When 8 Megapixels Just Isn't Enough

squidfrog writes "AP has an interesting article on a half photochemical, half digital process to produce 2.6-gigabyte photographs at 'more than a thousand times the size and resolution of those generated by a typical digital camera for consumers.' 'A vacuum pump ensures that the film is flat to within one-thousandth of an inch, and a dual-mirror device keeps the film parallel to the lens. Sand bags strapped to the camera and tripod prevent the machine from shifting, and a reinforced aluminum cradle maintains the parts of the camera in perfect alignment.' The images are apparently higher resolution than can be reproduced on available printing technology (5' by 10'), but the designer hopes to use an 18' by 36' digital display wall to reproduce the images at their best possible resolution in the future. The camera has apparently only been utilized for landscape photography thus far."

Comments

Room Sized Computers

[Deliveranc3]

Remember them, just another thing japanese business men will have on their keychain in a few years.

Digital?

[Janosh]

There is noting digital about the camera. It says in the article that the film is scanned after beeing proccessed.

Impressive camera

[donkeyoverlord]

The topic is a little misleading I was expecting that the camera somehow used film to store a digital image (makes no sense I know that's why I was interested). But what is really going on is Ross created a really stable, perfectly focused camera and then scanned the negitive in to make color corrections. The camera is not digital at all.

re:wow

[roll_w.it]

This guy(from the link at the bottom of the article) came up with his own large camera format. But looks like he's found other uses for those big pictures...

Released in 2006?

[JC-Coynel]

2.6GB files would be perfect to use as a wallpaper in Longhorn.

Re:which reminds me

[cei]

I mean its not as if you're going to get that sized film in a roll, is it?

Actually, I believe 9" film is still pretty standard for aerial photography. At least in the old days, they had to do so much overlap to compensate for the speed of the plane (vs headwinds) and other factors, they'd only end up with about a 5" square of new data in the middle of a frame, and they'd have to overlap quite a bit to stitch together an accurate map.

Interestingly, this is in part why RC paper was developed. Fiber photo paper stretched and shrank too much, and when you're doing things like plotting bomb trajectories, the accuracy of your maps is pretty important.

Or not.

Not Earth Shattering, But Advanced

[DonnarsHmr]

You're right. Nothing here is absolutly earth shattering. However, you're overlooking the extent to which the process has been taken. The film flatness is a HUGE issue at the enlargement rations at which he is working. Vacuum systems, while comercially available for medium format, are pretty much unheard of for large format cameras. The mirror alignment check is also a critical detail. Commonly used in telescopes, and within the last few years, enlargers, this is the first camera I have heard of that employs such a thing. Keeping the film plane absolutely perpendicular to the optical axis is, again, critical at these enlargement ratios because even an arcsecond of misalignment will produce visible defocus. The use of aerial film contributes greatly to the finished product. Aerial film has a MUCH higher resolution than standard films. The problem, as stated in the AP article, is using aerial film to reproduce a scene and produce a final print containing reasonable contrast and color values. This is where digital imaging comes in. The negative on the film cannot be used to make a "photo-realistic" print with conventional wet-process materials.

Oh, and it is highly unlikely that he "just stopped down the lens" At smaller aperatures, diffraction starts to become an issue and the resolving power is lowered dramatically. As for the sand bags, their purpose is likely twofold. Well, one purpose, two reasons. Obviously, they're there to reduce movement during the exposure. Part of this need is brought on from the length of the exposure time, but part of it also comes from the maximum allowable movement during the exposure. Take, for instance, the blades of grass. They're x millimeters wide d meters from the camera. From this, you can determine the degrees of arc that a blade of grass subtends. Moving to the back of the lens (inside the camera) you can work from the subtended angle and the distance to the film plane to determine the size of the blade of grass on the film. To avoid triganometry, consider that the entire vista before the camera is shrunk down to the size of the film, a small detail like a blade of grass is really, REALLY small on the film. If the film or lens moves by the size of the blade of grass on film, the blade of grass will be completly obliterated. If it moves even a small fraction of that size, it will be visibly unsharp. There's a reason holography is done on giant, sand filled isolation tables (no, I'm not implying that these photographs are resolved to somthing on the same order as the wavelengths of the light being recorded, I'm just saying thery're out there in the same freaky territory).

This camera isn't a new thing, it's an old thing taken to a place never before explored.

Re:Not Earth Shattering, But Advanced

[n6mod]

So many misconceptions, so little time.

OK, here goes:

Vacuum film planes are ancient tech for prepress cameras. At 20x24 and up, it's not just cool, it's an absolute requirement to use film (instead of plates).

Mirror alignment check to get the film plane and lens parallel? Useless in landscape work. Worse than useless. You don't *want* them parallel. You want the film plane vertical, and you want to tilt the lens forward (top away from the film) to move the plane of focus and *improve* sharpness. Otherwise the only way to get the depth of field you need is by stopping way, way down. And you're right, there are diffraction limits, (you obviously do telescope optics) but they don't start to bite you until at least f/45, more likely f/64.

"Aerial Film has a MUCH higher resolution..." Not really. The color aerial films only have 80-100 lp/mm resolution...pretty much the same as professional chrome film. They have wacky spectral sensitivities, because they're designed for data collection, not photorealistic images, and that's what forces this guy to scan the film and work in Photoshop. There are some very high-resolution b/w aerial films, but they really aren't that much better than something like Tech Pan. The real reason he's using aerial film is because he can get it in that size.

[Note to another poster: You do get it in rolls. In fact, that's the only way you can: 9.5 inches by 200-2000 feet. This guy is cutting sheets off one of those rolls...yet another reason he needs a vacuum film plane.]

Getting film this big is actually a real problem because nobody uses it. I checked out an 11x14 view camera from the cage over Christmas one year, and had to shoot Cibachrome directly because I couldn't get film. EI 6 and 30CC Cyan over the lens, but it worked....and let me tell you, contact prints look soft next to a direct Cibachrome.

Sandbagging view cameras is nothing new...and for all your discussion of the arc subtended by the image of a blade of grass...remember that the grass is likely to move. ;) I don't see any thing special about the "aluminum cradle" either, this looks like a classic studio view camera.

The camera isn't a new thing at all. It's a very old thing, in territory explored and abandoned decades ago, with a few bits of new tech to work around not being able to get the right old tech. :)

Now, with all that said. I do think it's very cool that there's someone out shooting 9x18" film. Big view cameras produce really amazing images, and I applaud this guys work. (I understand the problem too...Mt. Sopris is gorgeous, and all of my photos of it are really dull.)

The real problem here is that the article was written by someone with no knowledge of the subject.

-Z