Over the last 18 months, Rushes has been working closely with Dangerous Films to produce VFX for Human Body: Ultimate Machine, a four-part for Discovery US that begins showing in the UK on Thursday. In all, 400 VFX shots totalling 80 minutes duration were delivered. In the US the show is titled Human Body: Pushing the Limits

A graphic and in-depth exploration into the functioning and phenomena of the human body, this series reveals the human body as it has never been seen before. Real-life examples are used to illustrate its most diverse moments of strength, endurance and survival. From strength to sight, pain to sleep, heat to eating, the entire range of body challenges is explored.

Dangerous producer Tim Goodchild’s brief for Rushes was to create a CG Human that would show the wonder and beauty of the human body, remaining close to reality, while avoiding the potential for squeamishness on the part of the viewer.

Originally the intention was for the body to fit one presenter, however as the scripts developed and the work of R&D on the model progressed, it became clear that it would need to fit to both men and women, small and tall. Other issues such as potential nudity problems also had to be addressed as did the re-scaling of the anatomy and animation rig to fit the different cast.


Hayden Jones, VFX supervisor at Rushes, comments, "At this point, we took the view that we would need to spend longer ensuring that our model could be used in as many ways as possible. We developed a workflow that put constraints on the rig in order that any animator could pick up the model and know that it would only do, physiologically, what a real human body does. We split the model into various layers of bone, muscle, nerve systems and so on so that we could ‘dial through’ as many as we needed in the final composites."

The stories in the series place the body in extreme situations that couldn’t be filmed and so Dangerous went to Mike Tucker (series SFX supervisor) at The Model Unit to help solve some of the more difficult ones. To this end SFX models of caves, houses exploding, fireballs in forests and motion control were filmed at Ealing Studios in London.

Mike's unit also provided art department construction and SFX for all the studio filming in the UK. All the other filming took place in the US based out of LA and involved a huge range of different locations and stories across the series.

Hayden Jones adds, "It was such a great pleasure to work with Mike Tucker. He completely understands where his models will fit into the final composite and so the combination of his work with the live action and our CG gives the VFX shots a real depth and richness."

Louise Hussey, VFX producer at Rushes says, "This project is one of the largest that Rushes Film & TV has completed to date. We are incredibly proud of the standard of the VFX that we have produced, in that they fully encompass a broad spectrum of approaches and skills. Our previous work with Tim Goodchild and Dangerous has given us a short-hand and trust that allowed us to produce a large range of diverse effects with a relatively small team."

The 4-x-1 hour programmes were all offlined at St Annes in their Avid suites and Simon Brookes onlined the programmes in DS Nitris, conforming the programmes and creating the information graphics sequences that enhance the science of the series.


Rushes’ VFX supervisor Hayden Jones explains the technical processes: "Rushes were brought onto the project to advise at early script stage and produced various conceptual design and workflow tests. This proved that we could deliver a consistently high standard of imagery at HD, whilst keeping the whole process flexible enough to allow for additional shots all the way through the production process."

"Work was started on the visible human aspect of the VFX, and knowing that this would be the majority of the VFX workload, we wanted to get the pipeline as robust and easy to use as possible. In discussions with Tim Goodchild, it was decided to have three months VFX preproduction with two senior artists. This allowed for the model to be fully rigged ready for animation, and to create a visual look for the human model that Dangerous Films and Discovery could approve."

The rigging process, when virtual skeleton is placed inside the model to allow animators to bring it to life, was under taken by Rushes’ artist Mark Pascoe. The rigging was undertaken in Rushes main 3D package, Autodesk's Maya. Initially the muscle simulation techniques were looked into, to allow the visible human to realistically flex his muscles as the animators moved him around.

The Comet Muscle plug-in was chosen, as it had the most complete toolset for the purposes. Mark meticulously rigged and checked all of the 200 major muscle groups making sure they worked correctly and then switched attention to the other body systems. Using custom scripts driving deformers, Mark kept all the body systems packed together and as the production progressed and in order to fit the visible human inside many different sizes and shapes of actor, Mark devised a combination of Mel scripts to allow the main rig to be squashed and stretched to fit anyone. This was a major challenge as all of the scaling of limbs and torso had to take place whilst keeping the underlying human mechanics undisturbed.


The model was purchased from Zygote, who specialise in medical visualisation models. The models have a human dataset with ninety percent of the body systems that were needed to be shown however, the textures were not of sufficient resolution for many of the close up shots at full HD that Tim Goodchild had outlined. The job of creating a set of ultra high resolution textures was given to matte painter Alex Jenyon who used a combination of ZBrush and Photoshop, together with a library of medical photographs.

During the look development stages, VFX supervisor Hayden Jones also designed the shading and lighting pipelines. The project needed a very intense rendering exercise therefore careful measure had to be paid to not only achieving a high quality of imagery, but also to keep the rendering times down. Rushes chose Pixar's Renderman as the rendering package.

"Human Body was the first major project to go through Rushes Film & TV using the specially designed high dynamic range (HDR) pipeline, enabling us to light and render scenes quicker and with a much higher visual quality than previously possible with traditional CG lighting techniques:, says Jones. "We used many advanced techniques including ambient occlusion, reflection occlusion and selective specularity, and these techniques combined to create a rich visual style and a realistic rendering time."

At the start of production the animation team started making animatics of the major visual effects sequences. This not only helped the directors visualise key moments in their scripts but also started show them exactly what the visible human could do. The flexibility of the visible human allowed for the directors to come up with some shots they hadn't considered before and whilst creating these animatics the animators could test the rig under production conditions and add any features they needed.


Rushes provided an on set supervisor for all of the VFX photography, either on location or at Ealing Studios with The Model Unit. The supervisor advised on the best way forward to achieve each shot and collected all the data needed to create the shots once they were back at Rushes. Chrome spheres were photographed for lighting information, camera and lens information were recorded to enable camera tracking and measurements for all the actors were taken to help is fit the visible human to each person.

Once the plates were back at Rushes the shots were started with the animators loosely rotoscoping the action and sending quick playblasts back to the edit suite. Final animation was completed before passing it to the lighting and rendering team, headed up by Rushes senior artist Angela Noble. The chrome sphere information was taken and placed into custom light shaders.

The rendering was handled by Rushes’ 96 processor render farm. The render process not only computed the usual "beauty" pass, but it also rendered out 6 more helper passes designed to help composite the visible human into the plates.

The composite team used Apple's Shake as their primary program. However several complex 3D shots that took SEM photography as the plate also utilised Eyeon's Fusion. The compositors created a look that could be uniformly replicated throughout the shows, giving a consistency across the whole series. The extra passes were utilised in the comp to enable the compositors complete control, from small colour tweaks to virtual relighting the CGI.