New Camera / Gigapixel Optics Technology

[JeffGreene]

Clifford Ross, a New York City-based artist and photographer developed some photographic technology that has the imaging experts at Sandia National Laboratories scratching their heads. He's generating images consisting of a billion pixels (i.e. gigapixel images). This is an analog technology. He generates an incredibly dense negative that is then scanned. The link to the article is here.

After reading the article, click on "Image Detail" on the left side of the screen, after that, move the slider to the right to zoom in on the image. The distant detail that this camera can record is phenomenol. I can see, very sharply, four chairs and a table sitting on a deck on the far end of the lake.

What do you think?

 

[tripod]

Digital scan of 9x18" Kodak aerial film taken with a view camera. ? I wonder what the new technology is.

 

[pevelg]

I was shocked by the amazing quality... Up until I saw the camera used 😀:

http://cliffordross.com/R1/R1-shoot-5.html

 

[FrankS]

It's not the same, Fred. Your guy used a regualr digital slr and many images of the same place. The other guy uses a large format camera. I'm not sure that he makes multiple images of the same scene. He's jsut scanning a really big negative and massaging it with photoshop.

 

[David Goldfarb]

All hype. What he's doing isn't terribly different from ordinary large format photography, but he's optimizing conditions by fine tuning the alignment of the camera, I assume the lens is modern, using two tripods, vacuum back, drum scanning I assume, and post processing for maximum detail.

Without doing all that, there's still more detail in a LF neg than most people need. Here's a demo I posted a while back. Most people are pretty impressed by it, even though it doesn't come close to showing the potential resolution of the format--

http://www.echonyc.com/~goldfarb/photo/imviaduct.htm

This image was made with an ultralight 8x10" Gowland PocketView and a 50+ year old single-coated 12"/f:6.8 Goerz Gold Dot Dagor and probably a yellow filter on T-Max 100 film processed in D-76 (1+1), and then the neg was scanned on an old Agfa Duoscan at 1000 dpi. About the only thing optimal in this shot was the film. If I wanted more resolution, I could drum scan at 5000 dpi, and if I'd planned in advance, I could have used a heavier and better aligned camera, a vacuum filmholder (I own two of them), a sturdier tripod and head or possibly two tripods, a modern lens, and a higher acutance developer to have even more resolution, and none of these things are particularly exotic, and yet, without any of those refinements, the neg has more information than I need.

 

[JeffGreene]

.....
http://www.echonyc.com/~goldfarb/photo/imviaduct.htm

This image was made with an ultralight 8x10" Gowland PocketView and a 50+ year old single-coated 12"/f:6.8 Goerz Gold Dot Dagor and probably a yellow filter on T-Max 100 film processed in D-76 (1+1), and then the neg was scanned on an old Agfa Duoscan at 1000 dpi. ....., the neg has more information than I need.

David:

Your explanation makes great sense, I still don't see how he is able to embed the information into the image. He must be using an image database consisting of many of the same type images you identify. According to your explanation he has hot-linked many background images (i.e. a host) into mapped sectors of the initial image. Would he be limited to a certain image depth (i.e. five to six levels deep). It certainly is interesting technology! Thank you for your explanation. Am I interpreting it properly, or missing the boat?

 

[Wayno]

Here's a demo I posted a while back.

David: that's one for the rivet counters! 🙄

Consider me impressed.

 

[amateriat]

This is not exactly new, yet nevertheless intriguing.

But a lot less fun than running around with my RFs. Hell, this is why I ran away from heavy, tech-bloated SLRs.


- Barrett

 

[David Goldfarb]

David:

Your explanation makes great sense, I still don't see how he is able to embed the information into the image. He must be using an image database consisting of many of the same type images you identify. According to your explanation he has hot-linked many background images (i.e. a host) into mapped sectors of the initial image. Would he be limited to a certain image depth (i.e. five to six levels deep). It certainly is interesting technology! Thank you for your explanation. Am I interpreting it properly, or missing the boat?

There are two things going on here.

One is making the high resolution image, which is all that I'm talking about--fairly old-school technology--and that's the main thing that Clifford Ross is saying that he's "invented."

I think what your pointing to is the presentation method on the web, which is this--

http://www.zoomify.com/

Zoomify is definitely an interesting thing, but that's not what Clifford Ross is claiming is his.

 

[gavinlg]

Those examples on his page aren't that impressive. I've seen similar done with a canon 5d/1ds mk2. You just take 50 photos of a scene with a telephoto lens and patch them together with photomerge in adobe photoshop cs3. I do it all the time with real estate stuff, albeit with 4-8 photos.

 

[amateriat]

Those examples on his page aren't that impressive. I've seen similar done with a canon 5d/1ds mk2. You just take 50 photos of a scene with a telephoto lens and patch them together with photomerge in adobe photoshop cs3. I do it all the time with real estate stuff, albeit with 4-8 photos.

That all depends on what you mean by "impressive"...the technique, or the end result? I'm not much into doing a lot of stitching or other assorted jiggery-pokery when I could do the same, or better, in one shot. Yes, the "one-shot" method in this case requires a different sort of effort, but in the ultra-high-res stakes, no one here gets out alive. 😉


- Barrett

 

[gavinlg]

Barrett - I meant the end result. with resolution the process is rather irrelevant - the result is what matters.

 

[amateriat]

Barrett - I meant the end result. with resolution the process is rather irrelevant - the result is what matters.

True. That, and whichever process you gravitate towards. 🙂


- Barrett