It turns out that the themes for the year's Siggraph were Stereo 3D and Music performances.
Jean Detheux was giving a talk this year and was invited by Siggraph.
I just saw Pixar's 'Tokyo Mater' in 3D. I tried estimating the duration of each shot -
some establishing shots at 3 to 5 seconds - the rest 1 second or less. It seems to me this film was cut very much like a 2D film. It was 'worth it' to watch in 3D.
http://www.youtube.com/watch?v=d5gdy-6Y6RY
http://www.imdb.com/title/tt1330219/
I talked with Ramesh Raskar and John Tumblin about their upcoming book 'Computational Photography'. Ramesh is based in Boston and is working with the Cirque du Soleil on some project. I explained the brain project to them both and asked if they were interested in talking with Munro. I have their business cards. John suggested Marc Levoy might be interested. He is working on the medical imaging side of things.
http://web.media.mit.edu/~raskar/
http://www.cs.northwestern.edu/~jet/
http://graphics.stanford.edu/~levoy/
Friday, August 7, 2009
Thursday, August 6, 2009
Simulating the Balloon Canopy in “Up”
"The sheer number of dynamic bodies in the balloon canopy in “Up” required a reimagining of Pixar’s rigid-body simulation environment and development of concise rigs and exible artists’ tools"
Jon Reisch, Eric Froemling
Pixar Animation Studios
http://news.cnet.com/8301-13772_3-10249396-52.html
Challenges
lots of balloons – tens of thousands
lots of shots – close to 200
lots of action – popping, churning, withering , a character in itself.
Early testing – procedural motion
What makes balloons feel like balloons? – the bouncing. He tried particles – driving balloons using fluid - didn't work. He tried rigid body tests – they liked this
Existing toolset limitations – designed originally with Ratatouille. It became clear not good enough for all of these rigid bodies
step 1 - strip this whole thing out of maya – throwing in 125,000 bodies
Ballooning with 'Ned' – the system - I think it is proprietary
balloons on bottom of hot air balloon are smaller because camera is most often at the bottom
hot air balloon – so the larger balloons at the top of the canopy are still visible and not like funny pixels.
Simulating the strings – didn't do all of them
Building a canopy with shapers and noise to keep it slightly alive.
Canopy Spine Rig – needed more structure
Wilting – did have to cheat – the balloons wanting to come into frame
Hero Balloon Rig - actually simulated the string
Jon Reisch, Eric Froemling
Pixar Animation Studios
http://news.cnet.com/8301-13772_3-10249396-52.html
Challenges
lots of balloons – tens of thousands
lots of shots – close to 200
lots of action – popping, churning, withering , a character in itself.
Early testing – procedural motion
What makes balloons feel like balloons? – the bouncing. He tried particles – driving balloons using fluid - didn't work. He tried rigid body tests – they liked this
Existing toolset limitations – designed originally with Ratatouille. It became clear not good enough for all of these rigid bodies
step 1 - strip this whole thing out of maya – throwing in 125,000 bodies
Ballooning with 'Ned' – the system - I think it is proprietary
balloons on bottom of hot air balloon are smaller because camera is most often at the bottom
hot air balloon – so the larger balloons at the top of the canopy are still visible and not like funny pixels.
Simulating the strings – didn't do all of them
Building a canopy with shapers and noise to keep it slightly alive.
Canopy Spine Rig – needed more structure
Wilting – did have to cheat – the balloons wanting to come into frame
Hero Balloon Rig - actually simulated the string
The whole class creating a collaborative work
Animation in Education
Collaborative Animation Productions Using Original Music in an Unique Teaching Environment
Atlas' Revenge – music used in its entirety (original and royalty free music from teacher with opera piece at the beginning) – nice drama and editing
http://www.atlasrevenge.com/
The music, when it was composed first, gave the best results.
The Young Composer's Challenge – trying to work together with young animators.
University of Central Florida's Department of Digital Media
They did this in 1 year and 1 semester. They all have to learn every step -
each student might light 3 or 4 scenes. 25 students – all working together.
Large production number – if they don't maintain attitude – they are out.
The other students will tell the one or 2 bad apples to pull their socks up.
They do a type of Gong Show – pitch their ideas and the whole class agrees.
4 to 5 classes in every semester.
I wonder if Clare and Maral could enter her 3D animation for next year?
Darlene Hadrika dhadrika@mal.ucf.edu
www.youngcomposerschallenge.org
Collaborative Animation Productions Using Original Music in an Unique Teaching Environment
Atlas' Revenge – music used in its entirety (original and royalty free music from teacher with opera piece at the beginning) – nice drama and editing
http://www.atlasrevenge.com/
The music, when it was composed first, gave the best results.
The Young Composer's Challenge – trying to work together with young animators.
University of Central Florida's Department of Digital Media
They did this in 1 year and 1 semester. They all have to learn every step -
each student might light 3 or 4 scenes. 25 students – all working together.
Large production number – if they don't maintain attitude – they are out.
The other students will tell the one or 2 bad apples to pull their socks up.
They do a type of Gong Show – pitch their ideas and the whole class agrees.
4 to 5 classes in every semester.
I wonder if Clare and Maral could enter her 3D animation for next year?
Darlene Hadrika dhadrika@mal.ucf.edu
www.youngcomposerschallenge.org
Wednesday, August 5, 2009
Nvidia 3D Vision
Samuel Gateau from Nvidia
I have his slides on my computer. He made some great illustrations of how stereo works.
They discussed the 'how to' of applying stereo to games.
Nvidia 3D vision discover glasses
Nvidia 3D vision – active shutter glasses synchronized with the display
How it works –
3D game data is sent to stereoscopic driver
The driver takes the 3D and renders each scene twice
display 120Hz is 60Hz per eye
NVAPi stereoscopic module – provided to developers
developer.nvidia.com/object/nvapi.com
http://www.nzone.com/object/nzone_re5_downloads.html
Resident evil 5 for nvidia 3D vision
Masaru Ijuuin from Capcom
the look of depth
controlling stereo parameters
parallax budget
how much parallax variation is used in the frame
at 100*screendepth, parallax is 99% of the separation
between 10 to 100 * screendepth praallax vary of only 9%
at screendepth / 3, parallax is 2* separation, out of the screen
parallax is very large (> 2*Separation) and can cause eye strains – don't go over this distance
They had trouble with subtitling – For the next game their solution was to cut the subtitles.
There will be no 3D support on game consoles?
Will Wright
Will Wright was the keynote speaker. He created the SIMS and SPORE.
I overheard a conversation after the speech.
“Hi, I went to this graphics thing and missed Will Wright. What was interesting in his speech?”
The friend couldn't respond.
I thought “Oh boy, did you miss a great speech.”
Here are just a few ideas. His speech was headed “How we perceive things”. The brain figured prominently. He would be a great interview for Neuropolis.
He showed the website kittenwar.com He had uploaded a picture of his kitty that made his kitty look small and cute and another of his kitty that was a nice photo but not as sympathetic. His bad photo but cute kitty was more popular on kittenwar.com
In the entertainment industry, there have been individual silos of production. We are just now approaching the intersection of techniques and imagery and the way content moves around.
We don't have a concept of entertainment design. Is there a unified entertainment theory? There was a mood management theory (Zillman 1988). We watch movies that will improve our mood. http://en.wikipedia.org/wiki/Mood_management_theory The consumer knows what it wants, knows what is called , doesn't know how its made.
I overheard a conversation after the speech.
“Hi, I went to this graphics thing and missed Will Wright. What was interesting in his speech?”
The friend couldn't respond.
I thought “Oh boy, did you miss a great speech.”
Here are just a few ideas. His speech was headed “How we perceive things”. The brain figured prominently. He would be a great interview for Neuropolis.
He showed the website kittenwar.com He had uploaded a picture of his kitty that made his kitty look small and cute and another of his kitty that was a nice photo but not as sympathetic. His bad photo but cute kitty was more popular on kittenwar.com
In the entertainment industry, there have been individual silos of production. We are just now approaching the intersection of techniques and imagery and the way content moves around.
We don't have a concept of entertainment design. Is there a unified entertainment theory? There was a mood management theory (Zillman 1988). We watch movies that will improve our mood. http://en.wikipedia.org/wiki/Mood_management_theory The consumer knows what it wants, knows what is called , doesn't know how its made.
Monday, August 3, 2009
Database, algorithms, PIXAR, Disney and so on
Information Aesthetics - Consideration in Design and Data Analysis
Moritz Stefaner gave this talk on trying to understand the structure of a large data set.
In this case, the algorithm searched for citations of articles in scientific journals. He
wanted to find out how often an article was sighted, what each journal covered as
scientific topics etc.
Imagine if you were in charge of a space allocation study and wanted to know who emails whom most often. Imagine if you were asked to 'tag' or 'keyword' your images and wanted to see the most used or relevant ones. Imagine if you were paranoid and thought this was all too much information.
http://well-formed.eigenfactor.org/map.html
Automatic Coloization of Grayscale Images Using Multiple Images on the Web.
Associate Professor, The University of Tokyo, Takeshi Naemura
This talk described colorizing black and white images using similar color images on the web. This was an automatic process.
As I was watching the demonstration, I thought of the black and white 3D images of neurons
that Simon was working on. Could we use this tool to colorize the image?
He also presented a 3D live video system using lenslets in 2004.
http://www.siggraph.org/s2004/conference/etech/liflet.php?=conference
Moritz Stefaner gave this talk on trying to understand the structure of a large data set.
In this case, the algorithm searched for citations of articles in scientific journals. He
wanted to find out how often an article was sighted, what each journal covered as
scientific topics etc.
Imagine if you were in charge of a space allocation study and wanted to know who emails whom most often. Imagine if you were asked to 'tag' or 'keyword' your images and wanted to see the most used or relevant ones. Imagine if you were paranoid and thought this was all too much information.
http://well-formed.eigenfactor.org/map.html
Automatic Coloization of Grayscale Images Using Multiple Images on the Web.
Associate Professor, The University of Tokyo, Takeshi Naemura
This talk described colorizing black and white images using similar color images on the web. This was an automatic process.
As I was watching the demonstration, I thought of the black and white 3D images of neurons
that Simon was working on. Could we use this tool to colorize the image?
He also presented a 3D live video system using lenslets in 2004.
http://www.siggraph.org/s2004/conference/etech/liflet.php?=conference
Getting There
Travelling to Montreal on Saturday to stay with my Mom only to find that the 13 south bound was closed. 2&1/2 hours later..
5 am Airport madness, the automatic check in didn't work... Another lineup.
Spent a few minutes in Miami airport. My luggage spent overnight travelling to the wrong Hotel. I met up with Jean Detheux. His luggage spent overnight travelling as well... He is doing a live performance as an invited artist at Siggraph.
I haven't seen any evidence of Katrina. The French Quarter wasn't underwater. The people and the food haven't changed either. Mena's and Deanie's were friendly and the food NO good. How y'all doin' today?
http://www.deanies.com/restaurants/french-quarter/
http://www.menaspalace.com/ That's the picture above.
Subscribe to:
Posts (Atom)
