Tuesday, August 20, 2013

This film is the output of a two year part time Masters I studied at the Duncan of Jordanstone College of Art & Design, part of Dundee University.

I have used my research into the physics of the movement of light to produce a short film, which is an artistic representation of this invisible world of light that flows around us. With the wealth of work being conducted into the understanding of light by the likes of Dr Maria Ana Cataluna here at Dundee and Ramesh Raskar based at MIT in Boston, there is more and more interest in seeing the physical attributes of light. To achieve the results desired I also developed a technique of controlling the organic movement of computer-generated particles by using hidden control clouds that have influence over the visible particle systems. This allowed for large-scale simulations to be built in the ICE particle world within the Softimage software environment rapidly, and fine-tuned with the traditional node based controllers.

You can see the process of how I developed this project here:

vimeo.com/channels/568932

Thursday, August 8, 2013
Tuesday, August 6, 2013
Over the course of this masters I feel I have developed skills which I was unaware of at the outset. I came onto the Msc program with the goal of exploring animation techniques and visual effects, and personally over the last 2 years I feel that has been a great success. Now I am able to interpret a idea and not only visualise how I hope for the output to appear but also chart the journey that would be needed to achieve those results. The course has given me a wide appreciation of the different facets that exist in animation and vfx, as well as how much cross over there is between each department. You can not develop a rig for a character that has not been modeled to sufficiently accept the movements, when rendering if your textures and lighting are not optimised for your cameras it will be near impossible to produce a final image. But I focused primarily on particle simulations through out this masters, and specifically in the software environment XSi or Softimage as it is also known.

 

This software allowed for a more intuitive production of simulations than I personally found Maya did, but as with all software programs it came with its limitations. The main issue through out the masters is the lack of reference material for XSi that is available, the university only carry documentation for version 5 published in 2006. But all members of the Msc course do have access to the digital tutors online training resource which has been absolutely vital to my training in using the software as it has a wealth of tutorials covering many aspects of the software. The user guide built into softimage was also very helpful when delving further into the software. I also found the lectures given by Andy Nicholas on Softimage during semester 2 very helpful, as being shown in the flesh how the program functions was a useful learning experience.

 

All of these tools and experiences have helped me develop from having never used XSi before 2 years ago to being proficient enough, specifically in the ICE particle simulation environment, to produce the kind of results seen in the short film ‘Light and Sight: an interpretation of light’ made through out the masters.

 

This film was intended as the hook that I used to hang all of the teaching and development I produced through out my time on the course, but now looking back it has been more than that. I have thoroughly enjoyed exploring the science of light, and the physics behind it. The conversations online with others and reading through the papers published on the subject gave me a great appreciation for just how complex the world around us is. Being able to spend time in person with Dr Maria Ana Cataluna was an absolute eye opener, and listening back to our conversation I felt I needed to include her words in the final film. Subdued and washed out, the echo of our discussion rumbles in the back ground, reflecting the way her words and insights bounced around my mind through out the creative practice of producing the visible simulations.

 

Using real world footage, taken at a relatively remote location was also a challenge to collect but the location was ideal for what I had in mind. The technical challenge of transforming this raw footage in to particle simulations was also a massive hurdle. My calculations showed that I needed in the region of 2,073,600 (1920 x 1080) particles to achieve the transitional image that would allow the move from live action to particle world. The final cached particle simulation actually totalled just over 450,000. Which is why I originally used a particle duplication engine to up this to over the 2,000,000 mark, although I eventually only used these shots as a background padding, softened to remove the flicker which the duplication introduced.

 

Bringing all of these scenes together in Nuke using, the GPU accelerated render engine Exocortex Fury out of XSi, again allowed me to explore a new area of visual effects, and also helped me to push the look of the film further. The power of compositing applications like Nuke really did surprise me; there is so much capability when working with even half float EXR files to alter them in the comp.

 

Overall I am proud of the results that I have achieved over the course of this masters, not only in the main film produced, but also in the skills learnt which I can take into new projects now. My best work is still ahead of me, and I look forward to the new challenges that await.

Over the course of this masters I feel I have developed skills which I was unaware of at the outset. I came onto the Msc program with the goal of exploring animation techniques and visual effects, and personally over the last 2 years I feel that has been a great success. Now I am able to interpret a idea and not only visualise how I hope for the output to appear but also chart the journey that would be needed to achieve those results. The course has given me a wide appreciation of the different facets that exist in animation and vfx, as well as how much cross over there is between each department. You can not develop a rig for a character that has not been modeled to sufficiently accept the movements, when rendering if your textures and lighting are not optimised for your cameras it will be near impossible to produce a final image. But I focused primarily on particle simulations through out this masters, and specifically in the software environment XSi or Softimage as it is also known.

 

This software allowed for a more intuitive production of simulations than I personally found Maya did, but as with all software programs it came with its limitations. The main issue through out the masters is the lack of reference material for XSi that is available, the university only carry documentation for version 5 published in 2006. But all members of the Msc course do have access to the digital tutors online training resource which has been absolutely vital to my training in using the software as it has a wealth of tutorials covering many aspects of the software. The user guide built into softimage was also very helpful when delving further into the software. I also found the lectures given by Andy Nicholas on Softimage during semester 2 very helpful, as being shown in the flesh how the program functions was a useful learning experience.

 

All of these tools and experiences have helped me develop from having never used XSi before 2 years ago to being proficient enough, specifically in the ICE particle simulation environment, to produce the kind of results seen in the short film ‘Light and Sight: an interpretation of light’ made through out the masters.

 

This film was intended as the hook that I used to hang all of the teaching and development I produced through out my time on the course, but now looking back it has been more than that. I have thoroughly enjoyed exploring the science of light, and the physics behind it. The conversations online with others and reading through the papers published on the subject gave me a great appreciation for just how complex the world around us is. Being able to spend time in person with Dr Maria Ana Cataluna was an absolute eye opener, and listening back to our conversation I felt I needed to include her words in the final film. Subdued and washed out, the echo of our discussion rumbles in the back ground, reflecting the way her words and insights bounced around my mind through out the creative practice of producing the visible simulations.

 

Using real world footage, taken at a relatively remote location was also a challenge to collect but the location was ideal for what I had in mind. The technical challenge of transforming this raw footage in to particle simulations was also a massive hurdle. My calculations showed that I needed in the region of 2,073,600 (1920 x 1080) particles to achieve the transitional image that would allow the move from live action to particle world. The final cached particle simulation actually totalled just over 450,000. Which is why I originally used a particle duplication engine to up this to over the 2,000,000 mark, although I eventually only used these shots as a background padding, softened to remove the flicker which the duplication introduced.

 

Bringing all of these scenes together in Nuke using, the GPU accelerated render engine Exocortex Fury out of XSi, again allowed me to explore a new area of visual effects, and also helped me to push the look of the film further. The power of compositing applications like Nuke really did surprise me; there is so much capability when working with even half float EXR files to alter them in the comp.

 

Overall I am proud of the results that I have achieved over the course of this masters, not only in the main film produced, but also in the skills learnt which I can take into new projects now. My best work is still ahead of me, and I look forward to the new challenges that await.

Text to particle transition breakdowns.

Having spent time developing the technique of using a invisible control cloud to influence visible particles I have turned my attention to the text that features in my film. I wanted to lift this section a little bit, as using ‘standard’ text i felt fell flat amongst the rest of the piece.

To do this, as shown in the breakdown above, they were built in 5 stages:

Stage 1

Text is created in Adobe Illustrator and exported out as a jpg file.

Stage 2

Grid is built in Xsi to match scale of jpg, then jpg applied to grid as texture map.

Stage 3

Invisible turbulized cloud emitted from sphere and dragged across the grid.


Stage 4

Visible cloud emitted from texture map, filtered by scale to remove any black areas, and driven towards invisible cloud.

Stage 5

Passed through Exocortex Fury render engine and output as exr with Illumination, reflection, depth and beauty passes.

This setup works really well for me now, and allows for rapid production of simulations for a variety of different styles.

Monday, August 5, 2013

Pulling together all of the work put into the project now, these still images taken from the final edit show the consistency of pallet used through out.

Using Exocortex Fury 2 as my renderer through out the final scenes, I actually ended up omitting particle duplication technique that was the main reason for choosing Fury over Mental Ray as my renderer. Instead I focused on developing the look of the particles in the comp, but using a combination of layers that separated out the passes and then applying a diffuse and glow to the backgrounds.

I also incorporated the original particle duplicated passes if required to the background, but heavily blurred. This is because of the flickering artifact that particle duplication introduced, blurring would remove.

Overall though I am proud of the way these scenes have come out and achieved the look I set out to at the beginning.

The audio for my film has been an important feature right from the outset, and I wanted to create something that would reflect the images, story and help to carry piece.

I began by trying to find a music bed that would give the ambient empty sounds that i needed. For this I settled on Moby: Ana. As mentioned earlier in the blog, Moby runs a website that allows tracks of his to be licensed out to individuals for incorporation into art projects. I applied for this through the mobygratis website, giving a full breakdown of how I intended to use the audio and to my surprise was granted the permission.

As well as this I included a manipulated drone effect, this was because in early tests I used this as padding to give it some kind audio and found that people actually felt it fitted quite well. Slowing down a basic drone sound to 1 percent whilst maintaining the pitch gives it a groan and flat sound, and when cut and restarted it gives a really nice pop/scratch sound that fits in with the cuts in the simulation.

On top of all of this i worked on the audio i captured when interviewing Dr Maria Ana Cataluna. Tweaking the treble and applying a healthy reverb, which is then flattened in the eq settings to almost hide it amongst the music bed and sound effects.

Sunday, August 4, 2013
The Msc Uni webpage is now up, and I have started working with the image above as a poster for my project. I also rewrote a description of my project for use on the site:
I have used my research into the physics of the movement of light to produce a short film, which is an artistic representation of this invisible world of light that flows around us. With the wealth of work being conducted into the understanding of light by the likes of Dr Maria Ana Cataluna here at Dundee and Ramesh Raskar based at MIT in Boston, there is more and more interest in seeing the physical attributes of light. To achieve the results desired I also developed a technique of controlling the organic movement of computer-generated particles by using hidden control clouds that have influence over the visible particle systems. This allowed for large-scale simulations to be built in the ICE particle world within the Softimage software environment rapidly, and fine-tuned with the traditional node based controllers.

The Msc Uni webpage is now up, and I have started working with the image above as a poster for my project. I also rewrote a description of my project for use on the site:

I have used my research into the physics of the movement of light to produce a short film, which is an artistic representation of this invisible world of light that flows around us. With the wealth of work being conducted into the understanding of light by the likes of Dr Maria Ana Cataluna here at Dundee and Ramesh Raskar based at MIT in Boston, there is more and more interest in seeing the physical attributes of light. To achieve the results desired I also developed a technique of controlling the organic movement of computer-generated particles by using hidden control clouds that have influence over the visible particle systems. This allowed for large-scale simulations to be built in the ICE particle world within the Softimage software environment rapidly, and fine-tuned with the traditional node based controllers.

Shot breakdown of Sh006, with a replaced Iris and colour grading work.

Thursday, August 1, 2013

Coming towards the end of the Msc now at DjCad, and today was the last run through of our presentations prior assessment.

Good feedback all round, and now to make a few tweaks before final submission.

Sunday, July 28, 2013

Further work on Sh013 including a breakdown of the construction.

This is the final live action shot of the sequence. Constructed using a animated (rough) eye ball model to interact with the particle simulation. Then the output EXR files are worked up in Nuke, and finally comped onto the live action which is graded in a mixture of Magic Bullet Colorista and Resolve.