Thursday, 4 March 2010

Final thoughts and Evaluation.

We were very ambitious from the outset. We wanted to create a futuristic and realistic feel to the environment. We originally planned to build many more buildings in the environment. Looking back, I think we made the right decision to cut back on modelling to complete the project. Although this was one of many sacrifices we had to make I do not feel the project has suffered. In fact I feel we have accomplished our goal in producing a truly unique and consistent environment. We wanted the environment to be barren, desolate and isolated. I think we have achieved this with the environments large landscape of dunes and rocks. Although I was not entirely happy with the environment layer i created we had to make do with the programes available to us. Tom's use of lighting really improves each scene highlighting areas which we want the viewer to look at. The use of dust and sand layers add movement to the scenes. I am very happy with the visual style we have adhered to. Jared had concerns that we would not be able to maintain a consistency of style in the models we produced. However, through careful planning and co-ordination we managed to use the early concepts to our advantage and refer to them when building the models in maya.

Because we developed a storyboard early on in the project we were able to produce some sort of story which we could work around. Although a few of the scenes changed slightly the final product resembles the early sketches. Good planning has had huge benefits in helping us produce an environment which looked nice but also worked in a realistic sense. The good organization of the group and its commitment to the project payed off as we produced something which we set out to achieve from the very beginning. The group has had its difficulties but we have maintained a business like posture which has ultimately brought about the project's success. Another aspect which has been beneficial to the whole group is the fact that we have all helped one another when difficulties arose. That is to say that we have all had a go at different tasks whether it was lighting animating, texturing or modeling. The fact we tried different roles within the project means that we learnt a lot more than if we had assigned roles to different members. This also had a knock on affect on work flow. We did not have to wait for a team member to finish a job which would hold up production. Because we integrated with eachother as a team we assisted each-other to get the job done. This shows not only the friendship we had but the professional stance we had to the task in hand. Because we involved ourselves with different aspects within the project we learnt different techniques we otherwise would not have tried. 

The project itself was a huge learning curve as we had to combine many different skills to achieve a product which met the brief's requirements. We will take the aspects we achieved in this project and apply it to future projects. This is not limited to skills within specific programes but the way in which we conduct ourselves as a group when planning a project and participating in different aspects. Having a shared vision is key to the any project's success. The leadership of our director also played a huge role in bringing about the end result.

I am overwhelmed with the amount of work I produced while working on the project as well as its quality. I believe I have matured as a person as well as an artist that can convey the wishes of the director. My skills in various programs has improved ten fold especially within maya and zbrush. I hope to use these programs in the future when creating other projects. The project has allowed me to experiment with different programs and explore my artistic skills to a new level which will eventually be useful to an employer.

Although there were a few problems during production I am very happy with the outcome. The visual style is theatrical and immerses the viewer in an alien, hostile world. The sound effects and lighting enhance the moving images and bring new dimensions and weight to the environment. The work put in from team members has been outstanding and It would be a pleasure to work with such artistic and motivated people in the future. Thanks to Chris, Tom, Ben and Andy for all there hard work as well as encouragement throughout the project, nice one guys!

Sound

Once we had completed editing the scenes within after effects we then had to find approprite sounds to insert in the final film. This would add another dimension to the scenes and give some sort of weight to the objects within the environment. We acquired some sound effects from the sound effects hard drive from the shop. Many of the sounds were irrelevant to our project. However, we did find some ambient sounds like wind and sand. In order to find some more specific sounds we had to purchase them from various sound websites. Some sounds were costly, but sacrifices had to be made if the objects like the vehicles within our project were to sound realistic and believable. Although we purchased many sounds using money for our own pockets we acquired some truly awesome effects which we could sink to our moving images. This would enhance the environment as well as the story we wish to tell. We made the right decision in purchasing the sound effects as it has improved our work tremendously. 

Problems and solutions

The group had many problems rendering this week which was due to textures not being in the right place. We had to re apply textures for some of the scenes as some of the files were sourced on other group member's laptops. The problem was easily rectified. This was done by reloading the textures on the faces concerned. This hampered render time but the problem was resolved. We learned our lesson. In order for Maya to retain the textures when transporting files one must set the project within the maya directory otherwise the textures do not appear within the scene. This happened several times when group members transported files via memory drives. 

Once we had all the necessary scenes we then started to render the images as targas on the render farm. Targa files would enable us to display the scene as a layer within after effects. In after effects we could then edit the images and turn the static shots to moving images. We could also include smoke effects as well as adjust light levels and colours to suit our needs. We thought that by rendering the images on the render farm using cube we would be able us to quickly render scenes for their eventual transportation to after effects. However, we encountered many problems when rendering on the farm. The issues included files not being red properly by the cpus as well as the render farm taking considerable amounts of time to identify the files it needed to render. We also discovered that the render farm was only really compatible using mental ray. When we tried mental ray our scenes became very pixelated and unusable. To overcome this major setback, we decided as a group that we would render each scene on different laptops. The process was laborious but saved huge amounts of time. Whilst at home we could render of scenes and edit them during the day. By using all available laptops we maintained good, continuos workflow. This caused some issues when transferring targas and desktops became messy with files floating around.

I think we made the right decision by rendering scenes on our laptops instead of the render farm.
This was because as more people were using the render farm it became overloaded with information. If we had begun rendering in the final week, I think we would have struggled to finish the project. However, because of the teams good communication and Ben Keswick's excellent leadership we finished rendering ahead of schedule. Because of this, we were able to deal with any problems that arose in the post production process effectively. Although the quality of our work was reduced slightly because we did not use mental ray the work output was greatly improved. The decisions we made as a team enabled us to produce a film that met the brief's requirements.

we encountered some difficulties when it came to editing the targa sequences in after effects. These included loss of images because files were not named properly. I think if we were to do another project like thin in the future we will ensure that files are named correctly and were easy to find which would save time for other procedures.  

Saturday, 27 February 2010

Modeling the desert scene

Modeling the desert scene was probably the most important yet most difficult to achieve. There was great controversy within the group how the the landscape should look as well as its feasibility during render time. I proposed that the scene itself be modeled exclusively in Zbrush as I would be able to produce more realistic and detailed results. Although the quality of the models would be superior to other programs the polygon count would be astronomical. We may have problems when it came to rendering the scenes.

I started off by modeling the sand terrain on a disc plain. I used desert images as reference so I could replicate realistic mountains and sand dunes using custom alpha channels I created. I created these alpha channels using photoshop images which I sourced from the internet.

Once I was happy with the basic form of the terrain I then started work on the rock layer which I created using a 3d disc which I inserted underneath the sand layer. By using the subtools option in zbrush I would be able to create separate objects which would be fundamental when it came to texturing. Rather than texturing one plain where I would have to match the texture to the various components, I would texture each component individually. This would make texturing much easier and the overall finish would be more precise.

I was very pleased with the level of detail as well as the the quantity of polygons I was able to work with. I created the scene using a sub division level of seven which should work in maya. It was difficult to get realistic mountains protruding from the sand layer in zbrush, however I persevered using an alpha channel I created which created rocky outcrops. My main worry when creating the scene was creating realistic rocks and mountains. If the mountains were smoothed the rocky effect would be lost.

I knew I couldn't exceed the sub division level of 7. This would cause maya to crash and rendering the scene would be impossible. I had to realize that there were certain rules the model had to abide by otherwise it would hamper memory and texture editing.

I imported both models into maya to be textured. I had no idea of the problems the high res model would cause. Initially the models appeared to be working in maya's interface. However, when it came to applying the textures and aligning the objects UV's this caused maya to quit unexpectedly. Either the textures I produced or the objects resolution were causing system failures. I needed to rethink my strategy to produce the best possible outcome within the constraints.

I tried to texture the scene within zbrush itself, however, this produced limited results where the texture would simply wrap around the object and seams were visible. The result was awful and would be jarring on the eye. The only solution available was to simply reduce the amount of polys in the scene. This would mean that high detail would be lost and mountains would appear more rounded. However, sacrifices had to be made in order to produce a working scene which would work in maya. I reduced the scene to a sub division level of 5. This would produce a scene which still included some detail without using up huge amounts of memory.

My strategy had worked and I was now able to texture the scene and align the UV's with relative ease. If I was to create an environment again i would either use software like vue which uses displacement maps exclusively which reduces memory while maintaing high end results. I need to find out how to use displacement maps in maya as well as bump maps. These procedures would give good results especially when creating rocky environments. Despite the problems I have created a working environment that works for the project which allows for animation and importing to take place.

Here are a few images which show the terrain from a zbrush model to a lower res textured image. I have also uploaded the textures which I created using photoshop.












The desert terrain

I painted a couple of images concerning the environment in which the mine and its various components would be located. I focused on lighting and composition as well as the design of the mine itself. The concepts were originally done in black and white to get a realistic ambient light. When I was happy with the result I would then add a colour filter which  would help us when it came to creating lighting and matt paintings.

I only got round to painting a colour filter on one of the concepts, however this gave us some valuable lighting and colour reference.

I was overwhelmed at the quality of the images and how they suggested mood as well as a futuristic but industrial world. I wanted the atmosphere to be dark and saturated with beams of light only just passing through the dense cloud cover. I think I have achieved this within the images I painted. However, incorporating this look within a 3d world would be a challenging and difficult issue. We may have to evaluate the look of the environment and its atmospheric conditions once we model the scene.





Friday, 19 February 2010

Final Renders

Once I had finished the textures and decals I projected these onto the ship using planar mapping and transparent planes. I was thrilled when the desired effect had been achieved. I didn't think that the texture map would work. Hours of practicing paid off as I managed to fit the textures onto the model effectively. In some ares the texturing could be improved however, when the vehicle is lit this will help define the whole ship.

I tried to bump map the ship by using an existing texture. However, the result was undesirable and the ship appeared as though it had ripples all over the hull. I need to spend more time using the bump mapping tools to get the desired effect. But for now I am very happy with the result as it has fulfilled the groups expectations and my own. With the textures and decals now on the ship the whole model has now a realistic resonance about it. It not only feels realistic but also appears used and worn from the harsh conditions it works in.

Here are a few renders of the final build. 





Decals

I was pleased with the final outcome of the texture map I had created in Photoshop. I thought that the worn appearance of the metal befitted the story as well as the environment and its other components. The burn marks looked aesthetic and suggested the amount of energy needed to propel the ship through the air. The sheer scale of the object required engines that burned fuel constantly. The burn marks also suggest that fuel is not being burned efficiently but huge amounts of energy are being produced. The metal is rusty and worn with the constant battering of sand storms and the hostile atmospheric conditions. The group seemed to be very impressed with the texture, however more decals were needed to cover the ship. This would emphasize its commercial use. The Mining companies logo needed to be fitted onto the side of the ship. This would give the ship some identity and it would also suggest that the ship was owned by a very powerful organization.

I created some small decals which I would insert over the image using planes. I would then turn down their transparency. By doing this I could preserve the existing texture while at the same time ensure that the decal was the only image visible on the plane.