. : Elliot Forrest Animation : .

Collaboration With Fashion

After having the Fashion Photography students pitch their ideas to us as a group, it was evident that I would have enjoyed working on quite a few of their projects. With my other work in mind I decided just to work with Jacob on one project involving an idea loosely based around a choreographed dance sequence.

Jacob, Daniela and I met to brainstorm some ideas on both the technical aspects of proposed ideas and inspiration that we felt could serve as a loose guide to head towards. We found a few videos and bounced around the idea of having the dancer interacting with created geometry whilst cutting in various effects too. A video called nuance served as a pretty good guide to what Daniela had in mind and whist she didn't want to copy the video exactly whenever we chatted about what she had in mind, this was the style she kept coming back to.

As well as this, Daniela had another idea in which she wanted to film and have us colour grade for her. We had all decided after managing to book out the performance center that we'd try to acquire the Black magic to film on as we would then have raw footage to manipulate the colour information at an uncompressed level with a higher amount of control.

After deliberating long and hard over a track for the dancer to dance to we entered the television studio with lights, the black magic and a stereo. The idea for the piece was to shoot lots of different cuts and build a story with the clips in the edit so the stereo was predominantly for the dancer. Jacob and I gave input as we were shooting into some movements that could allow us to input creatively within the post production process. After transferring the footage onto my hard-drive, I put together a low resolution version of the clips in a single timeline within Da vinci Resolve. This process would allow us to easily transfer footage to and from Daniela over the web. We also put a frame counter on each clip so we could easily match and identify positions within the low resolution clips against the raw footage we would eventually be editing. The process of Daniela editing was quite a time consuming one. Jacob and I both dropped into the Digi suite at numerous times to see how the edit was going. To be completely honest this part I feel should have been done by myself and I did suggest this. Daniela made it very evident that she was not in favour of my suggestion and my keenness to assist her in the editing of the video was knocked at most suggestions. The edit to me was not working. Dynamically the video did not build from start to finish and the dance sequence was hardly ever on the beat. This made it very difficult when adding any effects as everything looked slightly out of position.

After retrieving the edit from Daniela, Jacob and I got to work at adding in some previously mentioned techniques to interact with the dancer. We had always intended to use lots of geometrical shapes and created all of these within after effects. We animated the paths, sizes and opacities to create something that could be perceived as an interaction between real and virtual. This process definitely would have benefited with some extra management. We had a minute and a half to fill and with the speed of a lot of the cuts it was really quite hard to sell this idea, especially with the motion of the dancer often not falling on the beats of the music.

We used particle effects, motion blurs and blend modes to create a varied composition. Unlike the nuance video, our composition was feeling very rushed. Each dance move was high in motion, lots of movement and not a lot of space for us to sell the idea of an interaction. Had the movements of the dancer been slowed down we would have been able to do a lot more. As we worked to fill the video, we met with Daniela a few times to discuss her opinion on what we had done and to my surprise she was happy with the way the it was looking. We signed the video off, forgetting that of course we were working at a low resolution and outputted to 1080p H.264.

The other shoot Daniela wanted us to help with required, UV light, UV body paint and another dance sequence. I felt a lot more promising about this one. Having said that my enthusiasm was soon damaged after being asked to turn up at 2pm and not filming anything until 10pm. The footage pulsated circumstantial to the phase of the UV light and the one light we managed to acquire was nowhere near enough to allow us to see properly through the lens. After filming and transferring the footage we came to the decision that, had we shot again more lights would be required to provide the desired clarity.

Having come out the other side of this project, I feel that I learned some valuable lessons. Until proven otherwise I know best. This is not an egotistical statement but for someone who doesn’t know the techniques and the requirements of the vfx process, then a backfoot really needs to be taken. That’s not me saying that she couldn't have been involved with myself editing but to listen to suggestions and take advice when given at least would have improved the composition tenfold.

Time management is not a joke. It's not funny when you have to wait around for 8 or 9 hours when it really isn't necessary to do so in the first place. Daniela's need for us to be with her at all moments was taking its toll. It was our duty to talk technicalities and help along a little with the aesthetic, we would have benefited from storyboarding the whole dance sequence, and ensure that the motion fell on the correct beats to enhance the animation parts.

This being said, Daniela's reluctance to take advice and manage our time poorly has only sparked a wariness of how to deal with difficult situations in the real world. It's not all pink and fluffy. I'm grateful for the requirements of a challenging client as it has enabled me to see clearly how important the pre-production process really is. Whilst my main interest lie within the post-production process it is still my duty, especially when dealing with clients, to be involved with all stages of the production, if not just for the end result, but for my own health as well.

11 Second Club

Project : Deep

Deep

Our production towards the deep project started with several meetings clarifying what scenes it was that we chose to do. Jacob and I had taken an interest in the opening castle scene and agreed to creating this whilst I agreed with Bradley to do the 3d hotel scene.

Production started with Jacob and I gathering some reference images to base our scene on. The idea was originally to build a set extension within Maya. After experiencing lots of problems with the projection mapping technique we had decided that perhaps it was best to achieve the end result by means of matte painting different layers within Photoshop and layering them within 3d space within after effects. We broke the scene down into 6 layers; foreground, front middle, Live action footage, back middle, mountain background and clouds. The live action footage obviously setting the speed at which we track into the castle we placed the layers in 3d space and set up stabilising the footage. As the footage was very shaky it proved quite difficult to eradicate all of the shake but actually enjoy the slight shake. It was our duty to allow the viewer to tell that the castle was a model and was not created in any other medium. After stabilising the footage we could then start on animating a virtual camera within after effects to match the move of the live footage.

We distanced the 2d planes to create a sense of depth and scaled them accordingly. We ensured that when creating the painted layers the resolutions were ample for scaling. To create the clouds we used a turbulent noise map and distorted the shape as to take on the sky above the scene. We animated the sub-scaling properties of the noise to create that effect of clouds lower in the sky passing larger ones. Once we had all of the elements, we balanced the levels and converted our colour images to grayscale. We added an adjustment layer to boost the contrast which instantly helped reinforce that hitchcock style piece we were hoping to achieve.

The production with Bradley started actually quite late in the game. Bradley worked well in producing a clean plate for me to use as a background in my 3d scene.

After creating the objects for the table I had made the decision of actually recreating the whole scene. The combination of real footage and cg elements I was struggling to marry in a convincing way. I therefore started to model and texture all elements of the hotel room, paying close attention to the texture of all things. The lighting was also an important aspect of the overall ambience and although I was not in the room when filming was completed I gathered that there was really only one light source in the way of the sun shining through the window.

I set up an area light ensuring that the direction was correct, it was casting physically accurate shadows and added a tinge of yellow to create a certain warmth apparent whilst I studied the footage. After all of my items were modelled and texture it was time to get on to the animation of the whiskey bottle smash. I initially tried to complete the process within Maya by using an nCloth material and designating the frame of impact, however I was experiencing lots of issues and had remembered that a few years ago I had played around with the physics engine within 3dsmax. I decided to export my whiskey glass and built a scene in max for the glass to interact with.

The way in which the shards of glass interact with other objects I felt was important to sell it actually happening. After attaining the desired smash I re-imported the baked animation into maya and lined the camera accordingly for where I was initially told by Bradley to put it.

I manually animated a single shard of Glass to fly towards the camera and fill the screen as I thought this would be enough clarification of the sparkling transition required into the sequin dress of the next scene.

After showing Derek the scene it was apparent that we had slightly different ideas. It was also apparent that the angle of the camera is in an incorrect position. I returned home and set to work at putting these things right. My scene is completely textured and animated however the rendering process is taking too long for me to be able to produce a finished product. Had my initial scenes been approved I would have been able to render straight away and produce an animation.

It did not help that my refraction bounces within Maya had to increase due to the amount of shards layering in front of one another. My whiskey bottle was appearing black originally as there were not enough bounces, as soon as I turned this up my render time went up from 10 minutes per frame to 2 hours.

Whilst I feel the time management part of the project could have been managed better I did enjoy the live brief scenario in working to the desires of another party.

A Light From The Dark

This semester was all about getting involved in team based projects and as became quickly apparent in the studio the third years had all started with their final film projects. Eager to get involved i put myself up for a few projects with the intention maybe to build some models, work on some animation, in fact, i was pretty happy to do anything that encapsulated the 3d pipeline. I quickly received a message asking if i'd like to create some models for a film called a light from the dark which was going to be created in 3d.

On meeting with the project leader Alex, we discussed my role within the team and he started me off by relaying some reference pictures:

So I started to block out the environment in maya. I didn't want to create unnecessary detail so I ensured that I put the detail only where the camera would be shooting. As the door into the Warehouse would be a focal point of the video I felt that it was probably a good idea to start there.

I blocked out the shapes of the door and surrounding area and thought about the necessary animation of the door denting as I knew that the geometry would need to be bent after the monster hits into it. After modelling the door area I continued to texture it. I used a combination of procedural textures within maya and a bit of help from Photoshop to create the door texture which I was told should be rusty and weathered.

I continued to model and texture the surrounding areas. The cladding on the wall around the door area, some filler objects including traffic cones and pallets, the light above the door and so on.

I continued to model and texture and light and build the scene as closely to the guidance i was given. Every week I turned in my progress to Alex and received feedback, sometimes critical and helpful and other times not. As i continued forward and the pressures started to build for everyone involved within the project I decided to use a lot of my own intuition and focused on just getting all of the content in there and to a stage that i was happy with but could revise if necessary.

I started to bulk out the corridor ensuring that i tapered the end where the door was to create the sense of entrapment.

I continued onto the warehouse now where the shelving would consume the majority of the space. I was told that on the shelves would be lots of boxes, crates, junk, bottles, nuts and bolts and so on. I obtained a reference image to use for a guide to building the shelves. I noticed that the holes that allowed the shelves to connect within the frame would be best achieved using transparency maps. This was important in the pursuit of keeping the poly count down, I knew that with the textures the scene was growing quite large and I'd been told to be careful of this.

I plugged a box ramp into the transparency slot of a blinn shader as well as the eccentricity slot to ensure that light was not diffusing seemingly off of thin air.

As well as, modelling, texturing, lighting, Alex suggested that perhaps I do some environment animating. I agreed as I had the best knowledge of the capabilities of the models and my first task was to animate the bend in the door as the monster hit into it. I used blend shapes as I thought this would be the easiest way to animated with the most amount of control. I created a duplicate door from the main model and deformed it in the areas that was necessary to portray the damage. I created a blend-shape deformer and key-framed the slider between 0 being undamaged and 1 being completely damaged.