Assignment Report

The fairy tale that has been decided was Goldilocks and the Three Bears, the reason for this was because it would allow for a huge amount of creativity to be used when designing the story. This fairy tale is in first person to give the illusion that the player is actually the character within the story. This was done with the potential future use of implementing it for the Oculus Rift.

This fairy tale will be made using multiple levels instead of an open world, the reason behind this is because through the menu the player can start at certain points within the story instead of playing the whole story.

Many decisions had to be made within the design phase, this included the layout for the chapters, and how the player would move within the environment. The decision was made that the player would have control to move around the environment but features like the eating of the porridge, sitting on the chairs and lying on the beds would be done through Matinee animations.

Creation:

Once a layout had been established it was then time to create the landscape for the chapters in the unreal engine. The texture that was used was a custom material that had two layers one grass and one mud, which would be used as a path. This material was created following an Unreal document, below is a screenshot showing the layout for the material.

Once this had been created the paint tool was used to add the mud layer as a path for the player to walk on, the next step was to create trees for the forest to fill up the levels the trees were imported from SpeedTree and then duplicated multiple times. The original idea was to use the paint tool inside unreal and use this tool to create the forest.

This method however was very processor heavy and made the gameplay lag badly. The trees were implemented with collisions to stop the player from walking of the path.

The next step was to create a basic structure for the outside of the house. This structure was created using the BSP blocks inside Unreal, and using subtractive blocks to create doorways and window frames, a wooden fence was also implemented to create a garden around the house, and this model was download from TurboSquid below an image showing the outside of the house.

Wooden fence – http://www.turbosquid.com/3d-models/free-pack-wooden-fences-modular-3d-model/600091

Below is the Matinee animation and Blueprint that was applied to the front door to be activated when the player collided with a trigger box.

Assets

Once the foundations had been created the detailed assets could be added, the majority of the assets inside the project were created using Autodesk Maya.

I decided that the next step was to create the difference chapters using the same tools that were used for the house. I manipulated the size of the boxes until the rooms were a realistic size.

A fire particle effect was implemented in the ‘Chapter’ level to give the room a more comfortable feel, a fire sounds effect was also added here to make the particle realistic.

Assets created include:

• Cupboards
• Table
• Porridge
• Spoon
• Lamps
• Beds
• Chairs
• Bowl

The models that weren’t created were downloaded from turboSquid.com or tf3dm.com

The only Blueprints added to this project were blueprints that triggered the matinee, sounds, the opening of doors and moving to between levels.

Matinee

Matinee animation was used to keep certain sections of the story to be played without interruption.

There is a matinee for the doors which plays when activated by a trigger. The bowls of porridge animation zooms in on the porridge bowls, the porridge goes down slightly on the first two bowls and disappears on the last bowl to give the impression that the player is eating the porridge.

A matinee was created for the chairs which shows the player sitting on them, while using a sound track to play the narration for the story. The last chair had key frames added to break the legs of the chair and topple it over.

The animation for the beds was done in a similar manner giving the impression that the user was lying down on the beds.  The voice-over for each section was done in a recording studio which then required using Audition to split the track into small sections to sync with the matinee.

A model of a bear from this website http://tf3dm.com/3d-model/brown-bear-17385.html was used, then it was imported into Maya and rigged/ animated and duplicated. There were some difficulties however with importing the animation into unreal due to some unforeseen problems. The video below shows the bears moving in time with the audio.

In total three animations were used in the story.

All the bears parts of the story was done in one level using one animation with many cameras made into a seamless matinee. Once the story was completed the user is taken back to the menu were they can exit the application by pressing ‘E’.

Sound

Multiple triggers were placed throughout the application that played certain sounds at certain points in the story. 3D attenuation was used to make sounds gets louder the nearer to the source for example the water by the bridge. Ambient music was also added to make the experience more realistic, (wind, birds)

All the sounds used were either third party sounds or sounds created using Audacity and Audition, unfortunately there were problems using FMOD as the files could not be exported

If this was to be done again an open world would be used instead as many problems occurred using multiple levels, Level of Detail would have been implemented on the trees due there being lag, to make it run smoother , and more sounds would have be implemented.

The scene was successfully packaged.

References

All third party sounds from http://www.freesfx.co.uk/

Footstep, Walk, Walking – http://www.freesfx.co.uk/

The rest of the audio came from https://www.freesound.org/browse/tags/sound-effects/

OR
WAS MADE IN AUDACITY AND AUDITION

Assignment Brief

This assignment requires the creation of a virtual 3D environment for a fairy tale story, there was a choice between:

• Little Red Riding Hood,
• Hansel and Gretel,
• Goldilocks and the Three Bears,
• Rumpelstiltskin,
• Three Billy Goat’s Gruff
• Puss in Boots

The fairy tale that was chosen was Goldilocks and the Three Bears. I was also required to produce an online journal/blog that should outline the main steps, technical, and creative decisions that were made in producing the story, as well as an evaluation on the effectiveness of these techniques and the artefact overall.
The word Count for this Assignment was between 600 – 1,000

This is were Post- Production & Special Effects Assignment Begins

Assignment Part 2 – Dynamics Scene

Here is my Completed and Rendered Environment displaying the following Dynamics:

• Fire,Smoke and Sparks
• nCloth and Destructive Environment
• Fluids (Water)
• Wind (tornado)
• Particle Emitter

I used the YouTube video shown in the link below for the tutorial on how to do the water that was in the pond:

https://www.youtube.com/watch?v=HIzxDmw88qY

I created a basic scene in Maya, which included a volcano and a couple of cliffs and a piece of land, I then proceeded to add the Dynamics I first started with the waterfall (see screenshot below on right) as I knew this would be the easiest, I created this in the same way that I created the water dynamics as shown in a earlier post (Dynamics).

I then created the nCloth shown in the first screenshot below, I created this my using the face cut tool on a flat polygon plane to make random cuts in the polygon, I then highlighted all of the edges for this plane and going into the nDynamics Menu Set and the nMesh tab, I selected the create nCloth tool, this made the plane a nCloth, however when I played the animation due to the gravity in Maya it would drop straight down straight away, to fix this I went into the Attribute Menu for the nCloth and adjusted the direction of the gravity after changing other settings when I now played the animation it no longer dropped straight away. Next I created a simple rock and by going to the same Menus as when I created the nCloth however this time selecting the tool called create Passive Collider so now after key framing the rock to pass though the plane, when I play the animation the result shown below is achieved.

After Creating the waterfall and nCloth I imported from my tutorials the Sparks from the Fireworks and the Fire from the Match Scenes. I then simply adjusted settings for each dynamic until I had something I was happy with. As shown in the Screenshot above on the left the fire is to show the eruption of the volcano while the sparks are smaller eruptions, the smoke shown behind the fire was simply created by creating a new 3D container and changing settings until the desired result was achieved.

Below is the screenshot showing a close up of the tornado, as in the video it is not clearly displayed, I not 100% happy with the outcome of this, as I don’t think it looks exactly like a tornado and if I could do this again I would try to make a better tornado, the twister shown in the earlier post called fuel,fire,twister is a better creation of a tornado as it looks more realistic.

The Final Dynamic I created was the actual pond this was by far one of the more difficult dynamics to do the link to the YouTube video shown above helped me create the final outcome and I am happy to say that I think it turned out very well and looks like water. I then created a Skybox to make all of the dynamics stand out more. Here is the hypershade window for my environment. As shown in the screenshot I have many different textures that have been used there is the texture for the volcano, water and land as well as the lava rock, the textures for the fireworks is also shown.

I am happy with the way this scene has turned out, however if I was able to do this all again I would take more time in creating all the dynamics so they would be at a higher quality and standard to what they currently are, I would also take more time in creating a more complex scene to better show of the Dynamics. I made the animation using the perspective camera however i created a new camera to edit the dynamics were needed.

Assignment Part 1 – Good/Evil Model

The two screenshots shown above show my Character in Mudbox, were it demonstrates both the multiple sculpt layers I used, as well as the reference image I used to get a basic idea of how I wanted my Models to look like, I added the reference image to the perspective camera meaning no matter were I turned the camera I could always get a clear view of the image. The first step I took was to do a base mesh for each section on separate layers, on the lowest sub division then move up the sub divisions adding detail as I go.

The Image to the left shows my model fully rigged along with all the names for which I called each joint, this allows me to see easily where each joint is located and fully animate my model. My model has both forward kinematic and inverse kinematics this is very similar to the last rig I did in a previous post.

Below shows a screenshot of me using the paint weight tool to make sure that when I animate my model the mesh does not deform due to the naming system I used it was easy to locate a specific joint without too much trouble. The other screenshot below shows all my paint layers for my characters.

These two images that follow show my finished good and evil models which was created using different sculpt layers and tools as well as different paint layers, each layer for a different section of the models one for good and one for evil, while the other image shows my level of detail group. This means that when the animation scrolls through the models will change the closer to the model based upon the settings shown in the screenshot to the correct model, the models all have different number polygons which was achieved by using the reduce tool. (see video below of final Level Of Detail) The distance for the Level Of Detail was achieved by using Key Frames

The two images below show different stages of the blend morph, for both the model and weapon. I used the blend morph that was previously shown in a other post and I used a new technique that allows for the colour to blend as well as the actual mesh.

I used the Blend Shape tool to manipulate the different models so they would appear at different times here is an example of the blend shape tool that i used in my animation.

Below are the complete videos for each section the heading before each video tell what the videos are about:

Blend Morph

Level Of Detail

Wireframe Mode

Shading Mode

I have used a variety of new techniques for example the blend colour, to morph the texture of one model and another, I have also used a bump map on the lower polygons to give the the appearance of the higher meshes without having all the polygons this save the render time. In mudbox I have used different paint and sculpt layers, to create my models, I have also used a reference image which I added to my perspective plane to help me sculpt my model into a similar appearance. I enjoyed this assignment very much and would love to to something like this in the future, if I had more time however I would look into applying blending normal maps and perhaps using more maps not just specular, bump, incandescence and diffuse .

fuel,fire,twister

Below are image showing the completion of the fuel and fire exercises:

Below is my Completed Twister using a 3D Container:

I found the first two exercises to be fairly easy and didn’t take too long to finish I was able to easily change attribute settings to get the desired effect, the twister task however took a while as I had trouble getting the twister to keep it’s shape, as when I adjusted settings like swirl to get a better tornado, I had to adjust other settings as well to keep it all in balance. I am still not 100% happy with the final outcome for the tornado however the one shown in the video above was the best I was able to achieve, without it completely losing it’s shape altogether.

Over all I am happy with the end results for this exercise however as I said if I could do this again I would try and make each tasks more realistic by changing a few more settings.

Motion Path and Emitters

Below is a video showing a model car that when it passes through a 3d container smoke will emit and follow the trial of the car, and smash through a simple wall at the end:

I created this by simply importing a model car from Mudbox in Maya then I created a motion path with sharp curves and attached the car to the motion path, once i had done this I set a key frame at the start of my timeline when the car was at the beginning of the path and then at the end of the timeline I set another key frame where the car was moved to the end of the path. Now that the car moves along the motion path I created a 3D Container that I positioned along my motion path I then created two emitters inside this 3D container and I parented these to the back two wheels of the car, the reason for this is so when the car moves inside the container smoke will start to emit from the emitters, after changing some settings until I got the effect I as most happy with for the smoke I created another container, I then created two more emitters and parented them to the car and repeated the same settings as I had from the last container. There is an easier why to get the emitters to from the last container to also work in the new container, however I was happy doing it this way.

After the containers had been created and smoke emits from the emitters I created a simple plane and cut the faces and then detached the edges and then removed the gravity so when the animation plays it would not fall down, then after adding a ncloth to the plane and a passive collider to the car when the car hits the plane it will break apart.

I am very pleased with this task if I was to do this again though I would implement the other method of allowing the emitters to work in more than 1 container instead of having many different emitters for certain containers.

Dynamics

Below are completed screenshots of the three dynamic scenes that I have created.

I created these three scenes using the particles tool in Maya and manipulating the settings.

For the model of the boy peeing on the wall I was given the model of the boy and I created the spray by creating a particle emitter using the tool create emitter located in Dynamic>Particles>create emitter. I changed settings such as the lifespan of the particles generated as well the type of render the particle will be e.g. sphere or cloud, for the spray I used streak.

The scene involving the fireworks was the most complex task, this was because I had to change many of the settings and add ramps to many components for the colours e.g. glow, transparency. I also had to add emitters onto the particles themselves to create the effect of explosions, this was also the first time that I have had to make sure I name every component to make managing them easier.

the final screenshot shows particles band how they are effect when gravity and a rough landscape are combined together, while gravity is used in all of the scenes to make them fall down and give the effect of them being effect by gravity, a plane is also required to stop them from constantly falling with is achieved by making using the tool ‘make collide’. In regard to the final screenshot these two features (gravity and the plane) were the first task that I completed to get me used to the tools, I created the particles using the particle tool and after adding a gravity and creating a fall off for the plane to add bumps I could see how the particles were effect by the type of plane. I changed the settings just like I mentioned in the previous tasks to give the effect of water.

I enjoyed these exercises very much I will definitely be using these tools again for future tasks.

Rigging

After having a mess around with the character (Zombie Bellman) from the movie hotel Transylvania, to get used to using the rigging, I created a human rig using the mesh for the human rig was the human model already created in Mudbox. Once the mesh had been imported into Maya I used the joint tool to create one of legs I positioned the root of the first joint at the hip, the second on the knee and finally the third on the ankle.

I used 2D views to add the joints and switched between front and side view to ensure the joints were in the correct place, after I created the first leg I made sure that it moved correctly and then using mirror joint I created on the correct axis another leg. I then repeated these process for the arms and using locators and the joint tool I created the joints which would control the back, I parented all these joints a single joint which I called Hip01.

Next I used the bind skin tool to bind the mesh to the rig so when I moved a root on the rig the mesh would also be effected, when I tried this I noticed that areas of the mesh that I did not want to be effected were also being manipulated. To fix this I used the paint weight tool to reduce the area a root would effect, at first I did not have a clue how to use the paint weight tool however after watching some youtube videos I felt fairly stupid for not understanding as it is really easy to do. Now that I have used the paint weight tool when I move a joint e.g. arm on the correct vertices are being manipulated on the mesh. Below is a image showing the complete human rig. I used both Forward kinematics and Inverse kinematics on the rig I can move individual parts to get the desired effect through forward kinematics and can move the whole leg easily through inverse kinematics.

Here is my rig with the paint weight tool active to show which areas are effected by certain joints and the image below shows a list of all the available joints the black is were nothing is effected and the white is what is effected this image is for the left shoulder joint:

I really enjoyed doing this exercise and I will defiantly be using both the rig and paint weight tools again for future tasks/ assignments.

Level Of Detail (LOD) & Reduce Polygon Count

Reduce Polygon Count I was tasked with reducing the polygon count for a model to help increase the render time, I used the tool reduce which located in Mesh>Reduce and select the option here as shown in the screenshot are different settings to help reduce the amount of polygons on a model, I had to move the slider to 100% and even then I had to repeat this process multiple times to get a amount of polygons that was greatly reduced from the original model, I also kept the original as shown in the screenshot it is the top field.

I reduced the model until I started to lose detail, the screenshot below shows the reduced model on the left and the original on the right, it is easy to see which is the reduce model and which is the original model.

I am very happy with the way this task and will definitely use this technique in the future for other tasks, if I was to do this task again I would take more time in reducing the polygon count.