment with 3D sculpting in preparation for returning to university for final year. Unfortunately, the circumstances that had occurred during my summer prevented me from acquiring time to do so, but I was determined that I would find the time to learn as sculpting felt to be more artistically friendly than modelling within Maya.
Personally, Z-Brush appeared to be the most efficient way for me to correct the proportional mistakes that I had created during the blocking out process. In example four, there is evidence of my attempt to block out the shape of the hair to create the necessary volume that was aesthetically endearing within the conceptual designs. The size of the hair and the complexity of the tight ringlet curls would be incredibly challenging to create using only Maya software. Therefore, it felt that the most appro
Therefore, it felt that the most appropriate choice would be to tackle the new software, as the technical tools provided within that software package would enable me to replicate the conceptual design of the hair, transforming it from a two-dimensional imagery to a three-dimensional reality.
The risk in learning new software is finding the time to perfect it. Whilst I felt that I was being innovative by using software that I had no prior experience with, yet could evaluate that it would be the most efficient way for me to achieve the conceptual design, it did slow down production speed significantly. However, the decision was made after weighing each option – to continue in Maya, and only achieve a somewhat successful replication, or to explore a new software package that, with practice and experience, would enable the team to have a true-to-form model of the boy.
(E.7 – Shown above, the result of the boy once he had been resculpted using Z-Brush Core)
To learn more about sculpting techniques and workflows, I watched process and tutorial videos, which are featured here within my research post. Learning by example aided my ability to pick up skills quite quickly, and through the process of trial and error – within a week and a half, I had managed to achieve the result which is shown above as example seven. Z-Brush Core was not the easiest of software to contend with, and during the sculpting stages I had to visit the Apple store quite frequently as some of the hot keys were inter coding with my Bamboo graphics tablet, meaning that my computer had to be reset manually by the geniuses – this resulted in more time set-back, which was frustrating as my own lack of proficient speed was costing the team enough production scheduling with the calculated risk of utilising new software.
It was near the end of the sculpting period when I ended up in a conversation about Scott Gill regarding the peer work process and evaluation stage, during which I expressed my concerns over the ability to rig the boy character. I had confided in him that I have had no previous experience with rigging and that whilst there are assistance packages such as Advanced Skeleton that would enable me to construct one efficiently, I had yet to find the time to research thoroughly into those services.
Scott kindly offered to create a rig for the boy, as he expressed that he would like to gain some more experience as it is something that he would like to specialise in. The team agreed that Scott creating the rig would ease some of the pressures that we were experiencing regarding the production schedule, and would greatly benefit using that time to work on other production elements. The next week and a half had been spent collaborating with Scott on the next stage of the modelling production: the clothing.
(E. 8 – Shown above, the boy outfit in various different poses)
In order to understand what it was that I was trying to create with the boy’s outfit, I felt that it would be appropriate to first sit and work out the garment from various different angles to explore how the design would work as a three-dimensional concept. The boy appears to a few years older than intended within these designs, but the priority of the task was focused on the outfit itself – perhaps applying these poses to blank models would have been the more appropriate choice, as that issue does deter concentration from the main objective.
(E.9 – Shown above, clothing practice with an older, and more recent version of the boy model)
During the periods that I had been left without my laptop computer due to it being repaired in the Apple store, I spent the time experimenting with character clothing. I began by fleshing out general shapes on an older, chunkier version of the model before exploring with delicate, lighter fabrics and materials on the more recent model. Once I had completed my character base sculpt in Z-Brush, the next stage of production was creating the outfit.
Scott explained the best way to optimise the boy model in order for him to rig it, which was to use the clothing as the body of the boy, rather than have an underlying mesh beneath the garments of clothing. He showed me an example of a model he had worked on during his time with Enter Yes (formally known as Black North) which really helped me to understand exactly what it was that he required.
The model was reimported into Maya (see example seven) before clothing was modelled around the existing shape. The clothing was modelled in stages, using the base shape of the body as a proportion guide. I had investigated the potential of creating fabric clothing within Z-Brush, but it required skillsets to be higher than novice level in order to create believable material. Therefore, reverting back to Maya for the finishing steps seemed to be the most suitable approach for my technical abilities and knowledge.
The process was time-consuming, similar to that of the rest of the previous stages. Clothing within Maya creates a unique property of challenges – the most significant is making block geometry and manipulating the mesh until the outcome somewhat resembles a fabric material. The most challenging areas were the parts of the outfit that had creases – the scarf and the socks. The material was constructed slowly, with a lot of intricate placement of vertices in order to create an aesthetically appealing shape. Matching the scarf tail to the snood geometry was a very intricate process, whereas the socks were manipulated with soft select to achieve a hastily pulled up piece of fabric, emerging from below the cuff of the boot.
Once the clothing had been completed, the final stage was deleting the body mesh from underneath the garments, and to model hands facing flat rather than in a rotated position, which I had done initially. The hands were a difficult piece of geometry to construct – our design had realistic digits. Therefore, modelling four fingers and a thumb that had to be small in order to appear childlike took a few hours, and two attempts in total.
(E.10 – Shown above, the first and second attempt at modelling the hands)
Once completed, the boy was given one last review from the team before being presented to Scott in order for the rig to be constructed. He was delighted with the way that it had been created, which had been pleasant feedback to receive.
(E.11 – Shown above, a turnaround of the completed boy model)
To summarise my thoughts about creating the model, I would have to express that I found it to be the most challenging task that I have undertaken so far within my university career. The process really pushed me out from my comfort zone and forced me to face my fears over my poorly developed skills regarding three-dimensional software to achieve the goal of creating our boy model for the team.
I feel that my technical intellect has benefitted from the process significantly. Whilst my process was slow and hindered from external events occurring in my personal life during the time of production, I am still astonished at just how quickly my skills seemed to have developed within such a short time frame. I was able to use my intuition to research and investigate which software packages would enable me to achieve the intended appearance of the conceptual design as a three-dimensional model, before taking a risk of learning that software within a short interval in order to successfully produce that outcome.
The process has enabled me to become more innovative within my working methods, as I have been able to integrate different software within my processes in order to become more proficient, that has enabled me produce an accurate replication of the original design, including the aesthetic choices such as the petite body and tight ringlets. The biggest hurdle that I faced during this procedure was speed – whilst learning the new software sanctioned the knowledge of sculpting and design to be implemented within modelling, the time that it cost the production schedule was rather significant. In future, I hope that the knowledge that I have gained from this experience will provide useful for increasing the speed of my workflow and that with continued practice, it shall lead to the production of a higher quality result.