MY FINAL YEAR DEGREE FILM

MY OWN GROUP"S OVERALL ANNUAL EVALUATION

2009-06-23T02:15:00.000-07:00

As a scuba diving guide I often get to guide a group of people that I have never seen before on an unknown terrain. This is a hugh responsibility. There have often been strong currents against us or we had other issues connected with mistrust, lackage of self-confidence or experience. There were several times that I simply wanted to stop and surface, but so far ... I have always pulled myself together and brought them all safe and sound to the end.

I have feeling that guiding my final year degree film team throughout the year was prety much the same thing. I was given a chance to work with people that I have never worked with before. There have been many troubles (from personal to technical) and there has been a lot against us, but we didn't give up.

Despite all the pressures we managed to finally pull ourselves together and create a beautiful film. It is true it is not finished yet, but it is done.

Thank you for that all! And I wish you a very successful future.

MY OWN 3RD TERM SCEDULE

2009-06-20T07:50:00.002-07:00

WEBSITE

2009-06-20T07:50:00.001-07:00

I am still in a process of building my own website, which will include the whole section solely devoted to our final film. I will of course credit all the people who contributed.

the website is called www.anjatolar.blogspot.com

and there you can see also other projects and associated work that I worked on through the years at Ravensbourne.

Other project promotional material:

Logo and business crads

NOT SO FINAL OUTPUT

2009-06-20T07:49:00.001-07:00

rough film

This is still not a final version of a film as I did most of the work myself and I am still improving it. There is lots left to be fixed. In general I simplified it, so it is more "understandable" for the viewer, but gaining descriptiveness it also lost its very abstract look from before, which, to admit, I slightly regret.

PRINTS FOR HARVEY to give a feel and COMPOSITING

2009-06-20T07:47:00.000-07:00

PRINTS FOR HARVEY AND COMPOSITING

Althoug one meter long prints is the last thing on earth that a busy worker would like to do when trying to finish the project...it kind of turned out great experience as it is for the first time for me too see how 2D animation fits on the top of 3D backgrounds (later on when compositing, 2D was actually planted "in-between' 3D rendered layers wit a help of camera tracking).

TITLES AND CREDIT'S DESIGN AND EFFECTS

2009-06-20T07:43:00.000-07:00

Titles and credits

I have several ideas for these. First I wanted to develop my own type, but then I found some nice ones on web and also compositing different elements I created the "peace and harmony" seal stamp, which is the part of film logo.

film promotional material. I did a few versions of title and credits, going from more simple to complicated and back again!

CREDITS NOW:

CREDITS BEFORE ( a bit confusing for the eye):

COMPOSITING AND CAMERA TRACKING

2009-06-20T07:42:00.000-07:00

Experimenting with animation techniques

The initial idea for this film was, to achieve the synesthesy of different animation techniques; 2D, 3D and live action. The reason for that was to achieve viewers involvement with the story, but also feel the tactile quality of the with material (paper). This lead from photo collaged backgrunds to 3D environments with 2D animation composited on the top. Live action was firstly expressed through photographed frames of hand movement. I had an idea of 3d crane being animated with live action hands, but after speaking to some experts and I excluded that idea as it would simply take all my time to be solved and I couldn't afford that. This is why I decided to completely abandon the idea of having a live action involved and simply concentrated on marriage between 3d and 2d.

camera tracking

There is no way to properly track the motion of 2D animation to fit 3D camera move. It was always jumping around, no matter how hard we tried. So I needed to do the camera tracking. Camera tracking is done with the help of a poly plane placed on the spot where would otherwise be 2D animation. This plane is assigned a texture with high contrast pattern that is easily definable from near or far for the tracker to lock onto. Tracking can be done nativly within After Effects, using its point tracking system, which gives varied results. It would be better to use Mocha, which is AF plugin and offers planar mapping system and gives into account the whole image data (planar tracking instead of point tracking ).

I used tracking because the other option where I assigned the 2D animation onto a plane in Maya as an animated (.mov) texture. The problem was that Maya didn't take opacity into account no matter what movie file I used (regardless I turned on the transparency option).

compositing

All compositing was done in After Effects, importing 3D backgrounds in layers, enabling me to have bigger control over them. I discovered that I need to put on "premultiplied" option to keep the light effects the same as when rendered out,

2D animation was imported in editable layered PSD files, previously sorted out with animation option in Photoshop.

2D Frames of crane flight from the top, composited on the top of rendered layers from Maya:

Final look:

LIGHT LINKING, A PAINFL JOB

2009-06-20T07:40:00.000-07:00

LIGHTING

Similarly also supports the passage between two worlds. It is more theatrical and sharp creating more inevitable feel of hospital scenes and on the other side light brightness creating the atmospheric feel of dreamlike landscape. For the second one I used a lot of glowing effects and played around with the depth field options of the camera lense.

To achieve versatile light effects I decided to do light linking (Window- Relationship editors- Light Liking- Light- centric.) in every scene, which was absolute pain as once we assign one object to the certain light, we also need to exclude all the rest of the objects. This needs to be done right after modeling as if we bring another object into the scene at that point- that means that all the lighting needs to be re-linked. Linking also enabled me to assign fog to just certain lights (objects ) and the same with reflections in the water.

A lot of objects (lights in the hospital scenes for instance) are actually enlightened within themselves with the help of texture attributes like glow, incandescence and specular. shading, so they aren't really enlightened by separate light, which saved us rendering time. Light linking also doesn't provide any shadows, which is why everything needed to be rendered out on layers in order to drop shadows in After Effects.I also noticed that in case I use matted option when importing layers into AF, the lights get completely different as if I use " matted- premultiplied" option.

A BOMB EXPLOSION AND OTHER VFX

2009-06-20T07:38:00.001-07:00

vFX (Bomb explosion and particle animation)

After Charlie tried many times to do it in 3D, I decided to simply do it in After Effects and it took about 10 minutes to do.

The bomb explosion made out of hands I made in After Effects, using watercolor images of hands in varied poses (I used rotoscoping to draw them). I used this as a reference texture for particle effect in AF. I used vector mask in a shape of mushroom cloud to mask out any stray particles to maintain the atomic bomb explosion look. The animation of glow was applied on top of this effect.

This same effect was used in certain parts as a transition between images, to give burnt out effect.

I also animated ripple effect on the water when character jumps over the puddle.

I also used "fast blur" effect to soften the edge of some imaginary scenes.

Ripple:

PROBLEMS WITH MAKING CRANES DOLLY SCENE

2009-06-20T07:37:00.000-07:00

The worst scene for me to render was (still is) the hospital scene where camera dollys through a waste amount of appearing cranes. The reason for that is that Simon Butler instead of animating the cranes with one same reference which he could simply intsance to get more copies...actually duplicated the crane within reference folder several times and after that also within scene. This means that every new object has its own texture, which (beside the enormous polgon count) made scene so wast that it is constantly collapsing, even if rendered with render farm. I was trying to render it for more than a week. e also didn't stay clear with naming the objcts, which made the thing even worse for me to arrange renders and adjust glow on the objects!

Also I have to mention that all the scenes I have received from no matter whom weren't clearly named and arranged (despite my request for it in annual plan and meetings) so it was a great nightmare for myself to do anything further in them.

ANIMATING THE CAMERA MOVES

2009-06-20T07:36:00.000-07:00

Camera animation

It took me a while to fully understand the difference between simply zooming into the scene and actually panning into it. It looks so much better (for instance: the light effects change according to camera move) to pan than zoom that I decided to actually animate camera in 3d and than find a solution to composite the 2D animation on the top.

I also learnt how to make room optically bigger than it really is with altering the camera lens options (animating the depth of the field).

I actually tried also to import 3D object (.obj file) from Maya into AF and animate the camera move there, but it slowed down the rendering time and work itself so much, that I rather decided to drop the idea and rather wait for better times of Adobe package.

VISIT TO SCIENCE MUSEUM- BIRD'S FLIGHT

2009-06-20T07:35:00.000-07:00

Visiting a science museum I noticed a great stimulation of Bird's flight which could help us to animate the crane.

There I also saw for the first time one of the objects from Hiroshima, which melted in atomic bomb attack in 1945.

2D ANIMATION-INKING, COLORING, ROTOSCOPING, USING PHOTOSHOP TO ANIMATE, COMPOSITING IN EDITABLELAYERS

2009-06-20T07:34:00.000-07:00

2D ANIMATION DRAWING INKING WATER COLOring

The reason I decided to do animation with ink and watercolor is not only because they are my favorite techniques, but also they express the traditional quality of japanese art. Water color infuse a subtle and very versatile quality to the general style. They are very unpredictable and each frame differs from the previous one which creates the effect of glitching movement and so more lifelike feel.

The whole peace is based on the contrast between red and blue, the traditional geisha colors of the characteristic for the time. Red means life power, blue divinity, sublime.

rotoscoping

To draw animated body sequences in some odd angles was too much even for my artistic imagination. So after photographing and filming the references, I printed some of them out and redrew them frame by frame, fitting the style to the one of the main character. This was a really hard job as character is very deformed. Some of other references I used to study the change in facial expressions and shading.

After I Drew the sequence with the pencil, I scanned it in and line tested it on college computers- until I found a better solution- animating it in photoshop. After nearly losing my mind, tracing all the animation I did with the pencil on an animation paper onto the watercolor paper with ink- I found a better solution. I scanned in all the frames, altered them in photoshop to give them more inky look , cleaned them, aligned them and printed them out on the watercolor paper. Once painted I scanned them back in and send them to the group members to cut them out and align them again.

ANIMATING IN PHOTOSHOP- CUTTING IT OUT, ALIGNING, COMPOSITING

First of all I learnt how to simulate light box when one doesn't want to spend more than £10 on it! :-)

As said before I helped myself with rotoscoping reference when animating some of very complicated perspective angles. After painting the animation by hand (the process explained in Rotoscoping folder), I scanned it in. I cleaned the white background out with the help of MASKS in Photoshop. These enabled me to simply paint back in case I wiped too much out, while using magic wand or rubber, wouldn't give an option to return once done.Once cut out- I aligned the objects together and opened the animation timeline in Photoshop. This feature was available already in CS3, but it became even more appropriate to use in CS4. It allowes quick line test and once the position and duration of frames is sorted out, can be easily exported into any postproduction software in a shape of layers.

3D ANIMATION- FIXING OTHER PEOPLE's WORK

2009-06-20T07:33:00.000-07:00

FIXNG 3D AniMATION

I haven't been really satisfied with any animation that other members passed back to me. Simon's fish animation (by the way he duplicated fish instead instance it again! And as soon he wanted to change the curve on which the fish moves - in order to get different timing- the animation cracked- the problem was in joint orientation of the fish) and Dan's Clock trees needed minor refinement( they bent a bit too high, so the animation looked a bit odd for feathers). I changed both animations working within graph editor (changing the lengths and heights between frames, reshaping the curves, so it floated more smoothly.)

FILMING REFERENCE

2009-06-20T07:31:00.000-07:00

SECOND REFERENCE PHOTOGRAPHS AND FILMING

because film was more defined on that stage we decided to make another photo and film reference as to help clarifying animation of the body in some rather odd camera angles (top down, from the underneath, run from behind) in case we need it for rotoscoping. This time we got a more willing actress and dan played all the rest of the family members :-)! For crane flight I found some reference of flight sequence in Science Museum. I actually took pictures of each sequence of the film.

The beginnings of animation

2009-03-19T21:59:00.000-07:00

I tried out some movements with rough pencil animation:

also I prepared the fish to be animated with help of REFERENCE OBJECTS:

I would advise that to everyone who has a massive scene, but in this case the file in which we work always has to be accompanied by the file we are referencing from.

RENDERING

2009-03-19T21:53:00.000-07:00

rendering

Firstly I tried out different rendering options in Maya, to see what happens to the lights, sharpeness of the edges, glows and reflections. As Can be seen on the picture references, Mental Ray behaved a bit strange with the see through objects, although it gave some interesting scattered look.

Maya Hardware left edges pixelated.

Maya Software Rendering seemed to behave the best of them as it left surfaces smooth, object's edges crisp and kept the transparency. I gave other members of the group to try out other options of rendering which could help us to achieve profuse look of paper (like subsurface scattering and ambient occlusion for instance), but it all turned out bad so I decided to simply go for Maya Software.

Mental ray

Maya software seems to leave the surfaces soft and crisp definition still:

Maya hardware made surface pixialted:

We made the decision to render out all in separately editable layers in order to have more control over them in compositing phase. For that I needed to sort out all the proper naming of the objects, assign them to the layers and then create rendering layers.

TEXTURING

2009-03-19T20:43:00.001-07:00

TEXTURING

At first we decided to texture all the objects with one same white paper texture, so the objects would be subtly defined solely with different light qualities. After trying so, I found out that this wouldnt't work and would also make a discrepancy with the colorful character animation composited on the top. So I decided to include a bit of ink drawing on all the objects (to shade them), which was an enormous amount of work, which didn't really pay off. It simply didn't look good, it was blunt and dull and so I finally decided to texture them with different materials. Different qualities of texturing represent the dihotomy of two different worlds: the world of her imagination and the real world of hospital. First one is expressed through softness and lightness of paper, the second through roughness of cardboard and bump.

Color range also contrasts from one world to another, while being more bright and divinely innocent in the first, more dark and earthly in the second.

UV MAPPING and UV SNAPSHOTS:

After modeling I needed to sort out the UVs of every object in a scenes that I made (There was just Louis who actually sorted out UV's for the scene I assigned him, and Moe for the fish I gave him to be rigged, all the rest were done by me). They are mainly all polygons, except few NURBS and subdivs. Texturing NURBS is much easier as the texture automatically wraps around the object. If UV's needed on a NURB object it is the best to convert (modify- convert- from NURBS to subdivisions).

Texturing POLYGONS:

Texturing NURBS:

NURBS are very smooth surfaces where the texture nicely wraps around, so they don't need some special treatment when UV mapped.But this can be also a problem when we want some more defined texturing of different parts of the same object. To do so e need to convert them to Subdivisions, which have the same texturing options as Polygon and at the same time kept NURB's characteristics (the way light hits them is very smooth).

Texturing SUBDIVISIONS:

As one might notice the fish is already rigged (by Mohammed Warsame), which usually wouldn't allow us any incisions into the mesh, especially not uv mapping, but if this is done already, then it does not matter when the texture is attached.

It was a really hard job to find the best way of texturing these 3D models without loosing subtle grading quality of ramps and just the right amount of glow that I assgned to the objects right after they were being made. They looked so nicely, that I really wanted to keep them, so I had to find the solution which would include these plus assgning the brand new paper textures to the objects.

This is the bomb city that is just simply ramped:

I tried out how it looks if I attach one single paper texture as we wanted to do in the first place.

Final (after finding the right solution):

But the outcome wasn't satisfiable, so I decided to include some drawing also into the texture of paper origami backgrounds, so I preserve continuity. This is a really long procedure as I have to print out each uv snapshot in a very high resolution (1024 by 1024 dpi- so bigger than A3) and finally- DRAW on every each one of them! When these were assigned to objects...they again didn't really satisfied my expectations as the resolution was still low, although I draw them enormous. They were also keep crushing Maya, no matter how small I made them.

It seemed that assigning hand drawn textures will be easy, but everytime I tried to assign new texture it collapsed Maya, so I tried out some new approaches to texturing:

(Take into the account that we would like to keep ramps, glows and lighting effects from Maya on our objects + give them papery look).

For generating origami paper textures I sometimes this cool Photoshop plug-in (called IMAGE SYNTH) which makes pattern out of a chosen part:

I FOUND 3 SOLUTIONS TO OUR PROBLEM:

1. LAYERED SHADER

There is a possibility of LAYERED SHADER, when you can actually layer different textures one on top of another (for instance ramp on top of drawing and paper bump map).

THE PROBLEM: I would like to keep those beautiful ramps and glows at 100% and also the new under-layer paper texture at 100% strength, but with layered texturing this isn't possible. Each one takes a percentage of the whole.

2. EXPORTING FROM MAYA AS .OBJ FILE(object file, which is recognised by most of 3D software), IMPORTING INTO PHOTOSHOP CS4 AS A 3D OBJECT.

It is also possible to import the object directly into After Effects, but this extremely slow down 3D camera.

PROBLEM: This tool is amazing an it is a miracle to actually have an option of dealing with 3D object in 2D software. You can actually paint on the texture straight away. But the problem is that this option i still in the process of development and it does not allow importing NURBS. When imported, they re badly deformed. ALSO another problem is that any texturing done in Maya with it's default textures (ramps for instance) is here lost and needs to be redone.

3. CREATING A PSD NETWORK.

This is a great option offering direct response from Photoshop. The object is connected in a network and all the texturing aspects (like bumpmap, transparency, colour,... depends what you choose to be accessible from a list) can be quickly altered in Photoshop and simply updated in Maya with clicking on "update PSD network" . This option is great for someone who would like to have more control over the texture.

Before (just ramps):

After:

THE PAPER CRANE

2009-03-19T20:18:00.000-07:00

The PARER CRANE was quite hard to model although id doesn't look hard at the first sight. I modeled it from one single plane the same way as if I was folding it.

This was my reference:

Texturing the CRANE:

UV mapping in this case was very hard as it is very organic. I managed to sort out the UVs this way:

After printing out the UV snapshot as PNG 1024 by 1024 dpi file (this means huge file but it is the only way to have drawing nicely seen in the 3D space) I am in a search of a texture that would resemble the fact that Sadako used any aper she found o do the paper cranes- either some nice wrapping paper or some printed texture with fonts and signs resembling medication box (se used any material she fond in a hospital).

On the end I decided to paint myself more traditional pattern, which would reflect the traditional gift packaging, but also be very subtle, so the whiteness of the crane wouldn't be too disturbed on a look:

Crane Renders:

MODELING

2009-03-19T20:09:00.000-07:00

MODELING- general

Once we made a decision that the backgrounds will be done in 3D (I didn't want to loose the opportunity to learn more about maya and improve skills), I started to experiment with different modelling techniques in order to achieve the look of paper, with all its quality of fragile thinness, profusion, organic folds and lightness.

Modeling with ILLUSTRATOR FILES

I have been trying different ways in how to model flat objects with holes in Maya and then one day after loads of experimenting, came up with an idea:

To CREATE a selection in Photoshop- make a path, add the path to a new layer and export it as an * ai file (illustrator), because this way the drawing becomes accessible as a curve when imported into Maya. Then we choose the outern curve, then inside ones and go to the surfaces- planar (option menu gives us chance to decide the poly count for example). I advise to use QUADS, not triangles as they deform surface.

Reference object:

3D Model:

MODELING WITH POLYS, NURBS AND SUBDIVS

2009-02-11T13:26:00.000-08:00

To achieve origami and paper cutout feel, I need to mix different modelling techniques. Flat objects are mainly done with create polygon tool and then given a more spatial form with with lattice tool. 3D objcts are mainly polygons or NURBS.

( Maya- animation menu- create deformers-lattice (define points))

This is the same scene, but here are background planes already finished. The lower part is mainly done with nurb modelling as it is much easier to reshape into a papery look and also the surfaces are much more smooth when the light hits them.

PUDDLE TOP SCENE:

The lotus flower and fish body are made out of polygons. the folds on fish body were very hard to make. I used cut faces tool and adjusted each edge separately, to give it paper feel. All the rest of parts of fish are NURBS, except eyes. The lotus leaves are done with cut polygon tool in 2 halves, then merged together. The lotus petal was made out of polys and then duplicated with a special option (edit- duplicate special). To achieve organic paper look (all the folds, transparency, fragility,...) I used NURB modelling with the compilation of Polys (the fish body is all made out of polygons using "cut polygon tool, to add edegs- this was ...hell of a torture!)

The lotus leave is also poly...but this time done in 2 halves with create polygon tool, then combined and merged, then formed into an organic shape with lattice tool.

Here we go: 2 more scenes that I have to model as Ajdin Barucija in the middle of sudden decided not to co-operate on my project (in week 7!) and despite dissertation hand-in, I have to replace them in one weekend:

I am planning to make very complex LIGHTING in these scenes, a lot of light linking, but before that even more complicated thing- UV mapping and texturing!

PHOTO REFERENCE AND MOOD BOARD RESEARCH

2008-10-23T17:33:00.003-07:00

FIRST REFERENCE PHOTOGRAPHS

Were done at college with a Japanese student. I was trying to catch some moments from the film to clarify my vision of the actions, camera angles, deformation of body in different perspective angles, the foreshortening, the facial expressions, positioning of the character in the space,the general feel and colour.

A few days ago I met a really beautiful Japanese girl called Ayano. She resembles the spirit of Sadako in a form of modern Japanese teenager.

I was very happy when she said she would love to pose for reference pictures for my animation:

For the run and the fall:

For the mourning face:

For making if the cranes:

Fir the hand puppet- crane:

The sorrow:

The connection between the "Bomb sequence" and the "Sorrow sequence":

I need these references to see how light behaves when hitting the skin, the Japanese features, facial expressions, the movement of the body.

To study the body movement even better I decided to take camera out and film the action (the run) from different angles, which will help us to decide about timing for animatic and wull aslo serve as a great drawing reference in 3rd term.

http://www.lasieexotique.com/mag_hina/mag_hina.html

Reading about Sadako's life, visiting the clothing department at Albert Victoria Museum, also researching the pictorial material I found out some interesting facts about clothing culture at that time:

http://en.wikipedia.org/wiki/Yukata

http://www.vam.ac.uk/collections/asia/kimono/index.html

school uniforms:

expensive kimono

hospital clothing

FIRST STORYBOARS- ROUGH ANIMATIC, BUILDING NEW ASSETS AND FINAL ANIMATIC

2008-10-23T17:33:00.001-07:00

New Animatic

http://www.youtube.com/watch?v=4BAnYGbkYZI

It was just then that I tried out how it works to put collaged background together with hand drawn body, I also decided to change a few shots to clarify the story a bit more.At that time I still had an idea of putting different animation techniques together (2D, 3D and live action).

Building new assets (backgrounds and character poses)

I redraw all the characters positions with ink and watercolors, to get closer to the preferable look of the film.

Backgrounds were done in Photoshop, using different photo references, collaging them together into a completely new composition. I used only materials that resembled japanese origami or paper cutouts to express the idea of fragility of the subject matter.

ROUGH ANIMATIC

We decided to do the first version of the storyboard. Despite nice look, all its in-between cuts of rhythmic montage made it confusing and rather hard to follow for someone who is unfamiliar with the story.

ROUGH storyboard

After building a Mood Board and group discussion of the two mayor ideas for the film I decided to storyboard them both in pencil. In the first version of the idea we would start with character dying in the beginning and the whole film would be flashback on her fight for achieving her wish to run again. In the second one the time would progress more analytically, we would show her past in the beginning, then progress to the hospital scene where she dies. So first one more optimistic, second more pessimistic version.

THE FABRIC RESEARCH AND CHARACRTER DESIGN

2008-10-23T17:32:00.000-07:00

CHARACTER DESIGN

After an extensive research ( tenderness and beauty of japanese girls in Mulan, sculptures and portraits of French orientalism, puppets and Geisha culture, where I was specially interested in make up, behaviour and clothing, stylist illustration of Dave Mc Kean, Slovenian artists like Rudi Skocirand Gabrijel Stupica's arte brutta portraits of his beloved daughter expressing childis purity and innocence.Jared also advised me to see Japanese Butoh Dance, a traditionally "performed" in white-body makeup with slow hyper-controlled motion.) All these inspired me and so I did a few versions of the character, which varied from very stylist cartoon illustration to more naturalistic approach. I actually draw Sadako full body size as I wanted to get the real feel of the character, compare it to the corpulenceof my body size and proportions(she was 135 cm tall and had only 35 kg). This helped me to define the general look of the character and its animation.

She was drawn nude as it is easier for me to define bone strcture and complexion, contour and how far I can reach in abstraction of the shape.

Research:

FROM ESSENTIAL IDEA TO THE FIRST PITCH

2008-10-23T16:52:00.000-07:00

FROM ILLUSTRATIONS OF ESSENTIAL IDEA to FIRST PITCH

The Idea of Film derived from a memory of a book that I read as a child. The essential illustrations were made randomly, using reference photographs from internet and my own vision of the story.