Interview with Animator
Kenny Frankland,
Creator of
"Mistabishi-Printer-Jam"
Watch it now!
Kenny Frankland first got our attention during the 2007 Aniboom Awards. His 3D animation music video "
Red Ocean" a peaceful journey into the deep blue, won for best music video. The publicity it receieved soon got Kenny on the music video making track and it wasn't long before he was approached to create "
All Hell is Breaking Loose" for London Elektricity. Now he's back with some more candy for your eyes and ears, with Mistabishi's "Paper Jam". Are any of you old enough to remember those good old printers that took forever and mad tons of noise? Ah the good old days....
View Kenny Frankland's Portfolio
What inspired your video?
The music itself was the main inspiration. Obviously the image of a printer came into my head first but I also liked how the start of the tune is quite structured and things get more intense and disjointed as the track plays through. I saw this actually happening to the printer.
The more layers the song gained, the more objects should be presented in the video and the more cluttered the image should become. The idea of its metamorphosis into the creature just came from the noises in the track. The samples of the printer heads sounded like something transforming but I wanted to avoid making it robotic and went for an organic creature instead.
What techniques were used?
The whole piece was created and animated in
3DS MAX, with texture work done in Photoshop and post effects on compositing in
After Effects. The animation was all keyframe and no plug-ins were used. Once you get a feel for the structure of the beats it becomes a quicker process. For example the key frames of the main drum segments were always 2,18,36,52,70,87,104,121. If I stuck to this structure, things fell into place nicely.
How long did it take?
Five weeks - from start to delivering the master - though I was working up to 18 hours a day to meet this deadline!
Do you have a favorite sequence in the movie?
I like the transformation scenes. I was dreading these, but they turned out ok.
Can you tell us a bit about the process of turning sound into visual?
I started by taking the track and cutting it up into individual numbered sections then wrote down what I saw happening in those sections and what objects are introduced. I then made a very rough animatic using sketches, rough renders and camcorder footage of one of my old printers.
Once I had an idea of the flow of the piece I did research on the things I wanted to make and then modeled the room, printer, creature and objects, then started animating to each of the individual sections of the track I’d cut up. I’d make sure I had stuff to render overnight and once these sections were rendered I placed them into the animatic so I could have constantly changing stuff to show Hospital Records. When all renders were finished I put them all together, added effects, TV noise and the ‘printer garbage’ to gel everything together.
Watch Mistabishi Paper Jam
Were there any big challenges?
My biggest challenge was always going to be render times. I only had five weeks to produce the video and don’t have access to a render farm - the cost of outsourcing rendering would have been 10x the budget so I had to think of a way to have nice speedy render times, but not sacrifice the look of the piece. I wanted that soft semi realistic painted look, but that means using depth of field and soft shadows and they can take an age to render.
To get past this I used a lot of planning and a lot of post render options to gain the effects I wanted. I started by modeling, texturing and lighting the complete room then baked all the textures so the soft shadows became part of the original texture. This meant I could then render the room as a background plate without using any lights. It took a while to setup but ended up reducing my render times from 4 minutes a frame down to 1 second! I then rendered the printer, creature and objects on a shadow/matte material so I got the shadows then comped the elements together.
The depth of field look was achieved by rendering a ZDepth pass of the scene which is then used later to calculate the distance for post blurring. Using these techniques slashed my render times and gave me more control over each element.
What do you see yourself doing in the future?
I would like to keep doing what I'm doing now (freelancing) but with a more regular flow of work. At the moment I'll do a dull but well paid corporate animation job for a couple of months, then either a music video (which doesn't pay so great but gets me fantastic exposure) or work on my personal film for a month (which doesn't pay at all!)
My goal is to get my film finished, maybe split up as a series of shorts and try and generate an income via that. I have many stories I want to tell with these characters.
Watch "Mistabishi Paper Jam"
View Kenny's Portfolio