Working on the Head Model
Overall I found the head model very challenging and uncomfortable, although it showed me a lot of new techniques. Honestly the biggest issue I had with the model was trying to work on my own face. I have always found it difficult to model/draw organic models as well as having a dislike for drawing self-portraits. With that said I did learn a lot of new techniques and enjoyed the time I spent creating the face.
Looking at the model I find it hard to see myself. The parts I dislike the most are the chin, cheek bones and neck. I think that the chin is far to pointy to be mine, and the cheek bones need to come out more, giving the model a more rounded cheek. The neck is only a small issue but I found that it came out to wide and had a few unknown bumps in it.
The parts of it I like are the nose, eyes and hair. The nose is still a little too straight but it looks to be the same size and have the same bridge. The eyes are the same size and I feel they are the right distance from the nose. The eyes still need work done on them and detail added but the basic figure is there. The hair also seems to follow the right mold, showing similarities to the real hair style and size. The only issue I have with the hair is the white light highlights it seems to show.
To conclude, I feel that I have created a low quality head model, with some likenesses to my own head. If I where to repeat the model I would defiantly spend more time on the models shape before textures where applied, as this caused a big issue when trying to adjust it later. One last thing I would do is add a second and third hair and fur modifier to the model and create realistic looking eyebrows instead of the textures ones that can be seen in the final render.
Working on the Guildhall
I found this part of the module the most interesting, fun and educational. I found that during the time I spent developing the models and the animation I learnt several good techniques as well as tips including;
The slice tool
The pro Boolean tool
How to use Rayfire
How to use lighting effectively and the specifications
How to use textures more effectively and give them properties
How to animate with multiple tools and camera paths
How to use the network render (Backburner) for Autodesk
Learning all of these skills and more has greatly increased the speed at which I can work as well as the quality of the final product. The only two things I would have to mention negatively with this project was the time restriction and the lack of equipment. When we started the project we were given six weeks to complete all of the models as well as the animation. This was difficult to keep to and we ended up taking another week to finish off the animation. I felt that to complete the high quality models we should have been given six weeks alone, with another six weeks for animation, rendering and video production. The other problem we had was the lack of equipment, broken computers, nowhere to render most of the time and no computers that could work with Rayfire. This caused constant problems throughout the project, but is respectably unavoidable.
The rest of the group seemed to work well with odd moments of weakness, but as a regular group member I can understand that stress levels and issues can cause arguments at any time and the best thing to do is relax and come back to the discussion later. I feel that we all worked hard and well, notably Jack Cassie who did a good job leading the team and working hard to learn the plug-in Rayfire with me and create the animations. The group and I worked hard to complete all of the models, animate the finalise the final product, and I could ask for no better.
The things I would do next time would firstly be to try and use a better range of computers, that give all of us access to Rayfire as apposed to just me and Jack. I would also like to have worked more on the lighting, as this was a big issue. We originally had two renders of the product, one with global illumination and one using the daylight system. The problem we had was that the global illumination copy looked unreal, but the daylight copy had to many beams of light, giving a realistic lighting effect but smudging the walls and creating white light in points where there shouldn't have been. Me and Jack both spent around twelve hours over the period of two days trying to create a better lighting environment with no success. This would have made the HD renders look more impressive and realistic.
In all I felt that the group worked extremely hard and I personally learnt a lot, as well as gaining the motivation to learn even more. In my opinion the project still looks amazing and one of the highlights of my educational lifetime. It could in respect use some work in places and alterations to the lighting, however given enough time anything can be improved and with that in mind, it's finished
Module Evaluation
Wednesday 21 December 2011
Tuesday 20 December 2011
Finished Head
Below is a render and video of the finished head, an evaluation will follow.
 |
| Final Render |
Using Morpher To Animate
To demonstrate my skills I wanted to use the morpher modifier to animate my model in some form or another. To do this I first had to turn off the hair and fur modifier as this always wanted to take the priority slot and I could not find a way to animate it with the hair still on the model.
After the minor set back I then went on to create a very simplistic animation whereby the eyes move from side to side. This was done purely to show I have the ability to use morpher and animate models and to save time by not doing a more complex movement. To do this I made two copies of the original and then rotated the eyes. The important thing when using morpher is to not add or remove any vertex or polygons, as the morpher will only work as long as the only change that can be made is the positions of the vertex/polygons.
The final animation I created can be seen below.
 |
| The head and morphers |
After the minor set back I then went on to create a very simplistic animation whereby the eyes move from side to side. This was done purely to show I have the ability to use morpher and animate models and to save time by not doing a more complex movement. To do this I made two copies of the original and then rotated the eyes. The important thing when using morpher is to not add or remove any vertex or polygons, as the morpher will only work as long as the only change that can be made is the positions of the vertex/polygons.
 |
| Morpher looking right |
 |
| Morpher looking left |
Finishing the Head
With the project time coming to an end I had to decide to show i had the ability to use other tools in the program instead of creating detail on the model. For this reason I finished the model's texture and the ears and began to add hair using the "Hair and Fur" modifier. The head itself looked unlike myself and had several issues, most notably being that the ear textures did not match the head, as well as the texture containing hair. If I where to re-create the texture I would make the entire head skin colour, and then only using the modifier to add hair.
The hair tool is surprisingly difficult to use, as I had many problems cutting the wrong parts of the hair off, and moving it in the wrong direction. After working on it for some time however I managed to create hair that could resemble mine, in respect to the length and basic shape. The one down side I had was the hair seemed to give off a white light. This may have been a highlights problem with the render, but without more time it is difficult to say for sure.
 |
| Finished texture map |
 |
| Ear texture attachment |
The hair tool is surprisingly difficult to use, as I had many problems cutting the wrong parts of the hair off, and moving it in the wrong direction. After working on it for some time however I managed to create hair that could resemble mine, in respect to the length and basic shape. The one down side I had was the hair seemed to give off a white light. This may have been a highlights problem with the render, but without more time it is difficult to say for sure.
 | ||
| Using the hair modifier |
 |
| Changing the hair colour |
 |
| Render of Hair |
The Final Steps
The last few days me and the group have been spending time rendering both the scenes and then the frames again inside Adobe Pro (The video editing software were using for this presentation). As we did this I continued about re-rendering some of the scenes again with different lighting specifications to improve the quality.
As I did both of these things I also used my time to create the credits scene for the final presentation, as well as showing the others in my group how to put text overlays onto the moving images. These make our presentation look more professional and gives it a modern feel.
Monday 19 December 2011
Network Render
To render the scenes effectively we decided to look into network rendering. This was so that we could have high quality renders, but not have render times that go into double digits. The main idea behind this is simply having all of the textures and models needed on one master computer and then having others linked to it via server (using there IP addresses).
This allowed us to render full scenes in 30 minutes that would have taken 3-4 hours on a single computer. The way the master computer works is by allocating each slave computer an image to render, and then assigning it a new one after it finishes. With this process being repeated on 30 - 40 computers, its understandable how effectively this can work. The bonus to this is that if one of the renders crashes, the image it was given gets re-allocated and sent to the next slave.
This allowed us to render full scenes in 30 minutes that would have taken 3-4 hours on a single computer. The way the master computer works is by allocating each slave computer an image to render, and then assigning it a new one after it finishes. With this process being repeated on 30 - 40 computers, its understandable how effectively this can work. The bonus to this is that if one of the renders crashes, the image it was given gets re-allocated and sent to the next slave.
Animating
After all of the models and textures where finished we moved onto animating the scenes. The main task I had in animating was creating the camera angles for the HD shots, including there path constraints. I also learnt how to effectively light the scenes with the daylight tool from Jack Cassie, and in return started him off with the basic commands in Rayfire. (The Rayfire animations where given to Jack to complete as unfortunately the program requires high-spec computers, of which only he owns).
The way I completed the camera movements was mostly done using splines and path constraints. To do this you simply draw out the path you with the camera to follow, and then apply the camera to said path. The camera will then go from start to end of the path, allowing the user to use AutoFrame to focus of the camera shot when necessary.
The lighting was created using the daylight system, where by you use a beacon to indicate the sun, and then "portals" that you can places inside of windows and doors to allow light through. This helps create realistic shadows and lighting. The only issue with this is that as a group we found the images would look grainy, and there would be clear points where the light beams could be seen, ending on walls. This made some points in the images overly bright with dots, and then making darker areas such as walls look very dirty and blotchy. For example in the image below there are a few spots, notably the computer and celling where the light spots can be seen. The blotches can be seen on the wall surrounding the window and again further down the wall.
The scenes I animated where:
Library
Shop
Museum
Guild Room
The way I completed the camera movements was mostly done using splines and path constraints. To do this you simply draw out the path you with the camera to follow, and then apply the camera to said path. The camera will then go from start to end of the path, allowing the user to use AutoFrame to focus of the camera shot when necessary.
 |
| Path constraints and camera |
 |
| Daylight system used with "portals" (Placed lower then usual) |
The lighting was created using the daylight system, where by you use a beacon to indicate the sun, and then "portals" that you can places inside of windows and doors to allow light through. This helps create realistic shadows and lighting. The only issue with this is that as a group we found the images would look grainy, and there would be clear points where the light beams could be seen, ending on walls. This made some points in the images overly bright with dots, and then making darker areas such as walls look very dirty and blotchy. For example in the image below there are a few spots, notably the computer and celling where the light spots can be seen. The blotches can be seen on the wall surrounding the window and again further down the wall.
 |
| Library Render |
The scenes I animated where:
Library
Shop
Museum
Guild Room
Subscribe to:
Posts (Atom)