830 likes | 977 Views
Astronomy Visualization The State of the Art. 207 th American Astronomical Society Meeting Washington, DC January 11, 2006 STScI Public Lecture Series February 7, 2006. Astro-Viz Special Session. Frank Summers, John Stoke Bryan Preston, Lucy Albert, Faith Abney
E N D
Astronomy VisualizationThe State of the Art 207th American Astronomical Society Meeting Washington, DC January 11, 2006 STScI Public Lecture Series February 7, 2006
Astro-Viz Special Session • Frank Summers, John Stoke • Bryan Preston, Lucy Albert, Faith Abney • Call for Viz – 46 submissions • Jury Duty • DVD – 40 visualizations, 2 hours • Session – 9 visualizations
Scientific Visualization in the Age of Buzz Lightyear, Yoda, and Gandalf
Simulation Illustrate point Complex physics Simple geometry Simple lighting Simple camera No compositing Exact / approximate Intellectual Left brain Accuracy Animation Tell story Simple physics Complex geometry Complex lighting Complex camera Heavy compositing Whatever looks good Emotional Right brain Aesthetics Hollywood vs Academia Fantasy Verité
Flying Through Valles Marineris Dana Berry Skyworks Digital Animation
“Valles Marineris” Animation Case Study Animation by Dana Berry and Kees Veenenbos Executive Producers Paul Gasek, Ann Druyan and Steve Burns
The Cosmic Background • Production began in 1978 through 1979 at a cost of $8.7 Million by KCET Los Angeles. • 13 Episodes first broadcast in 1980. • Audience eventually reached well over half a billion viewers in 60 countries. • The series won both Emmy and Peabody Awards. • The spinoff book from the series spent 70 weeks in the New York Times best seller list, including 15 weeks as the number 1 best seller. • Proved that art and science can peacefully coexist.
The Cosmic Background • Several iterations exist. • Tenth Anniversary edition, with an epilogue added in which Sagan updates the science of each show. • 20th Anniversary edition was released with some updated imagery. • The Arts and Entertainment Channel scheduled to broadcast 20th anniversary edition on the evening of September 11, 2001. Consequently the show was shelved and never aired. • The Discovery Channel arranged with rights owner Ann Druyan and SkyWorks Digital to recut each show and swap out nearly all of the graphics. • Series aired on the Discovery Science Channel in fall of 2005 and will be repeated in 2006. The fall broadcast increased the size of the Science Channel’s audience 6-fold. No plans for DVD distribution.
The Task: • PBS non-commercial format 58 minutes to be edited for commercial 44 minute format: 14 minutes to be trimmed from each show. • Prune scientifically outdated portions of the show. • Swap original graphics and animation with newer, more scientifically correct visualizations. • Produce 164 new animations 5 months within small, finite budget. • Minimize impact on other concurrent in-house productions.
Valles MarinerisGraphics Update Case Study • 3 kinds of science visualization: • Data visualization, simulation: reveals information hidden within a data set. • Pedagogical animation: teaches us something. • Experiential animation. Emphasizes emotion, drama and art. Cosmos animation is experiential.
Valles MarinerisGraphics Update Case Study • Make decision to replace existing studio model footage. • Plan production strategy and choose production tools. • Storyboards • Source images • Division of labor • Build accurate 3D model of Valles Marineris Topography. • Add correct color for ground, lighting and atmospherics. • Choreograph Camera Move. • Develop Render Strategy.
Valles MarinerisGraphics Update Case Study Original fly-through was generated from a studio model. (Don Davis, Adlolf Schaller, Rick Sternbach, John Allison, Justin Segal) In its day this was state-of-the-art special effects. Camera begins at high altitude and descends into the valley.
Valles MarinerisGraphics Update Case Study • New Animation was totally data driven MOLA data projected into a deformation plane. • Camera move roughly follows original studio model shot.
Valles MarinerisGraphics Update Case Study Tools Used: Terragen V1 – Generate frames After FX – Compositing, noise reduction, motion blur and color correction Resources Used: Mars Global Surveyor MOLA data set for topography Rendering Hardware: one 3.6mhz Sony and four 3.2mhz Hewlitt Packards (Average render time 46 minutes per frame, 3,000 frames. (cpus spent 1 month rendering this shot in hi-def).
Valles MarinerisGraphics Update Case Study • The Deformation Map, (right) where the deeper the hole, the darker the gray tone. Image on the left shows the deformation map projected and shaded.
Valles MarinerisGraphics Update Case Study • Camera Path roughly the same as original, except the camera spins around to look back as we rise above the canyon.
Valles MarinerisGraphics Update Case Study General look of the sequence has more earthlike feel. The canyon feels broader, wider than in the original version.
Valles MarinerisGraphics Update Case Study • The MOLA data was not high enough resolution to sustain illusion as camera approached the ground. • Soft fractal noise added to MOLA data to help sustain illusion. • Final render showed jitter problems caused by the fractal noise. • Solution to jitter was painful: cut fractal noise down to .001% of its original value, scale up original render by 200% and scale it back down and add motion blur in After Effects.
Cosmos, 25th Anniversary “Evolution is a fact, not a theory.” -Carl Sagan Show 2, One Voice in the Cosmic Fugue
Strange New Worlds Robert Hurt Spitzer Science Center
The Art of Science • Artist’s Concepts Are Powerful Communications Tool • A picture can help explain complex result at a glance • A movie can tell more in short clip • Aesthetic appeal can stimulate interest to read more • Developing the Concept • Identify key science themes of result • Artist, PA rep, & PI discuss visualization • Converge on single image to tell story • Develop “bullet list” of science results for image • Ongoing artist/scientist interaction can speed convergence on final product
Case Study:Transiting Planets • The Data: • Graph of light curve from two stars showing dip when planet passes behind star • The Story: • First light ever directly detected from extrasolar “hot Jupiter” worlds
Case Study:Transiting Planets • Goal for the animation: • Visually explain the significance of the data graphs • Science Bullet Points: • Planet is close to star, Jupiter-sized but very different from Jupiter • Visible light: only reflected light from sun • Infrared light: planet is self-luminous • Improved viewing contrast in the infrared • Geometry of the transiting system: we see planet pass behind star • Dip in light curve implies we see light from planet (blended with light from the star)
Case Study:Transiting Planets • First Sequence: • “Dramatic” shot showing system. • Start in familiar visible light, then transition to infrared view • Show best-guess infrared “color” • Planetary atmosphere derived from atmospheric simulation of “hot Jupiters” • Second Sequence: • Maintain infrared view through movie • Start with familiar overhead view with orbit for reference • Pivot down to our line-of-sight, drop orbit once the geometry is clear • Overlay time-synchronized graph as planet passes behind star
Galactic Center Fly-in Andrew Hanson Indiana University, Bloomington Priscilla Frisch University of Chicago
Galactic Center Fly-InAstronomy: Navigating to Compact Radio Source Sgr A* at Center of Milky Way GalaxyComputer Science: Correctly Visualizing Galactic Scenery across Huge ScalesAndrew Hanson (Indiana University)Priscilla Frisch (University of Chicago)AAS, Washington DC, January 2006 2006 AAS Washington DC
Galactic Center Data Cover an Immense Spatial and Spectral Range • Task: Zoom in from 8 kpc to 3 Light Days away from the Galactic Center Black Hole. • Spatial scale: Over 6 orders of magnitude • Spectral scale: Radio to IR to X-Ray 1010 A (90 cm) 105 A (20 mm) 1 A (10 keV) 2006 AAS Washington DC
Zoom from 8 kpc Solar Distance to 3 Light Daysaway from the Black Hole at the Galaxy Center Axel Mellinger Milky Way Galaxy Montage (many kpc) 2006 AAS Washington DC
Align Infrared, X-ray, Radio Images with uncertain common point sources and very different scales 2006 AAS Washington DC
Rosetta Stone Images needed for Alignment when Point Sources Lacking • Hard to match diffuse object locations • Exact pointing hard to get for large FOV images • IPAC gallery has Combined 2MASS-MSX image R=MSX 6-11 mm, G=2MASS Ks, B=2MASS J. • MSX Image: R=21 mm, G=13 mm, B=8mm • Chandra gallery image has RGB radio, infrared, X-ray image: 21 mm -> 90 cm -> 10 keV 2006 AAS Washington DC
2MASS-MSX(IPAC gallery)MSX(IPAC gallery)X-ray,Radio, MSX(Chandra gallery) 2006 AAS Washington DC
Task: Cross Register Diverse Data Sources • Match both diffuse and pointlike objects • Align small subimages deep inside enclosing images • Match coordinate system and epoch • Adapt computer graphics methods to 6+ order-of-magnitude • Flyin: final pixel is 1/millionth of the starting pixel 2006 AAS Washington DC
Cross-RegistrationSolution Example Dusty Bowshock (Keck) Dusty Bowshock (Keck) Ionized Gas Minispiral (VLA) Galactic Center (Paranal) 2006 AAS Washington DC
VISUALIZATION AS A TOOL TO UNCOVER THE SCIENTIFIC CONTEXT OF THE GALACTIC CENTER SYNOPSIS: This is a pedagogically stimulating exercise because data are embedded in their full multispectral and multiscale context, creating a rich landscape marked by galactic landmarks derived from all colors of the spectrum. Acknowledgments: NASA grants NAG5-8163, NAG5-11999, NAG5-13558. Many individuals helped with the galactic center science, including Tom Geballe, Mark Morris, Mark Reid, and Andrea Ghez. 2006 AAS Washington DC
Hubble: Galaxies Across Space and Time Frank Summers Space Telescope Science Institute
627 M pixels 19,464 pixels 32,195 pixels GOODS CDFS
Resolutions Dimensions Total Pixels