Summer Drawing

I am still working on setting up a drawing session just for IFDM, but till that happens you will find several resources around town. Since my specific interest is figure drawing, that's where we will start. Fridays at the New Mexico Art League are all day drawing. There are two sessions (morning and afternoon), you can go to one or both. The Harwood also has drawing sessions in the evenings. There was (and still is to my knowledge) a Tuesday session 6-9pm in the old church building that is also figure drawing. Till I can get something going for us, please use these opportunities to get yourself going!

Swanky Retro-Future Diamond Heist

Here's a preview for a neat green-screen based action flic. All the effects were done in After Effects. Could we do better?

Via Boing Boing

Extracting Images From Movies

We now have access (or soon will) to nice (as in clean) moving pictures thanks to our Panasonic AVCHD camera. Now, we need to hack those images! What we want is a simple utility to extract the frames from our .MTS files. It is important to realize that the .MTS file is really just an MPEG4, therefore if you are having problems loading it, you might just try a new file extension. After reviewing several tools I have found one that works just as we need, FFMPEG. It is a cross platform, command line tool that can convert nearly any movie format to any other, including a simple frame dump (what we want). Much thanks to Jeffery Bowles (an Advanced Graphics 413 superstar) for a nice, short description of getting it going on a mac:

MTS files are just mpeg4 files. The command line utility ffmpeg will let you extract the frames:

Building:

svn checkout svn://svn.ffmpeg.org/ffmpeg/trunk ffmpeg
cd ffmpeg
./configure
make && make install

Example usage with an HD camera

ffmpeg -i 00000.MTS -s hd1080 -f image2 waa-%03d.png


Note: the %03d bit above means each successive frame should be numbered using 3 digits. That is, the first frame will be waa-001.png the next will be waa-002.png and so on, up to our last frame waa-999.png.

First Movie

The Library of Congress has begun placing important historical footage on YouTube. Above is the VERY FIRST moving picture, of a man sneezing. Created by Edison Labs in 1893, it used a device called the Kinetograph. This research cost only $13,000 in today's dollars.

Via Wired

Aiptex Camera Test Shot

-- 400x246 reference image of DSC checker from their website.

-- Testing the Aiptex mini-hd camera. My office, 3pm south and eastern light. 720p-30fps.


You can click on the images to see them full size. Overall, it isn't clear that the Aiptex HD really provides much advantage over SD. This is a case of you get what you pay for. The reference image (top) is effectively SD resolution, whereas the Aiptex shot is 720p, which one has better image quality? How do you rate image quality? What are the factors you are using to judge?

Tilt Shift Photography

Bathtub IV from Keith Loutit on Vimeo.
Tilt Shift Photography is a technique for capturing extreme depth of field (among other important effects) in the real world. Notice how everything in the video seems like a miniature model. Why?

Average Faces

In class, we talked about how you can emphasize the uniqueness of a face by accentuating it's deviation from some canonical average face. Face Research has a nice demo that allows you to create your own average, for example: The average male/female, white/black etc. Also note that there is an explanation of the process that they use to create the average faces linked to the page, it's worth a read. Think about why these average faces are pleasing or interesting...

High Dynamic Range Images (HDRI)

High Dynamic Range Images allow us to capture a virtually unlimited range of light intensities (from very dark to very bright) in a single image. This requires more than 8 bits of precision for each color channel, typically 16-32 bits per channel. We can create HDR images of real life by taking multiple images with different exposures and combining them. This can be done in Photoshop using File-Automate-Merge to HDR. Give it a try! Since we can't actually display HDR images on a computer monitor, we have to "tone map" it so that the aspects of the image that we are interested in are visible. HDR images are essential for realistic computer generated images mixed with live action. In the near future, HDR photography will be common place.

Color Theory

Color Theory check it out! You should know this.

Audiofile

Thru You is a very interesting project that produces original break-beat songs using entirely user-generated content from You Tube. Check out I M New as a start.

Consider the complexity, organization, and musical talent required to compose music this way. How do you think it was done? What tools were used. Do you think the song could have been composed in video editing software directly? Seems like they may have used Ableton Live or even an old school sp12, with the video produced after the fact...

via Boing Boing Gadgets

Dr. Manhattan process

From Wired we have a description of the process for bringing a unique digital character to life. I really like that they were able to unify the glow effect (the character must illuminate the real life characters) and the tracking.

Color Space

Color defines a space. How we use this space depends on what our media can represent (consider the difference between our eyes and the monitor) and what we need it to express. Typically in computer graphics we first think of the RGB color space. This space is useful from an engineering point of view (we need red, green, and blue emitters to make a color pixel), however it can be very difficult to work with as human/artist. What other color spaces are there? How is the HSV color space different from RGB? Why might a human find this better to work with?

Just Awsome

Slow Motion Punches to The Face from Eduardo Wydler on Vimeo.

Think about how one might model these dynamics. What kind of deformations would be required to make a character do this?

In case you hadn't heard...

February 18, 2009

Albuquerque Journal

3-D May Spell Jobs for N.M. (UNM involved)
By Michael Hartranft, Journal Staff Writer

New Mexico stands to gain nearly 100 new high-technology jobs under two commercial partnerships with the state's supercomputer program, including a pioneering venture with film studio giant, DreamWorks Animation.Backdropped by Encanto, the state's supercomputer at Intel in Rio Rancho and the heart of the New Mexico Computer Application Center, Gov. Bill Richardson announced that the Cerelink Digital Media Group of New Mexico is teaming up with DreamWorks to explore “cloud computing” or remote rendering for film production. The project would create about 30 jobs in its early phase.The governor also introduced the Insigniam Innovation Discovery Center, a health-care consulting business that recently moved to Albuquerque from San Francisco. It plans to use the supercomputer to develop solutions for improving health-care quality and value, Richardson said.Insigniam plans to hire 60 to 65 people this year and could expand to 300 over the next four years, managing partner Lee Akay said. Former Intel employee and Cerelink managing partner Richard Draper said the trend in computer-generated films is moving toward 3-D. As such productions require millions of computing hours, there is a need for cost-effective computing capacity. “The answer is New Mexico,” he said. Cerelink and DreamWorks — whose films include “Shrek” and “Kung Fu Panda” — will jointly explore developing technology and infrastructure for the studio to be able to render computer-generated 3D films in New Mexico. The NMCAC worked with UNM, the state Department of Information Technology, National LambdaRail and Cerelink to test an ultra-high-speed link between New Mexico and Hollywood. “As far as we know, no motion picture has been rendered remotely, so this is truly an R and D effort,” Draper said. The initial phase will take a few months. “At that point, when DreamWorks and Cerelink are satisfied that part works for remote cloud-computing, then that's when we start looking at the next phase, which is to hire folks to start doing animation here as well as the rendering that would be done from the studios in Los Angeles,” he told the Journal. He said about 30 jobs would be created during that time. Draper said Cerelink will not use the supercomputer in the first phase, but that it holds potential for animation companies down the road.

Introduction to Adobe Scripts

Trevor Morris has a nice introduction to scripting the Adobe Creative Suite applications. Read up! We will be using Adobe tools for post processing our animations. (Note: this tutorial has two parts).

Here is another easy tutorial with nice examples.

The Uncanny Valley

"The uncanny valley is a hypothesis that when robots and other facsimiles of humans look and act almost like actual humans, it causes a response of revulsion among human observers. The "valley" in question is a dip in a proposed graph of the positivity of human reaction as a function of a robot's lifelikeness."

This post is in reference to a comment in class about the "corpse effect". Here is a clip demonstrating how to achieve the corpse effect using photoshop:

Tron

Here is a nice clip of the 80's CG Movie "TRON" set to some classic 80's music.



Notice: The objects (e.g. light cycles) were created using CSG, also try to guess how one might achieve a live-action glow effect nowadays.

Graphics Primitives

In class, we discussed several "classes" of graphics primitives. Here is some more information on each:

• Triangles which can be tied together to make a mesh.
  
• Splines which can be used to form surface patches, for example NURBS.
  
• Subdivision Surface which is similar to Splines but have a very simple and intuitive construction mechanism.
  
• CSG or Constructive Solid Geometry, which builds complex objects from simple primitives.
  
• Volumes which represent not only the surface of an object, but its interior as well. (Good for smoke, clouds, etc..)
  

What are the advantages and disadvantages of each primitive? What other primitive types or variations could we add to this taxonomy?

Sphere Gallery

Spheres are the most basic graphics primitives. Here is a gallery of some not so basic sphere designs. Have a look, this gallery might inspire some novel "bouncy ball" animations.

Learning some Photoshop

I really learned how to use Photoshop doing graphic design and typography. Here is a long list of text effect tutorials for Photoshop. Do one! and discover some new Photoshop features and techniques.

ImageMagick

ImageMagick is a free tool for manipulating images. It generally works on the command line, but also supports python (perhaps it's better to say python supports it, whatever). You can locate and install ImageMagick multiple ways. A quick web-search will help you figure out how you need to do it. I run OSX so I used "macports" to install it: sudo port install ImageMagick Once you get it installed make sure you can use it by typing convert -version at the command line. You should see some information about ImageMagick (including it's version, I see 6.1.8) and the options that convert takes. If you don't see this useful information you need to keep working.


There do exist python bindings for ImageMagick, however after about four hours of hacking I couldn't find compatible versions of ImageMagick, Magick++, PythonMagick, and Boost to make it work in python. Perhaps I'll come back to this topic later, but for now we will focus on using the Python Imaging Library (PIL).