I've been excited for FMX pretty much since I booked the tickets and as last week crept closer still that energy built up to a crescendo on the late night train to Stuttgart. The conference is an opportunity I had hoped Jak and I would be able to attend and thanks to Gareth we were fortunately allowed to. FMX its about VFX in film, animation and games with this year a big focus being VR. Talks focused on topics ranging from pipelines and production to art direction and rendering engines, a nice diverse selection for us to dip our toes into when we pleased. It was with a great deal of anticipation, excitement and a slight lack of sleep that the two of us walked through the doors of FMX on Tuesday ready to have our monds blown at some CG goodness.
Double Negative started the morning off and didn't disappoint with a presentation of the physics behind Interstellar, specifically on the Wormhole and Black hole and the resulting wibbly wobbly timey wimey they entailed. This involved a miniature genral relativity 101 to get everyone up to speed on how gravity alters space-time and how we can visually see these effects in satellite imagery and observations on Earth. As some of you may already know Interstellar lauded itself over how accurate the film's physics were and indeed the talk made this abundantly clear delving into the struggle that the studio went to in creating the visuals for the film. While an all new gravitational renderer is perhaps the most immediate result of this it sounds like that by the end of the project most of the team had a sound understanding of space-time and gravitational relativity. It was very cool to see the process and challenges that DNeg went through to create, what I feel at least, are some of the most stunning space imagery in cinema and it's amazing to see how the visuals and scientific underpinning come together to create this. After watching the film I initially felt that there must have been a lack of artistic direction in creating the shots, essentially the team would be driven primarily from the formulae used in the render engine, which for the most part was the case as expected but there were one or two instances where a more creative approach was taken.
Following on from this Weta Digital gave a small talk on their in-house renderer Manuka which was used for a handful of shots on the recent Planet of the Apes and the last Hobbit films. The high end CG stuff is an area that over course of my time in BMW has grown but I would be lying if I said I had anything more than a basic understanding of terminology or how things work under the hood. With talks like this I was able to get a good handle on a lot of the surface stuff, raytracing, shadow mapping, etc however once it started to get a little more technical and especially once the room was opened for questions and people were inquiring on specific areas or issues I could feel the discussion begin to go a bit over my head. What was shown off however was vey impressive especially the hair and fur results in Apes and interior testing of the software.
Jak and I then went along for a breakdown on some of the visual effects (VFX) for the TV series Sleepy Hollow by Pixomondo which pretty much looked at the timescale and how the selected shots were created. While nothing incredibly new it was cool to see the full picture of a development rather then the short breakdowns that VFX studios usually put up. I think it's fairly well known that VFX timescales can be pretty gruelling and it felt that in TV productions this was felt the most with quick turn-arounds and a fast rate of output to meet a weekly show's demands. This doesn't exactly advertises itself as sounding too enticing but the fast turn-around and the variety of projects and tasks that a studio are required to do is something that's quite appealing to me. For many VFX houses a similar experience can be found in their advertising arms and with an increased level of creative freedom for the most part in advertising it is an area that I'm thinking of looking into for the future.
After lunch there was perhaps one of my favourite talks of the whole week regarding the production of The Lego Movie by Animal Logic and the numerous issues and challenges they faced over the course of development. It was great to see their research not only during pre-production but continuing all the way through until near the end of production once the look development was complete. Cue numerous shots of lego in the sun. It's interesting to see the process of closing in on the final look of the film both technically and artistically but the major issue that dogged the production was AL's render engine, which I believe they said was Renderman (a render engine by Pixar) which pretty much had to be massively overhauled with an in-house plug-in/rework to meet the deadlines as well as keeping the overall artistic quality to the high level that the studio was gunning for. I think they also noted that they had been able to render out quadrillion or something polygon scene using their changes in an amount of time that wasn't totally ridiculous which was pretty awesome to see if not completely unlikely to happen in in a production anytime soon. It was cool to see the comparison between the real lego versus the final CG lego in the film and how near identical they eventually got it to be but most of all it was really interesting just hearing the challenges faced by AL and how the team went about trying to solve the problem without making any compromises with the final product.
Pixar then took to the stage discussing some of the new tech in Renderman which was rolling out into their upcoming films, first of which being Finding Dory and in a similar vein to the earlier Weta talk some of the stuff did a little above me but I did grasp most of what was discussed. The new models and textures were shown for Dory and Nemo's dad who can't remember the name of, nor can I be bothered looking it up sorry, which were looking pretty fine and as a result I'm eager to see how the final film looks.
The keynote for Tuesday followed which was twenty years of Weta and the talk mostly revolved around the Hobbit and the LoTR and the differences between the roles Weta provided for Peter Jackson. It was a really interesting insight to see the differences in tackling both trilogies and how a large number of solutions on Lotr such as forced perspective with Gandalf and the Hobbits or the extensive use of miniatures couldn't translate onto Hobbit due to the improved quality of the image. The presentation bounced between similar topics between the two film series such as Gollum and Rivendell before transitioning to focus more on the recent Battle of Five Armies film which while I'm still to see the film I have seen what I think is most of the VFX breakdowns for the film and what was shown in Stuttgart was more of the same really. I think what I enjoyed most about what was showcased was the large insight into Andy Serkis's role within both productions as Gollum and the development of Mocap over a number of film franchises. It gave a great insight into the development of not only Serkis's performance but also the animators who keyframed a huge portion of both trilogies while keeping true to Serkis's vision for the character.
With that being the climax as such of the day what followed in the evening was about an hour of short films that the speaker had collected over the years and the variety of what was shown was very enjoyable. The talk acted as a nice wind down for the day after a hefty physics lesson to start with leading up to Weta demonstrating their numerous achievements in CGI and film over the last two decades. I felt it was a great start to the event however there was perhaps one thing that bugged me when I noticed it. DMU proclaims itself the best game art course in the UK however while Teeside, Wales, Bournemouth and other British universities were present showcasing their work to the attendees signing up prospective students and hoping to snag a guest lecturer or two, De Montfort was nowhere to be seen which is something I found a bit off. Personally I feel it's a great opportunity for the course to get it's name out to students and industry leaders alike and allows students the chance to attend a variety of talks and meet peers and potential employers which in my mind is an utterly fantastic opportunity. It's something I'm following up on and potentially I might push for the university to attend if theirs enough student support for it which I can't see why there wouldn't be. I don't know, it just seems like a win win if the course puts itself out there that I'm genuinely surprised that they haven't already.
No comments:
Post a Comment