Astronomy Buff

How Do You View a 4 GigaByte Image?

by Tony on December 21st, 2007

All I can say is, you’re gonna need one hell of a monitor for these images…

I was in a meeting yesterday listening to a presentation from some folks at the University of Indiana discussing issues surrounding their One Degree Imager and was fascinated by the instrument.

Sometime in 2010, a 32,000×32,000 pixel camera (that’s a billion pixels folks!), will be hung off of the back of the WIYN Telescope, a 3.5 meter instrument in Arizona. The WIYN was always designed for a 1 degree field of view, but has never had a camera to take advantage of it, that’s what the One Degree Imager (ODI) is designed to do.

“The One Degree Imager will have two big advantages over other astronomical cameras,” said Catherine Pilachowski, chair of the IU Astronomy Department. “First, it’s big. ODI will have a billion pixels, in a 32,000 x 32,000 array. Because it’s so big, ODI will be able to take pictures of a big region of sky at one time, an area of sky more than four times the area of the full moon. Most astronomical cameras only look at a much smaller piece of sky, so ODI will be a big gain for IU astronomers.

“The second advantage is that the detectors in ODI will be able to partially correct for the distortions of Earth’s atmosphere,” Pilachowski said. “Using a special technology, the camera will slosh the image around to follow the small but fast motions of the atmosphere that cause the telescope’s image to blur. With ODI, IU astronomers will sometimes be able to take pictures that are nearly as sharp as those produced with the Hubble telescope, but over a much larger area of sky.” (Full Release)

The instrument is designed as a general use instrument, meaning that it needs to be designed to accommodate lots of research interests. In contrast, the Dark Energy Camera (DECam, 8 times smaller than ODI at 500 MegaPixels), that’s part of the Dark Energy Survey, can be designed specifically to meet the needs of the survey and no one else, a much easier proposition.

What’s weird about the ODI project is that they’ve apparently procured money to build the camera, and that’s underway now, but nothing is being done to handle the data that come off of it.

Right now, they’ve gotten some support from the University of Indiana folks to ‘volunteer’ some of their time to try and develop a pipeline, archive and visualization software, but clearly they are going to need more support from the collaborators if this is to take off. Are all the images just gonna fall on the floor? They gotta go somewhere or there’s no point in building the thing.

Money is tight everywhere these days, I guess. You do what you gotta do with what you’ve got.

I was really impressed by the Indiana guys, they are doing a lot with no resources. I REALLY hope that NOAO, a 40% partner in the project (Indiana is only about 17% yet they are taking the lead on this) will step up with some leadership and provide some programmers for software development.

Anyway, The Dark Energy Survey Data Management team (of which I’m a member), offered any software we develop to help out the pipeline needs since that was one area the two projects overlap. We have to move, process and archive hundreds of GB per day, and we’ve already solved problems that they will have with moving images roughly 8 times larger than ours.

Their biggest challenge will be in visualizing the images. Which brings me to my question raised in the title: How do you view 4 GB images? Obviously, not all at once, but apparently, that’s how the intended customers want to view them. The solution is not as simple as just throwing fast hardware at it because the project wants the data to be useful to everybody, not just those with good equipment.

The High Resolution Imaging Science Experiment has dealt with similar issues, they have dealt with trying to visualize 20,000×40,000 pixel images.

The University of Indiana guys have access to some really cool tools, like this data capacitor, which amounts to one really big, really fast remote disk drive. They also have some nice grid computing facilities at their disposal. To the extent that they can solve the visualization issues, they will be in the cat bird’s seat in terms of being on the forefront of astronomical visualization because the imagers astronomers are putting on the back of their telescopes aren’t getting any smaller….

Technorati Tags: dark energy, dark energy survey, one degree imager, wiyn

Share This

Related Stories

Next Up: Mars Opposition 2007
Why Magnification Doesn’t Matter
Rather Depressing View of the Earth
Binoculars + Astronomy = Tripod
Through Your Telescope: Orion Nebula

POSTED IN: telescopes