Pixar’s walking teapot is always the best swag of SIGGRAPH. Every year it is a different theme. This year is Mr. Potato Head in honor of Toy Story 3.
Why a teapot? Because it has become the standard model for showing off computer rendering techniques.
The line to get the teapots was ridiculous this year. They give out a batch each of the 3 exhibit days. I stood in line on the first day and they ran out by the time I got close.
The next day, I got in line over an hour early, and I was still several blocks from the start of the line. I did get one, though.
Based on the certificate of authenticity, they give out 1,500/day.
Stereoscopic 3D (S-3D) was big at SIGGRAPH this year and left me thoroughly impressed. For me, it was the biggest news of SIGGRAPH 2009.
What is S-3D? S-3D creates the illusion of depth, like you would see with a View-Master. Instead of a single image, each eye has a unique perspective of the image.
Why is S-3D hot all the sudden?
Digital projectors are common in theaters now (TI’s DLP and Sony’s SXRD)
It is easy to place a filter system in front of digital projectors that allows each eye to see a different image with special glasses.
RealD has most of the market currently. I don’t have any stats, but I believe I heard in the SIGGRAPH panel that over 1,000 theaters are RealD-capable now and they have more than 90% of the market for 3D cinema. To use RealD, theaters need a special silver screen to reflect more light and keep the light polarized. They also need a RealD filter device in front of the projector to polarize frames for the left eye and right eye. The glasses are cheap to produce and can be recycled or replaced.
Dolby 3D uses color filters instead of polarizing light. Each eye can only see certain colors, but amazingly appears as a full color image. Dolby 3D uses a standard white screen. A big advantage is there is no special screen to buy and thus 2D movies are not impacted by a theater’s move to 3D-capable. I’m not sure how a 2D movie looks on RealD’s silver screen, but I’ve been told it does look different than using a standard white screen. The Dolby 3D glasses are not cheap like the RealD glasses, and theaters must be careful not to lose the glasses.
I’ve seen both RealD and Dolby3D and I can’t pick a winner…both look really good.
Movies are obvious candidates for S-3D…but I saw other forms of entertainment that *really* shine in 3D. In the SIGGRAPH panel, I watched a demo reel from 3ality Digital that showed scenes from an NFL game, an NBA game, and a U2 concert. Seeing a live event in 3D is *very* compelling.
Two scenes stood out in my head:
A running back trying to break through a wall of linemen and eventually forced out of bounds and into the 3D camera made the entire audience react as the running back kept getting closer to us. We had the same reaction as if we were standing there preparing to get hit.
Watching the kicker kickoff. Seeing the ball from the opposite end zone is nearly impossible with the sea of people in the background…but in 3D, the ball popped out in front and was very clear, even though it was very small.
I can’t wait for movie theaters to start hosting NFL/NBA games in 3D…I’ll definitely shell out some cash to see that.
I’m a believer in 3D for movie theaters…I’ll definitely go see every movie in 3D that I can.
But what about in your home?
That is getting very interesting very quickly.
JVC had a 46 inch LCD on display at SIGGRAPH that uses the same technique as RealD…and it looked really good. The bad news: You lose half your resolution because each frame contains the left and right images instead of just a single image. That means you *cannot* do 1080p currently in 3D.
Panasonic should have a 3D-capable set by next year. Sony has not made a 3D announcement yet, but was very active in the SIGGRAPH panel…so you can draw your own conclusions.
So you get a Blu-ray player and TV that supports HDMI 1.4…what about 3D content? There is *no* standard for 3D yet…so anything you buy now may be worthless when a standard is decided.
My advice for 3D in the home: wait. It is coming, but it looks like it is still a few years away.
As soon as 3D for the home is standardized…I’m jumping in…I’m hooked on this stuff.
The SIGGRAPH panel left some good websites to follow for information on this topic:
The last section was the Curated Reel, work invited by the Computer Animation Festival Chair. My favs:
ILM showed clips from Star Trek, Terminator Salvation, and Transformers. No link for this, but you can see it on Friday’s at noon if you come for a tour.
Pixar showed its short “Partly Cloudy” (shown before “Up”)
It turned out to be Vanilla Ice! What better way is there to get people excited about your new 3D software than with a washed up white rapper?
It’s fun to goof on Vanilla, but he did put on a good show.
I got *really* close…I had inside info on who was going to be the performer, so I wanted to be in the front row/center so I could get a fist bump from V-Ice.
The club filled with a bunch of smoke, so I knew something was about to happen…
I figured it was going to be the opening act. A guy emerged from the smoke and tried to shake my hand. I motioned to do a fist bump instead…which confused the guy at first and then he played along. I didn’t realize until the smoke cleared that it was actually Vanilla Ice. I totally dissed his handshake offer and instead made him do a fist bump. Hey…that’s how I roll.
My buddy Spaghetti (a.k.a. Trey) took all of these photos. Here he is with Vanilla in the background…
You can see my green shirt…which is covered in water. Vanilla Ice has a large supply of bottled water…not because he is really thirsty, but because he uses it to spray on the crowd. You can see the water in the bottom-left corner of the next pic…
The girl standing next to me was a reserved engineer from AMD. All it took was a look from Vanilla Ice and the next thing I know, she is on stage with a bunch of girls making out. It was like he brought out a pole at a strip club…amazing…
I am still practicing that Vanilla Ice look…so far it is not working.
Also in attendance, Vanilla Ice’s good friend, porn star Ron Jeremy (next to the girl in white)…
It was a fun time (although I learned nothing about Softimage except that they throw great parties). Maybe they already knew they were going to be sold at this point?
As is a tradition at SIGGRAPH, women were difficult to spot. See if you can find any…
I started off the day in a technical papers session entitled “Image Collections & Video” which included 4 papers. This is probably my favorite session at SIGGRAPH…it always has new tech that will probably impact me in the near future.
The first paper was “Factoring Repeated Content Within and Among Images.” This is a new image compression technique that looks for repeating patterns in an image (an “epitome”) and only stores the pattern once. A “transform map” reconstructs the original image from epitomes.
Results of this technique compared to JPEG 2000 show significant improvements in image quality using the same amount of data (although images also had significant repetition). Also, this technique is useful for minimizing RAM utilization because you can render parts of an image without completely recreating the full image (see “Requires XXX for rendering” in photo above).
Photo Tourism takes a large collection of photos (from Flickr, for example) of a subject and combines them to allow 3D interaction.
The Finding Paths paper presents more natural ways to interact with these photos.
The best navigation depends on the types of photos. If photos are taken from many locations of the same subject, then “orbit” is useful. If photos are taken from mostly a single location, of various subjects, then “panoramas” are used.
Path planning used for the transition between photos takes into account routes where photos exist. Thus, the transition from outside a building to inside doesn’t go through a wall, but instead follows the path where people took photos via walkways.
Appearance stabilization fixed issues with different colors of photos of the same subject.
Check out the video above to see this in action…it is amazing.
Last year at SIGGRAPH, a paper presented a technique called “seam carving” that retargets the dimensions of an image without distortion. The least “important” parts of the image were removed to shrink the image, or duplicated to grow the image.
With this technique, you always have video that fits your screen without black bars (letterbox) or distortion (scaling). I prefer letterbox on my TV, but on a small screen (like a cell phone), I can see where this technique would be useful.
This technique captures a 2D texture representation of an object in a video (an “Unwrap Mosaic.”
Once you have the texture, you can update the texture to add/remove/change features. The modified unwrap mosaic is then reapplied to the video so that the changes look like they are part of object. Check out the video clip above to see it in action.
These types of operations typically require recreating a model in 3D to make the changes and then reintroducing the changes back to the video synchronized with the original object. Unwrap Mosaic greatly simplifies this process because all the work by the user is in 2D using a image editor.
Coming to your favorite compositing package soon (hopefully!).
The next papers session I went to was called “Parallelism.” The main paper I wanted to hear was about Intel’s new GPU called “Larrabee.” The paper is here.
There is a lot of speculation about what Larrabee is and isn’t and how it may or may not change the graphics hardware industry.
From what I understand, Larrabee is a graphics card. But unlike all other graphics cards, its GPU is based off an x86 CPU.
The advantage of Larrabee: Software that works on a PC can be compiled to work on Larrabee unchanged. For GPU’s (like Nvidia’s GeForce series), PC software must be rewritten to take advantage of the GPU or avoid its limitations.
As you can see from the chart above, traditional GPU’s (DX8-DX10) let you program 3 stages of the graphics pipeline (vertex shading, geometry shading, and pixel shading).
Larrabee, on the other hand, allows full programmability of the graphics pipeline or a completely different graphics pipeline.
Of course, everything sounds great in theory. It will be interesting to see how perception changes once we see some real hardware.
Here’s my prediction for Larrabee…
GPU’s will be faster than Larrabee for graphics that fit in the standard graphics pipeline
Larrabee will be faster than GPU’s for tasks that don’t fit into the standard graphics pipeline (like ray tracing)
Larrabee will be faster than a CPU for tasks that involve lots of processors
Larrabee fits nicely in between a GPU at one extreme and a multi-core CPU on the other.
Larrabee will be most important to people that use racks of PC’s now to do their work
Larry Seiler from Intel happened to sit at my table for lunch one day at SIGGRAPH. He presented the Larrabee paper at SIGGRAPH (pictured on the left).
I asked him where the name “Larrabee” came from.
He said Larrabee was from the TV show “Get Smart.” The chief’s assistant, Larrabee, didn’t get much respect. The Larrabee project has been in the works at Intel for a while and didn’t get much respect initially.
I *really* like the Fast-Forward session because it quickly gives you an idea of what papers will be interesting and helps you plan out your schedule.
But, the Autodesk User Group is important for learning about the new things that Maya and 3ds Max can do.
What to do? I ended up arriving early at SIGGRAPH and picked up the Full Conference DVD-ROM the day before the conference started. I spent the day studying which papers I should visit. Problem solved! No need to go to the Fast-Forward session.
Industrial Light & Magic on Iron Man and Indiana Jones and the Kingdom of the Crystal Skull
Lucasfilm Animation on The Clone Wars
LucasArts on The Force Unleashed
DreamWorks Animation on Kung-Fu Panda
Blizzard Entertainment on StarCraft II and Diablo III
The Third Floor on previsualization
Sony Pictures Imageworks’ Rob Engle presenting a sampling of the SIGGRAPH® stereoscopic 3D screening
Demonstrations of Autodesk® 3ds Max®, Maya®, MotionBuilder®, Lustre®, Mudbox™, Toxik™ software, and more
An after party featuring the interactive art of Massive Black and ConceptArt.Org.
One of the more interesting observations about the event was how little information there was about 3ds Max. The host of the User Group essentially glossed over the new release of 3ds Max 2009 because "it had been covered elsewhere." The focus was definitely on Maya.
It has been 3 years since Autodesk acquired Maya. Autodesk already had 3ds Max. So what happens when the #1 and #2 3D software packages are owned by the same company. Kill Max? Kill Maya? Keep them both? Merge them? It's an interesting conversation topic and everybody has their theories on the fate of Max and Maya.
I think the SIGGRAPH User Group made it pretty clear what Autodesk's direction is: Maya *is* the 3D package for the film industry. Check out the Autodesk User Group t-shirt they gave out...it is all about Maya (Maya dragon logo, 10 is for the 10th anniversary of Maya/10th version of Maya):
Autodesk announced they acquired REALVIZ (image based modeling, motion capture, panoramic photography). REALVIZ demonstrated their motion capture software called Movimento. They showed how the software could extract motion capture data from 4 handheld cameras on a set (not using a motion capture stage).
The *coolest* part of the user group was a screening of several movies in 3D.
We were all given some Dolby 3D glasses. These glasses were different than any 3D glasses I used in the past. The lenses appeared to be clear (not red/cyan anaglyph). The lens were not polarized (I put two lenses on top of each other and rotated them...if they were polarized and not aligned you won't be able to see through them). They were not LCD shutter glasses.
You can see from the pictures above that white light gets either a red or green tinge to it. I also noticed I could get *either* lens to do red or green depending on the angle I looked through the lens. When watching the film, you see perfect color.
So what are they? I found this article that explains how Dolby 3D uses color filters for each eye.
I was *amazed* by how good the 3D glasses work. It is *much* better than anything I've ever seen before. Do yourself a favor and check it out. Here is a list of theaters that support Dolby 3D.
The coolest 3D demo we saw: Star Wars Episode 4 (the original). The film was converted to 3D by In-Three. They showed the opening sequence with the Star Destroyer chasing the Princess Leia's ship.
Disney's Glago's Guest. A short that will proceed the upcoming Bolt (Nov 21st). Definitely check it out...it is well done.
The User Group started at 6:30 and ended after 10pm. That's almost 4 hours! I loved it...but I wished it was half as long...especially since we don't have time to grab dinner between the last class at SIGGRAPH and running over to the User Group. And there was no food at the User Group.
Next up was the party. It was also in the Shrine Auditorium, but in another room that I didn't know existed.
I was *starving*...and the party had plenty of great food. Everybody was running to the food as if they had been stuck on a desert island for several years.
Massive Black gave the party a unique look with various stations with live models painted by artists. Huge screens let you watch the art as it was created.
The course covered data structures for points, lines, regions, bounding box hierarchies, rectangles and surfaces/volumes.
Hanan's website has some nice Java apps that help you visualize how the data is organized. I couldn't run them in IE8, but they worked in Firefox.
I plan on referencing this information the next time I need to pick a 3D data structure. Hanan has a book that covers these topics in more detail that I'll probably pick up.
Ed Catmull, president of Disney Animation and Pixar, spoke candidly about his experiences running Pixar and Disney.
The talk was fascinating. I wish I had the whole thing on video. I jotted down some notes on what Ed said.
Why Pixar is successful...
What makes Pixar so successful is that Pixar focuses on the story.
BUT...*everybody* says they focus on the story. So saying "we focus on the story" is really meaningless.
If focusing on the story isn't what makes Pixar successful...what is?
Ed came up with two more possibilities: good people and good ideas.
Which one is more important?
Good people.
Why?
Good Idea + Bad People = Bad Implementation
Bad Idea + Good People = Fix Idea or replace idea with a better idea
An example of this...
Pixar had their "A" team work on Toy Story and then Bug's Life. "A" team worked well together and were very successful. "B" team started working on Toy Story 2. "B" team didn't have the same "magic" as "A" team. Toy Story 2 wasn't working and was in trouble. When "A" team finished with Bug's Life, they took over Toy Story 2 and started over (new script) with only 8 months until the movie was scheduled to finish. The movie was finished on time and was a huge success.
Every 1st version sucks. The next iteration sucks less. Keep iterating until it is actually good.
The software is used internally by Pixar and sold externally as a product. Pixar once tried to keep new features internal before releasing them to customers so Pixar would have a competitive advantage. Net result: customers stopped trusting Pixar and lost faith in the product. Releasing all features fostered a better relationship with customers and meant new features were vetted by a larger audience which improved the product.
Peer reviews...
At Pixar, people have to present their work to an audience on a regular basis in an unpolished state. Because of the frequency, people quickly get over the embarrassment of critical reviews. Also, the whole group learns from the criticism.
Desired Qualities in employees...
Have to be a filmmaker
Have to be honest
Have to be able to criticize people you may idolize. Once you are successful, people give you a "free pass" because they assume you know what you are doing. Everybody needs criticism to perform their best work.
Crises...
You can't be successful by avoiding crises...because they always happen. *How* you respond is what determines success.
This class dealt with how to interactively (defined as >= 10 fps) manipulate large models. The Boeing 777 data set was used for several of the examples, which has 470 million triangles and over is over 30 GB on disk.
Performance approximately logarithmic to number of triangles
Disadvantages
No hardware support
Slow
As models get more complex, ray tracing becomes more viable compared to rasterization.
ReduceM is a data structure for geometry designed for ray tracing. It minimizes the footprint for viewing large models at interactive rates.
Far voxels are a way to represent lots of data by a simple cube. Each face of the cube is a texture of what the model would look like from that vantage point.
I'm always curious about the attendance and exhibitor numbers for SIGGRAPH as a barometer for computer graphics.
Here is what I could find from this year's SIGGRAPH dating back to 1996...
Year
Location
Attendance
Exhibitors
1996
New Orleans
28500
321
1997
Los Angeles
48700
359
1998
Orlando
32210
327
1999
Los Angeles
42690
337
2000
New Orleans
25986
316
2001
Los Angeles
34024
303
2002
San Antonio
17000
225
2003
San Diego
25000
240
2004
Los Angeles
27825
229
2005
Los Angeles
29122
250
2006
Boston
19764
230
2007
San Diego
24043
230
2008
Los Angeles
28400
237
I got these numbers from financial reports or the home page for that year's SIGGRAPH (here is SIGGRAPH 99). For 2008, I took the attendance from the final presentation of SIGGRAPH and I counted the number of exhibitors in the show floor guide booklet.
Here are a couple of charts to look at these numbers graphically...
SIGGRAPH had its largest attendance ever in 1997. Then there was a downward spiral until 2003. Since then, SIGGRAPH has been fairly stable.
Here is a video of the coolest thing I've ever seen at SIGGRAPH. It looks like liquid that is "alive"...you have to see it to believe it. People sat and watched in amazement seeing liquid do things your brain says it can't do.
I'm looking forward to hearing about DirectX 11 and Intel's new Larrabee graphics card. Both should have presentations at SIGGRAPH.
There is always way more stuff to see/do than you have time for, but that means there is never a dull moment.
Who's going? Let me know and we can meet up.
I plan on posting to twitter while I'm at SIGGRAPH. If you are going, you should get a twitter account and do the same. It's easy and you can read/write updates via your mobile phone. Let me know what your account name is and I'll track your updates.
I'm taking my new laptop with me and plan on posting blog updates at night.
I registered for SIGGRAPH 2008 this morning. This year it will be in Los Angeles August 11-15.
Registration opened yesterday. Register before July 4th for the best rates. Become a SIGGRAPH member and save $50.
One notable difference this year: no Electronic Theater viewing day selection in the registration. I actually can't find any mention of the ET. The closest was this quote about Computer Animation Festival changes:
Computer Animation Festival For SIGGRAPH 2008, the festival has adopted a new format. Each day of the conference, it presents competition screenings, showcase screenings, and panel discussions with filmmakers, instructors, and artists involved in the creative process. The traditional Animation Theaters will not be available for SIGGRAPH 2008.
I hope this doesn't signal an end to the ET...it is one of my favorite parts of SIGGRAPH.
Sounds like several things have changed (or at least changed names).
Randy Pausch is a professor at Carnegie Mellon University. He is losing his battle with pancreatic cancer and is expected to live for a few more months.
CMU has a lecture series entitled, "The Last Lecture." It is supposed to be your chance to talk about things that matter deeply to you. For Randy, it really may be his last lecture.
I first watched his 10 minute reprise of his lecture as a email forward...
After watching it, I wanted to see more. Here is the full video (76 min) from his original talk, "Really Achieving Your Childhood Dreams..."
The second video is more sincere and drawn out and I recommend viewing it.
Also, I was surprised to see a co-worker in the second video, Tommy Burnette. Tommy got a nice shout out from Randy for following his dreams to work on Star Wars.
I've seen Randy's work at SIGGRAPH's Emerging Technologies. He's worked on a VR project called "Alice."
In his "last lecture," Randy announced one of his childhood dreams of "being Captain Kirk"...since then, J.J. Abrams gave him a role with a line in the upcoming Star Trek movie.
The first was course 22: LucasArts and ILM: A Case Study in Film and Game Convergence. I'm directly working on this project, so it was fun to hear about it from a 3rd person perspective. Probably the most interesting for me was hearing about the history of how we got to where we are today.
The next presentation was called "Spor(T)." This was hosted by EA. Most of the presentations were about tools used to create the upcoming "Spore" game. The final presentation was about real-time motion for sports...thus, you get unified title "Spor(T)."
Spore is interesting because of the procedural aspect...much of the animation and art and design is based on algorithms instead of static pieces of art. This means you can get a large variety of results that the game designers may never have seen or planned for.
They showed off a planet editor and a creature editor that let you quickly build very detailed models with very little input. They also showed how they used the Halton sequence to distribute items (like trees) in a random/natural fashion, yet very reproducible (so you can return to the same place after you turn your console off).
The sport part of the talk dealt with how to animate a soccer player realistically with most of the motion computer generated (instead of motion captured). They use a modeling tool called ANT. ANT allows you to tweak the look of the running style (how much do you lean forward, how much do you swing your arms, etc.) in real time. ANT lets you say you want a player to track a ball with his eyes. As the ball moves out of the players view, the player will automatically rotate his head left or right. Once the head has turned as far as it can, it automatically turns to the other side to continue tracking. All this happens automagically...no animation by humans required.
I found this on youtube. It's an overview of some of the papers presented at SIGGRAPH (voiced by Jim Blinn). It's a nice summary of what was presented this year:
Wednesday I spent most of the day walking around the show floor. I was *really* disappointed...there were few interesting things to see. In the past I could have spent days learning about new products and services...this year I didn't see anything I hadn't seen before.
I use the show floor as an indicator for who is gaining in importance in 3D graphics and who is falling off the radar. The bigger the booth, the more important you are.
Based on the floor plan above, the biggest booths were:
It is interesting that 3 of the top 5 booths were there to recruit people, not sell products. That is definitely a change from previous years (when the floor was more interesting).
Also interesting was who was not present, but usually is...
Apple
Adobe
Intel
Microsoft
Dell
PDI/Dreamworks
From that list, I was most surprised by Adobe and Apple. Normally Apple has a huge booth promoting Final Cut Pro, Shake or Motion. Same goes for Adobe. They recently released CS3, and I would have figured they would have been showing it off at SIGGRAPH. PDI/Dreamworks was at SIGGRAPH for the job fair, but not on the show floor. Here is a list of all the exhibitors.
We had a large booth at SIGGRAPH because we are hiring a lot of people (see here, here, and here). We also had some really good presentations. I sat in on the John Knoll/Pirates 3 presentation that was packed...
Lucasfilm hosted a party in the Gaslamp District that evening. It was pretty easy to find...just look for the place with a bunch of Stormtroopers out front...
Two girls I met at the Autodesk party (Dorina and Laura) won a record single of "The Imperial March (Darth Vader's Theme)" during a contest at the Lucasfilm party.
The girls wanted me to autograph their record...which is ridiculous because I haven't done anything. They still wanted my autograph so I signed it...but I told them I could get some guys that actually worked on Transformers to sign it. I saw my bball buddy Nick Woo at the party and asked him to autograph the record...
I also saw Jeff White there and he signed it and then took a picture of me with the girls...
After the Lucasfilm party, I headed to the SIGGRAPH reception at the Marina Park to get my one free drink. After the drink, we went back to the Gaslamp District to hit the Softimage party.
The Softimage party was at the House of Blues. Initially, the line to get in was over 2 blocks long and it wasn't moving. We went to another bar and hung out until I got word from The Intern (Noah) that the line had disappeared and they were letting people in.
The HoB is a cool venue. They had a woman DJ spinning some groovy industrial dance tunes with go-go dancers on either side of the stage. Other than those women on stage, it was SSS (Standard SIGGRAPH Sausage-factory)...
I started the day with "Transformers: Giant Frickin' Robots." Several ILM guys were on the panel (Scott Farrar, Russell Earl, Scott Benza, Jeff White, Richard Bluff) to talk about how they did the effects in Transformers. It was packed...I had to go to an overflow room that had audio of the panel and video from the effects sequences...but no video of the panel. The best part for me was seeing the reference video of the actual martial artists fighting and then see the equivalent sequence preformed by the robots.
I stopped by the show floor to see what Nvidia was pimpin'. The most interesting was Mental Images' "Mental Mill." Mental Mill is a shader development environment where you graphically connect the output of nodes to the input of other nodes to get a shader. It creates MetaSL shaders that can be converted into C++, Cg, GLSL, and HLSL or other shading languages using the developer kit. The full version (cost...no idea) has a shader debugger! There is a free version (Artist Edition) that is shipping with Nvidia's FX Composer, but it does not include the debugger.
FX Composer has a shader debugger planned for version 2.4 (they are currently on version 2.0). Nvidia has no plans to support GLSL (only Cg and HLSL) in FX Composer...bad news for GLSL (from my view, GLSL is really struggling).
HP was showing laptops that have 3 button touch pads and 3 buttons for the pointing stick. I don't know why laptops have skimped on the 3rd button for so long, but I'm glad to see this change.
I went to the Highlights From SCA (Symposium on Computer Animation) sketch and also the Highlights From UIST (User Interface Software and Technology). Each sketch was like a "best of" summary of their last conference.
UIST (pronounced "wist") had the most interesting presentation I saw that day. The paper is "Summarizing Personal Web Browsing Sessions." It is like a TiVo for web researching.
An example they give is looking for a hotel. You can use hotels.com to find out how much a hotel is, availability and its address. You can provide the address to Google maps to see the hotel's location. You can provide the hotel name to yelp.com to get a review and a rating. That requires three different web sites for a single hotel listing.
Using the software presented in the paper, all this information can be placed on a single "card" so that you can focus only on the information you care about. Once a card is setup, just click on other hotels you are interested in on the hotels.com page and their cards are immediately filled in with maps from Google maps and reviews and ratings from yelp as well as the hotel price, availability and address from hotels.com.
It is a simple concept, but a *huge* time saver! I could definitely use this. The videos do a great job of showing this in action. I tried playing these in the browser, but audio would not work. When I saved them to my hard drive (right-click, Save Target As...) and played them, I got audio. Check 'em out:
Just like TiVo lets you watch what you want and skip what you don't (i.e. commercials), this project shows you the information you want, and bypasses the ads. I look forward to trying this out!
The pre-show had 3 graphics "celebrities" play old vector-based coin-op classics on the big screen. I didn't catch the name of the first guy playing Asteroids or the last guy playing Star Wars. The middle guy was Ken Perlin playing Tempest.
Most of the good clips are not available online, but a few are...
Travelers: Snowball
HP Hands "Paulo Coelho"
En Tus Brazos
Sears Tools "Arboretum"
Burning Safari
U2 and Green Day "The Saints Are Coming"
Happiness Factory
Vigorsol: The Legend
Portal
The Itch
Equilibrio
Raymond
ILM showed a clip on how some of the effects in Transformers and Pirates 3 were done. We show this clip on Friday's at noon as part of our "Friends and Family" demo reel. So if you are coming to visit...make sure you can stop by for lunch on Friday.
I walked past the PDI/Dreamworks SIGGRAPH party at 9pm and there was a line two blocks long to get in. The party was at Stingaree in San Diego's Gaslamp Quarter. I was meeting up with some friends for dinner and then we were going to drop by the PDI/Dreamworks party.
We arrived at Stingaree around 10pm. There was no line, but they were at capacity. The party was on the roof and they were not letting anybody in until people left the party. Surprisingly, not very many people were leaving (go figure, free booze and sushi).
So myself and several co-workers were the first in line to get let in once other people left the party.
As the line formed behind us, it was fun watching people walk up to the bouncer and say the equivalent of "I'm with the DJ," only to be shot down and sent back to the line.
Then a blond girl from HP showed up.
The HP girl spoke with the Dreamworks recruiting manger about getting into the party without waiting in line. She said HP is a partner of Dreamworks and that means she should not have to wait in line.
I listened as the recruiting manager explained that nobody is more important than anyone else and that the reason for this party is to get potential candidates inside. Without the candidates, there is no party.
The HP girl refused to accept this...she was from HP and she should be let in NOW!
She was denied and sent back in line.
The recruiting manager apologized to us all about the wait and said she really wanted us all inside.
Then a second blond girl from HP arrived. She looked like an older woman that was trying to relive her glorious clubbing days. She had the same "we don't wait in lines"-attitude about her, except she was more hard-core about it. I'll call her PHP (Past Her Prime).
The first blond explained why they were waiting in line to PHP. PHP was outraged and brought the first blond with her to demand they be let in NOW!
The bouncer explained the situation again, but PHP wanted to talk to a manager. A rep from Dreamworks began to talk the blond girls.
Now mind you, all this is happening outside with a line of potential Dreamworks candidates listening to every word.
Then I heard the magically line that made me start recording...
"If Dreamworks values its relationship with HP..."
Oh no you didn't! You did not just claim that the corporate relationship between Dreamworks and Hewlett-Packard is contingent on you getting into this party without waiting in line!
The conversation went on for about 10 minutes before she was finally let in. It really didn't matter...enough people had left that they began letting us in anyway.
I don't know who this girl is...but in my experience, people that claim they are really important...aren't.
So the net result: There is a line of industry people that got to watch HP tarnish its image. Well done!
The video doesn't capture the best parts of the conversation, but it does give you a feel for what happened. You can see the two blond girls talking to a Dreamworks representative.
At the user group meeting, we got some nice long sleeve (great for SF!) black shirts with a Maya logo on the front, Max logo on one sleeve, and a Combustion logo (I think...it looks like a flame).
One of the speakers was the head of marketing for Chrysler. She brought along a Dodge Demon that was unveiled on stage like this was an auto show. Very good-looking car.
Max and Maya have now use years instead of versions numbers. The latest offering is Maya 2008 and Max 2008.
Max gains lighting/shadows in the interactive 3D viewer. No need to re-render when placing lights...very cool.
Maya gains some rigging features that made the crowd very happy. A new widget for controlling the view is in the upper right corner. It lets you switch views (e.g. left to 3D view), or rotate your view in screen space (turn upside-down, for example).
ILM was the final presentation with how we did Transformers using Maya. Jeff White was the presenter.
After the User Group, everybody (it seemed like a few thousand people) headed to the USS Midway aircraft carrier for the party.
We started out on the flight deck where all the airplanes are kept. They had a fireworks show right in front of the USS Midway. After the fireworks, we went downstairs to an area that had flight simulators we could ride (I passed...I've already done enough of that!).
It was an open bar until 11:30pm (we got there around 10). They also had free pizza and desserts.
There was a DJ, dance floor, and go-go dancers...
It was quite a sausage factory, except for this one area with several cute girls. I walked over and asked one girl if she'd like to dance. We danced a little and I asked her if she uses Maya or Max. She didn't know what I was talking about. I asked her what she does, and she said "actress." I asked if she had done anything I would know, and she said "yes." I said, "Let me look you up on IMDB on my cell phone"...and she was there.
Her name is Zoe Quist. She played the memorable part of "crowd onlooker" in War of the Worlds. I told her my company did the effects in that movie. She said, "What company is that?" I told her "ILM"...to which I was greeted with a blank stare...she had no idea who ILM was. Oh well...at least she was easy on the eyes.
It turns out she was hired by CafeFX (did the effects in Pan's Labyrinth, co-sponsor of the party) along with almost every other cute girl on the dance floor to make it less than 100% computer dorks on the dance floor moving non-rhythmically to the music.
I started the day with a papers session called "Image Analysis & Enhancement." The first talk ("Image Deblurring With Blurred/Noisy Image Pairs") was about how to use a blurry long exposure image with a noisy short exposure image to get a sharp, clear picture with great color reproduction. My camera has image stabilization...it is not blurry, but it does contain a lot of noise. This technique would give you great pictures even when you hands are shaking or in low light situations. Very Impressive...I want this in my next camera!
Two other papers in this session were related. The first, "Photo Clip Art", showed how to add objects (like people or cars) to a photo by using MIT's LabelMe photo database to find candidates that matched the correct size and color properties of the original photo. The idea is to mimic the clip art added to PowerPoint a presentation, but for photos.
The next paper session I went to was called "Image Slicing and Stretching." The first paper ("Soft Scissors: An Interactive Tool for Realtime High Quality Matting") showed how you could just quickly trace the outline of a cat, for instance, and all the hairs that blend into the background were selected correctly.
The next paper was called "Seam Carving for Content-Aware Image Resizing." Check out the video to see it in action. Basically, they just get rid of the least important parts of a picture until it is shrunk to the desired resolution. To increase resolution, the least important parts of the picture are duplicated. The results are surprisingly good and fast. This only works for images, not video yet. Could this be the end of black bars and stretched/compressed people on TV's?
"Image Vectorization using Optimized Gradient Meshes" is a paper that deals with converting a photo into a vector format (like what is used by Adobe Illustrator or Flash animations). I have dealt with shapes as vectors, but never photos. With Windows Vista supporting vectors in the UI (to support arbitrary dpi settings) and the rising popularity of flash, storing information in vectors will become more important.
Later, I went to the sketch "Highlights from I3D". One paper presented "Tile Trees"...an alternative to normal texture mapping. The advantage is that you don't have UV's to adjust to fit a flat texture on a surface. It also does not waste any space of a texture map (like in this texture atlas).
I started the day in Course 3: Sketch-based Interfaces. This course was about using sketching (i.e. using a pen interface) as input to a computer.
This technique throws out the idea of using the ubiquitous WIMP (windows, icons, menus, point and click) interface in favor of using a more natural human interface...drawing with a pen. The computer does more work so that the human can work the way he is accustom to working.
They showed an example of a 3d modeling package. It simplified the user interface by providing the user with only a single 2d view. The computer assumed any drawing was meant to be on top of another object on the screen. So if you drew a 2d rug, the computer would assume it was meant to lie on the floor, and would create a 3d rug for you. What if you wanted the rug to float in the air? You simply draw a shadow in order to position the rug in the air...pretty slick!
This course really makes you rethink how you should write computer software. They used some techniques that were clearly easier for people to learn, but were nothing like the software we use today.
They listed a few SDK's for getting started with sketch-based interfaces:
Tablet PC SDK
C#
Ink as a first class object
Somewhat limiting
Window Vista's Windows Presentation Foundation
Vista-only
Doesn't have all that Tablet PC SDK ahs
SATIN
Java
Uses Rubine's Recognizer (Rubine 1991, an oft-cited paper)
I'm tempted to get a Tablet PC just so I can play with this stuff.
Next, I headed to the Emerging Technology area. I wanted to see Microsoft Surface, which I heard would be on display there.
There was a large group of people around the Microsoft Surface area. I waited for a while until I could get close enough to try it out. They didn't show anything new...everything they showed I had seen in their promo video.
I did get to play with the video puzzle that used real glass squares that the Surface would project portions of a movie on. Your goal was to reconstruct the movie (some of the squares were upside down, which caused the movie to be played inverted).
I asked about the way Surface could ID objects played on it. It uses circles of varying size and position as a bar code. I asked why they didn't just support bar codes and the guy said that the camera couldn't make out the details of a bar code.
In one example, you place a credit card on Surface and drag items from your bill to your credit card to pay. For this to work, the credit card needs to have the special ID circles on it. I asked if I wanted to use my existing credit card, could I make my own dots. He said there is nothing special about the dots, so you could print your own pattern on an adhesive and add it to anything.
The One Laptop Per Child (OLPC) was on display and I got to play with it a bit. It is very different than typical laptop...it would take a little getting used to. I didn't play with it enough to get a feel for where this project is going.
They had some e-ink on display. It looks just like paper...but it is a video screen. It looks perfect from any angle and the text is very sharp and easy to read with the high contrast. They had an example of a display that could be rolled up like a scroll...try that with your LCD!
Later, I went to the Fast Forward session to get a preview of the upcoming papers. It's a great way to figure out which papers you should check out.
I checked into a hostel today (my first). It certainly is a bit different than a normal hotel. So far...I like it. It works just fine for me. I don't think any of my former girlfriends would have approved of this place. People are very friendly and want to talk to you. I'm in the middle of the Gaslamp Quarter, so I'm surrounded by restaurants and night life...and I'm only 4 blocks from the convention center.
Here are some pics of my room. It was $30/person per night and my room is for 2 people. So I paid $60/night to get a private room.
My main concern was the bathroom...I wasn't looking forward to sharing a bathroom and a shower. It looks like it won't be too bad after all...here are some pics of the bathroom...
The outside (see the first pic) has a painting of a guy trying to commit suicide...not sure how that fits into staying at a hostel, but I'll let you know after I've been here a week.
Tomorrow I leave for SIGGRAPH 2007 in San Diego. I just setup my laptop with Windows Live Writer, so I should be able to post from SIGGRAPH. Should be fun! SIGGRAPH starts on Sunday.
I just registered for SIGGRAPH 2007. This year it is August 5-9 in San Diego. Online registration is here. Register by June 29th for the cheapest rates.
This will be my 11th SIGGRAPH! If you have any interest in computer graphics...you have to see SIGGRAPH.
I decided to try staying in a hostel (my first). I have a private room for $60/night just 4 blocks from the convention center.
Some people were talking about driving down. It is 500 miles from San Francisco to San Diego. I got my airline tickets for $150, which is about what I would pay in gas to drive to San Diego and back. When you consider driving down will take about 8 hours versus a 1.5 hour plane trip...it is an easy decision.
I got the Full Conference ticket for $800. I am going to the Electronic Theater on Monday night.
TIP: Always do the "Credential Mailing" option ($15)...it is no fun to wait in block long lines with everybody else trying to get your badge so you can go to your first class. Let the badge come to you!
Here are some photo I took while I was at SIGGRAPH Boston.
The photo above is from the SIGGRAPH Teapot Exhibit. I have the black teapot (SIGGRAPH 2006), gold teapot (SIGGRAPH 2005), and blue teapot (SIGGRAPH 2004).
While I was waiting to get my Walking Teapotswag from Pixar, I saw a computer graphics living legend....Edwin Catmull. Ed is the president of Pixar. He is responsible for several important innovations in computer graphics:
An interesting bit of trivia about Ed...he created the first 3D computer generated graphics in a film. A digitized version of Ed's left hand is used in the 1976 film Futureworld.
I wrote a bit about the ET pre-show previously. I copied it below from Day 2 to save you the trouble of clicking on it...I know...soooo considerate. Here is the Etch-A-Sketch in action. Here is how we erase a giant Etch-A-Sketch.
My buddy John Verostek joined me for the Electronic Theater tonight. We were handed a paddle with a green side and a red side when we entered the theater. When you walked to your seat, you could see on the screen the layout of the theater with green and red lights representing every person holding up their paddle with the green/red side facing the screen. You could find your dot on the screen and flip your paddle to make your light turn red or green. We did some fun interactive games with this gimmick. The best was the giant Etch-A-Sketch projected on the screen. The audience was divided in half. One side controlled up/down by displaying red/green. The other half controlled left/right by displaying red/green. If the audience was split 50/50, then nothing would happen. Mostly green would move up, mostly red would move down. Same goes for the audience with the left/right controls. We were given the task to trace the outline of some shapes: a square, a triangle, a circle, and (of course) a teapot. And how do you clear an enormous Etch-A-Sketch? You shake it, of course! We all shook our paddles in unison and the screen would clear. It was a lot of fun...even though they had some problems registering our paddle information in the section I was sitting in.
The first game we played was bouncing a beach ball around. When the ball came to your area, you were supposed to flip the color of the paddle. Here is the video of it in action. We were still learning how it works at this point...so I think you'll only see one bounce and lots of "buzzing" (meaning you didn't bounce it in time).
Another game was pong with a dog as a ball. The goal was to hit the dog past the other teams "fence" so it would fight their cats. I did this game at a previous SIGGRAPH...I want to say New Orleans in 2000. Here is the video.
This video is of a video wall that is made up of small channels from satellite TV. The camera turns whatever it sees into a mosaic of television channels on the video wall.
These next two are reprints from Day 2, but now I have the video.
This video is from a black room littered with thousands of receipts in piles all over the floor. Each receipt was a few lines of text from a personal ad (like, "I'm looking for a girl that...". Above, in the ceiling, is a printer that prints out a new phrase from a personal ad on the internet and lets it float to the ground every 12 seconds.
Another art piece used the same "magical oil" that was demonstrated at a previous SIGGRAPH. A metal spiral poked out of the oil. The oil would jump out of the pan and cover the spiral and create wild spiky effects...like it was alive...you'll just have to see the video...amazing. Here is the video.
This video is from a game where you chase around "invisible" goblins. The only way to find these goblins is to shine a flashlight on them and follow their shadows. Once you track a goblin down, you use a vacuum cleaner to remove it (of course).
As silly as this sounds, it is very well done. The flashlight takes a computer generated 3D goblin and figures out how to orient it and then removes light from the flashlight so that it creates a shadow that looks like it belongs to a little goblin. The flashlight can be oriented in anyway, and so the shadow has to be constantly recalculated.
When you watch this game, you forget about all the technical issues and just see shadows of goblins running around. Well done! Here's the SIGGRAPH writeup.
This video shows a table with a blueprint of a house on it. You can grab a then LCD display and slide it over the blueprint to see the 3D version of the house. Unlike the previous post, you can rotate the screen and see cross sections from any angle. Here is the SIGGRAPH link.
This video shows a table that has different fruit projected on it. You can grab one of the "tiny displays" and slide it over the fruit and see x-ray slices of the fruit. In another example, the tiny displays had pictures of animated people. If you slid one character near another, it would jump off one display to join the other character where they would bow to each other (I'm guessing this is of Japanese origin). The official account is here.
This video is of a tank of water with computer generated fish projected in the tank. When you touch the water, the fish react. In the video, I think I moved too fast or confused the fish, because they don't seem to be running from me poking the water. There weren't any directions...so maybe I wasn't supposed to poke the water.
This video is from another water exhibit in the same area. A fountain of water creates a water "shell". The shell shrinks as you put your hand closer and then expands when nobody is touching it.
This one messes with your head. It is called "Morphovision." You see a model house in a glass box. The house starts to bend and warp...which makes you think you are looking at a video. But you are not! You can walk around the house and see it from any angle. Eventually, the trick is revealed...the house is spinning at a high rate of speed with a special strobe light that lights up different parts of the house so that it doesn't appear to be spinning. Here is my video clip of the demo...try to ignore the jibber-jabber from the boy geniuses. If you listen very carefully, you may be able to hear my eyes roll back into my head while I was filming.
This was one of the more impressive exhibits at SIGGRAPH. It was a very simple display, but it was 3D and it floated in the air. You could walk around the display and see it from any angle. In the video clip, you can hear the buzz from the electricity...it sounded like at any moment lightning bolts were going to shoot out and fry us all, a la Raiders of the Lost Ark.
Here's a link to more information about how this "True 3D Display Using Laser Plasma in the Air" really works from the people that did it.
Here are a couple of panoramic views of the show floor. The top pic is the view from the main entrance. The bottom is taken from the opposite side. Click on the photos to see all the detail (6k x 1.5k).
Google had a booth for the first time. They were showing off Google Earth and SketchUp. Microsoft had a tiny booth (big enough for 2 computer displays) to let people try out Vista and talk to their engineers. Intel, Apple and Dell were noticeably absent. Alias was sort of missing...they were bought by Autodesk. I would have thought Autodesk would have doubled the size of its booth and had dedicated stages for Maya and Max. Instead, they had a single large booth...about the same size as last year *before* they bought Alias. They had one stage that alternated between Max and Maya demos. The booth focused on the name "Autodesk"...no Maya or Max logos to be found. In fact, if you didn't know that Autodesk owns Max or Maya, you could easily have missed them at the show...which *never* would have happened in the past.
I made these these panoramic photos with my Sony Cyber-shot DSC-T33. I took three overlapping photos and fed them to image stitching software. When you make these, you should err on the side of using more overlap...like a third of the photo should be shared with the previous photo. That helps the stitching software line up the photos correctly. I used both Microsoft's Digital Imaging Suite 2006 and Adobe's Photoshop Elements 4.0. Microsoft's gives better results with almost no user interaction. Adobe's gives you more control, so you can line up pictures that Microsoft refuses to stitch together. For future photos, I'll probably use Microsoft's first, and if I have problems then use Adobe's.
The last day! The main thing I caught today was a COLLADA presentation. COLLADA is a industry standard file format that allows easy import/export of 3D data. As more companies support COLLADA files (.dae extension), the easier it is to add tools in your pipeline without concern for writing custom importers/exporters.
Remi Arnaud talked about the current state of COLLADA.
3 years old
Made the industry standard file format by the Khronos Group in January 2006.
COLLADA RT is a viewer that will display COLLADA content. COLLADA FX will load and display COLLADA FX files. Both are sample code meant to teach you how to work with COLLADA data. More information is here.
The COLLADA Refinery was briefly touched on. It is a tool that should allow you to massage COLLADA data into formats other non-COLLADA application understand. For example, you should be able to convert a COLLADA file to a list of triangles for a tri-stripper to read. The Refinery should be able to run in batch mode.
Dallas' DNA used COLLADA in the making of "The Ant Bully" via Houdini.
I spent most of my day today at the Nvidia Exhibitor Tech Talks.
The first talk was by Evan Hart, a long time ATI-guy who recently jumped ship to Nvidia. The talk was called "OpenGL Extensions for Next Generation Hardware." Basically, this talk was about new features in the hardware that supports DirectX 10, or Shader Model 4.0. This new hardware adds a new level of programmability to graphics hardware....a geometry shader. A geometry shaders, for the first time, allows a graphics card to create or delete geometry on the fly without any help from the CPU. This opens up many new possibilities: hardware accelerated curves/NURBS, fur, automatic terrain tessellation, and point sprites to name a few. There is no DX10/Shader Model 4.0 hardware announced yet. I asked if current hardware would support the new features with a driver update and the answer was "You'll need to buy new hardware." So if you are looking for a graphics card...wait until you can get a DX10 board. Since DX10 will only run under Windows Vista (not XP), I would expect DX10 hardware around the time of the Vista launch early next year. Nvidia was very careful not to give a time frame for when they would release their DX10 hardware.
The next talk, GPU Physics, was by Mark Harris from Nvidia and Chris (didn't catch the last name) from Havok. They showed how a graphics card could be used to accelerate physics via the Havok physics engine running on a graphics card (called "Havok FX"). In the demos, it appears you get about a 10-100x improvement in the number of collision dections between objects when comparing what you can do on a CPU verses what you can do on a GPU. So instead of having 1,000's of objects on the screen, you get 10's of thousands. Havok can run on any platform (source is given, I believe). Currently Havok FX is limited to using a single GPU (this will change in the future). Future plans include the ability to calculate on the fly how something breaks apart. For example, you could send the graphics card a wall that is basically a cube. If you hit the wall with a bomb, future versions of Havok will be able to automatically break the wall into pieces.
The next talk was about Nvidia's FX Composer 2.0. New features include support for GLSL and OpenGL (no Linux support). FX Composer uses COLLADA as its standard file format for import/export of shader and model data. FX Composer can be used for creating, debugging, profiling, and tweaking shader code via GUI elements (for artists) or code (for programmers). Should be released at the end of 2006.
The last talk Nvidia I went to was "DirectX 10 Effects." Three demos featuring Shader Model 4.0 effects were shown: fur, cloth (that can be torn in real-time), and metaballs. After each demo, we went through how the effect was achieved and studied the shader code.
I picked up another book today (I can't stay away from the bookstores). This one is one of the only new books I saw at the show...."COLLADA, Sailing the Gulf of 3D Digital Content Creation." I am a big COLLADA supporter...I'll have more to say about it tomorrow when I go to the COLLADA BOF (Birds Of a Feather).
The last meeting I went to was for the OpenGL BOF. They had 8 presentations in less than 2 hours...which was a bit too much info in too short a time. They announced with the downfall of SGI, OpenGL would be controlled by the Khronos Group from now on. They announced Vista support of OpenGL is on the same level as Direct3D, so OpenGL will run like it does currently under XP, and not a slower emulated version as was feared last year in order to make it work with Vista's Aero Glass. It sounds like OpenGL 3.0 will be the equivalent to DirectX 10/Shader Model 4.0. It is not expected until SIGGRAPH 2007 at the earliest.
One of the talks was about how OpenGL can better address developer needs. When listening to what people wanted and what is planned by the OpenGL group, I couldn't help but think that OpenGL is trying to be what DirectX is today, but expects to get there in 2 to 3 years (documentation, SDK, code samples, features, tools, etc.). And since DirectX is a fast moving target, I see OpenGL just falling further behind. The only real benefit that OpenGL has over DirectX is that it runs on non-Microsoft platforms (Mac, Linux, Unix, etc.). If OpenGL is not going to lead anymore, maybe it should just become API compatible with DirectX and then we'd only have to learn about one API. It is actually more expensive to develop OpenGL apps than DirectX. You get plenty of really good DirectX developer tools between the DirectX SDK and Nvidia's SDK (both of which are free). If you want a debugger for OpenGL, it will cost you. The only game in town is gDEBugger for $490!
The SIGGRAPH Reception was at the Seaport World Trade Center. I hung out with a friend and former co-worker, Eric Hirschorn.
My new favorite thing to do is ask the locals here in Boston to say, "Don't park the car in Harvard Yard." They talk so wicked awesome around here.
I ended up passing on the SIGGRAPH Chapters party, which was held at the ROXY from 9-2am. I wanted to be fresh for my morning meeting with NatualMotion.
I got a private meeting with NaturalMotion on a boat docked by the convention center. I already knew about Endorphin (which I really liked), so they were showing two new products.
One product is called "euphoria." It basically figures out how to animate characters on-the-fly, in real-time. Consider Madden Football. Instead of using motion capture data from real players tackling a limited number of ways, imagine every tackle being unique based on the position, size, and momentum of players. It also means you won't see one player's arm pass through another player, because collision dectection is handled correctly (this happens all the time in the current version of Madden). I would expect a future version of Madden will use euphoria.
The other product they demoed is called "morpheme." It still works with canned animation, but it allows you to easily blend animation together seemlessly. For example, if you have a walk cycle animation and a fall down animation, morpheme allows you to create the transition between the two animations automagically. This is a huge time saver.
I went to one Paper's session today: HDR (High Dynamic Range) and Systems. The three HDR papers all dealt with compressing HDR (2 for images, 1 for video). The HDR image compression (both versions) had 6x compression ratios. One version stressed image quality, the other stressed ease of compression/decompression. The HDR video compression allows compressed HDR video to play back as standard video on a DVD player or as HDR video with a special HDR player.
HDR is a big buzz word in the graphics industry these days...everything is moving to HDR. Games (like Half Life 2) are using HDR to get much more realistic/natural looking graphics. HDR monitors and graphics cards are coming. I expect after everybody buys all their movies in HD, then the HDR HD versions will come out. Watching a HDR TV verses a regular TV is almost as big a change as watching color verses black and white...it is dramatic. And you thought you were done once you got an HDTV! Wait until you see HDR HDTV!
The other paper was a presentation of Microsoft's DirectX 10 by David Bythe (doesn't really fit with HDR compression, but who cares). This article covers most of what David spoke about.
I spent the rest of the day on the show floor.
Autodesk announced Maya 8. Since Autodesk purchased Alias, now they own the two most popular animation packages in the industry: 3D Studio Max and Maya. Alias and Autodesk in years past would each have the largest booths at SIGGRAPH extolling the virtues of their products. This year they had one booth, and it seemed smaller than what either of them had last year. And now the booth had to split time between both products. Net result: both products were hard to find and did not stand out. There was no big signage for Max or Maya, unlike all the competitors (SoftImage, Lightwave, etc.).
Here are some of the new features of Maya 8. There were other additions, but I guess not important enough for their web site. There were two features that stood out for me. First was the ability to add GUI elements to the viewport. For example, you could add a slider that changes the LOD's of your models and that slider will be persistant on the screen no matter what tool set you are working with because the slider is part of the viewport. The other feature is the ability to override the viewport renderer. That means you can have your game engine do all the drawing (OpenGL or Direct3D). It gives you WYSIWYG for 3D modeling...your model in Maya will look identical to its representation in the game engine because Maya is using the game engine. Sweet! This can really improve the workflow of a 3D production pipeline.
Google had a booth at SIGGRAPH, which is a first for them I believe. They were pimping Google Earth and SketchUp. I already knew a little about SketchUp because my main man Spaghetti is using it. I watched a demo and I can see how SketchUp is really fast, easy, and fun to use. I wondered what limitations it had compared to Maya. Some I came up with after watching a short demo:
No top, side, front views (just perspective).
No NURBS, or SubD's
No rendering.
No animation.
Very simple toolset. May require exporting to Maya for advanced modeling.
This tool is not meant to replace Maya...it is meant to complement it. With SketchUp you can quickly flesh out an idea and rapidly change it. I was impressed with the interface and how intuitive it is.
Pixar had the best (only?) swag...their perennial walking teapot. This year's edition is black with red/orange flames (looks like a hot rod from the 50's).
Nvidia had a cool announcement...the Quadro Plex. The concept is that you put a PCI Express card in your PC. The card talks via a 6ft cable to another box, the Quadro Plex, that contains your graphic card(s). So now you can have a PC with NO graphics card...just a cable to a graphic card box. This solves some cooling and space issues. With a fully loaded Quadro Plex, you can use 4 GPU's. Nvidia supports one PC talking to 2 Quadro Plexes for a total of 8 GPU for a single PC. That is a lot of graphics processing power (which I know our flight sims would eat up and more!). Expect future support for even more Quadro Plexes for a single PC. There is no support for GeForce...just the more expensive Quadro. 2 Quadro Plexes can be connected together to make a unit that will fit in a 3u rack space.
I wore my "My LINUX box crashes more than my WINDOWS box" t-shirt today. It caused several people to stop me and ask what my shirt was about. The best was when I walked past the "Open Source Pavilion." I could hear the people working the booth talking about my shirt as I walked around and looked at their stuff. It was very tense, but luckily no fight broke out...this time.
A company called Sandio Technology had an interesting new 3D mouse. The mouse is based on a standard Logitech mouse, with two extra HAT switches on the side on one on the top. With these extra buttons you can move a 3D object along the x, y, and z-axes. You can also use the extra buttons to rotate a 3D object along the x, y, and z-axes. These extra buttons do not interfere with normal mouse operations. The price is nice...just $79. It is not as slick as my Spaceball 5000. However, I'd rather use this mouse than a mouse/keyboard combination for moving 3D objects and the price is much less than a Spaceball 5000.
I talked to JourneyEd about what it would take for me to buy their heavily discounted software. I asked if I could be enrolled in community college in one class, and they said that would be OK, as long as I have a student ID. I asked if the Maya 8 that they are selling is time limited to one year like previous versions were. They said Maya 8 is not time limited...it is the same as the full priced version. Same goes for the Adobe products. If that is the case, I'm going back to school! Students can get Maya 8 for $389 (reg. $6999) and the Adobe Production Studio Premium for $599 (reg $1699). How can I afford *not* to go back to school?
Microsoft has a small booth with a laptop running Vista Beta 2. There was an engineer there fielding questions and letting people try out Vista. Some things I learned from playing around and talking with the engineer:
Volume icon in notification area now indicates volume level (more bars indicates louder volume).
Button on task bar initiates "Flip 3D" - the 3D task switcher. Can also be started by Windows Key-Tab.
Mouse scroll wheel can scroll through apps in Flip 3D.
New Graphics Driver Model
GPU is virtualized and can be shared.
You can run multiple 3D games and 3D apps at once, even if they all want to use the graphics card exclusively. Performance kept your from doing this under XP...should be OK with Vista.
If a graphics driver crashes (the #1 reason computers crash in XP), driver will silently restart itself without rebooting. Open applications should continue to run as if there was not a problem. This may be one of the most important features in Vista.
You can update your graphics card driver without rebooting the system.
The first class I went to today was called "Physically-based Reflectance for Games." It was taught by Dan Baker from Firaxis and Naty Hoffman from Naughty Dog. Basically they talked about how to make objects reflect light in believable ways, yet still be fast enough for a video game. I believe the course had lots of good information, but they rushed through the slides so fast, it was hard to get any more than the high level concepts. After the talk, I decided to seek out a good reference book on reflectance.
I went to the SIGGRAPH bookstore (again) and I didn't find much selection focusing on reflectance...just two. I decided to go with "3D Computer Graphics", 3rd Edition, by Alan Watt. It has a nice chapter that details many of the subjects in today's class. It also covers lots of other topics I'm interested in (NURBS, ray tracing, shadows). It is an older book, but definitely fills a hole in my personal library.
Another book I picked up should help with picking colors that go together for my web site..."Color Index."
I checked out the art exhibit today. I got some video clips I'll upload when I get back. One bizarre creation was a black room littered with thousands of receipts in piles all over the floor. Each receipt was a few lines of text from a personal ad (like, "I'm looking for a girl that...". Above, in the ceiling, is a printer that prints out a new phrase from a personal ad on the internet and lets it float to the ground every 12 seconds.
Another art piece used the same "magical oil" that was demonstrated at a previous SIGGRAPH. A metal spiral poked out of the oil. The oil would jump out of the pan and cover the spiral and create wild spiky effects...like it was alive...you'll just have to see the video...amazing.
They had a special "teapot" exhibit in honor of the 3D graphics icon...the Utah Teapot. I have a photo of teapot origami, fashioned from a single piece of paper.
For the afternoon, I went to a course called "Spacial Augmented Reality." The main idea is that soon, projectors (like the ones used to show a movie) will be ubiquitous. The technology is shrinking to a point where they will easily be included in cell phones. The problem is, where do you point the projector if you don't have a screen? These guys have done research into how to make almost anything a screen. For example, you could aim your projector at a window covered with curtains. Using a camera and some special graphics hardware (a GPU), they are able to warp the video/image so that when you look at it on the curtain, it looks like a flat surface. They also do color correction so that the video/image looks consistent, even though the background maybe a checkerboard instead of the desired solid white background used for most screens. They take into account some surfaces may be closer or further away and require different focal distances so that it looks the same sharpness in the final output. Impressive stuff.
My buddy John Verostek joined me for the Electronic Theater tonight. We were handed a paddle with a green side and a red side when we entered the theater. When you walked to your seat, you could see on the screen the layout of the theater with green and red lights representing every person holding up their paddle with the green/red side facing the screen. You could find your dot on the screen and flip your paddle to make your light turn red or green. We did some fun interactive games with this gimmick. The best was the giant Etch-A-Sketch projected on the screen. The audience was divided in half. One side controlled up/down by displaying red/green. The other half controlled left/right by displaying red/green. If the audience was split 50/50, then nothing would happen. Mostly green would move up, mostly red would move down. Same goes for the audience with the left/right controls. We were given the task to trace the outline of some shapes: a square, a triangle, a circle, and (of course) a teapot. And how do you clear an enormous Etch-A-Sketch? You shake it, of course! We all shook our paddles in unison and the screen would clear. It was a lot of fun...even though they had some problems registering our paddle information in the section I was sitting in.
The ET was shown on the new Sony 4K SXRD projection system (think 4 times the resolution of HDTV at 1080p). They did a demonstration of 2k (1080p) vs 4k (2160p), but I honestly couldn't tell a difference. To be fair, I was sitting in the back part of the theater.
There were lots of good animations. A couple of beer ads were very clever (Guinness' "noitulovE" and Foster's Australia "Big Ad." Pixar showed "One Man Band", which is the short that preceeds Cars in theaters now. Pixar's stuff always stands out of the pack. A really cool concept video is "Doll Face." I always love the special effects reels (ILM: "Poesidon" and "Pirates of the Carribbean: Dead Man's Chest", Weta Digital's "King Kong", Sony Imageworks' "Open Season"). I think the one that blew me away the most was King Kong's "In a New York Minute" (separate from the King Kong effects reel). It showed Manhattan built building by building in 60 seconds. I went to the talk about how they did it (using procedural buildings), so I already had an appreciation for the work involved. The amount of detail was just awesome.
When I arrived at the Boston Convention & Exhibition Center (BCEC) this morning, I was reminded of why it is a good idea to pay the extra $10 and have your credentials mailed to you...the line to pick up your credentials was over 2 blocks long and not moving! You have to have your credentials in order to get in to the classes. Registration opens at 8am and classes start at 8:30am...a lot of people missed the first few hours because of that line!
I walked past a window into the show floor, which opens on Tuesday. It looks *way* small. I remember back in '97, it took me two days to see everything on the show floor. This year it looks like 2 hours will be plenty.
My first class was Interactive Ray Tracing. When you see interactive 3D graphics (video games) on a computer, they all use a technique called Z-buffering. Ray tracing is an alternative approach to 3D graphics that works by tracing the path light travels from a light source to your eye. It gives much more realistic results, but it is *very* time consuming. Computers are getting so fast that the idea of using ray tracing for video games is gaining steam suddenly. We still have a ways to go, but this could radically change how we work with 3d graphics. Adding extra CPU's (cores) really helps ray tracing.
I skipped the second half of ray tracing to see a class called "Procedural Modeling of Urban Environments." The idea is to create some simple rules (buildings have floors, floors have windows, top floors have a roof, bottom floors have doors), and then let the computer create models of buildings for you. I am currently building the Bank of America building, and it is taking a lot of time. With this approach, you could build a very detailed city very quickly. The city would not look exactly like the real thing, because it is using rules to build a building, not blueprints. It is an interesting idea and I can see it saving a lot of time for flight sims and games.
I picked up a cool book at the SIGGRAPH bookstore. It is called "The Elements of C++ Style." It a small book that talks about how to format your code so that other people can read it. I've just glanced over it, and I agree with their sentiments. Most of the things they write about I've learned over 16 years (has it been that long?) of writing and reading code...it would have been nice to have this from the start because no class teaches this. This is one of those books that I wish my team lead would hand to each member of our group. It will help make your code more readable and look less like a ransom note.
I have lots of pictures and some very cool videos from the Emerging Technologies display. I'll post those when I get home.
I am getting ready for my annual pilgrimage to SIGGRAPH (Special Interest Group for GRAPHics). This year marks my 10th SIGGRAPH in a row! I really love going to this conference. It always energizes me and makes me want to learn more about computer graphics.
This year it is in Boston. I've never been to Boston, so I'll spend a day and a half sightseeing before the conference starts. My friend John Verostek lives in Boston and plans on showing me around over the weekend.
I plan on taking a bunch of pictures and video, but I probably won't post them until I can get home to my desktop and image editing software. I'll jot down my blog posts in a notebook (the old kind that doesn't need batteries and won't get stolen) and transfer them to my blog when I get access to an Internet terminal. I could post via my cell phone...but that is just too painful...even with its tiny QWERTY keyboard.
I've added all the papers, classes, panels, and talks to my calendar. On average, I have 3 things I want to attend at any one time. On Sunday, I have 6 things! You can't see it all, which is part of the draw of SIGGRAPH.
Softimage
SIGGRAPH
Vanilla Ice
Seven Forty
740
Party
Larrabee
Images
Videos
Papers
SIGGRAPH 2008
Autodesk
Maya
3ds Max
Dolby 3D Digital Cinema
Stereoscopy
Hanan Samet
Sorting in Space
Ed Catmull
Ray Tracing
Massive Models
SIGGRAPH attendance
SIGGRAPH
registration
Randy Pausch
Last Lecture
Really Achieving Your Childhood Dreams
Spore
ILM
LucasArts
EA
House of Blues
Softimage
Lucasfilm
UIST
Transformers
Nvidia
Mental Mill
Summarizing Personal Web Browsing Sessions
FX Composer
Electronic Theater
San Diego
PDI
Dreamworks
Party
HP
Stingaree
Autodesk
Maya
Max
USS Midway
User Group
SiGGRAPH
san Diego
Emerging Technologies
Microsoft Surface
OLPC
E-Ink
Fast Forward
Shirts
Hostel
