Why do the rumormongers not understand explosive technology

[dbyrd]

Everyone wants to think that they are making investments in their cameras. I used a Hassy 500c/m for 20 years, and it was still competitive as far as IQ was concerned as the best medium formats when I sold it.

Now, when you buy a digital cameras you are buying consumable products. This stuff has no value as bling. A camera is like a Big Mac. If you spend several thousand dollars on a top of the line Canon or Nikon, you'd better use it and use it a lot. You don't score any points for owning the grooviest camera; you have to make pictures. In five years, cameras with comparable specs will be selling in Walmarts for the price of a dinner for two at an upscale restaurant; in ten years your telephone cam will make comparably hi-rez photos.

There is a new technology, based on theory that was published in 2004, and already with operational prototypes that will make present digital imagining technology obsolete. The whole question of how many megapixels will be irrelevant.

Take a look at this article in The Technology Review (an MIT publication):
http://www.technologyreview.com/read...rging&id=18293

"At the heart of this camera is a new technique called compressive sensing. A camera using the technique needs only a small percentage of the data that today's digital cameras must collect in order to build a comparable picture. Baraniuk and Kelly's algorithm turns visual data into a handful of numbers that it randomly inserts into a giant grid. There are just enough numbers to enable the algorithm to fill in the blanks, as we do when we solve a Sudoku puzzle. When the computer solves this puzzle, it has effectively re-created the complete picture from incomplete information. Compressive sensing began as a mathematical theory whose first proofs were published in 2004; the Rice group has produced an advanced demonstration in a relatively short time, says Dave Brady of Duke University. "They've really pushed the applications of the theory," he says.

Kelly suspects that we could see the first practical applications of compressive sensing within two years, in MRI systems that capture images up to 10 times as quickly as today's scanners do. In five to ten years, he says, the technology could find its way into consumer products, allowing tiny mobile-phone cameras to produce high-quality, poster-size images. As our world becomes increasingly digital, compressive sensing is set to improve virtually any imaging system, providing an efficient and elegant way to get the picture."

[dbyrd]

The Hassy 500C became available in 1957. The 503CW which is still available from B and H is basically the same technology. At $2700, I imagine it costs roughly the same in inflation adjusted dollars as it did 50 years ago.

This is from Wikipedia:

"In 1991, Kodak released the first commercially available digital SLR, the Kodak DCS-100. It consisted of a modified Nikon F3 SLR body, modified drive unit, and an external storage unit connected via cable. The 1.3 megapixel camera cost approximately US$30,000.


"In 1999, Nikon announced the Nikon D1, the first DSLR to truly compete, and begin to replace, film cameras in the professional photojournalism and sports photography fields. This camera was able to use current autofocus Nikkor lenses available at that time for the Nikon film series cameras, and was also able to utilize the older Nikon and similar, independent mount lenses designed for those cameras.

"A combination of price, speed, and image quality was the beginning of the end of 35mm film for these markets.

"In January 2000, Fujifilm announced the FinePix S1 Pro, the first DSLR marketed to non-professionals.

"In 2003, Canon introduced the 6.3 megapixel EOS 300D SLR camera (known in the United States as the Digital Rebel) with an MSRP of US$999. Its popularity encouraged other manufacturers to produce affordable digital SLR cameras, lowering entry costs and allowing more amateur photographers to purchase DSLR's."

The serious development of the DSLR market has been mostly in the last 5 years. Obviously, Nikon, Fuji, and the other contenders dropped the ball. If Canon had had any serious competition, things would have moved much more quickly. Now with Nikon and probably Sony making a serious play, the possibilities of the technology should come to market quickly.

[liquefied]

That new technology does not seem very promising to me. If I understand it correctly, it's just a very complicated and math-intensive form of interpolation which is the creation of new data based on guesses derived from original data. I would imagine a great majority of serious photographers would not be happy with their cameras simply making a guess at what most of their image is composed of.

I'm not really sure what your point is for the rest of this post though. I think we all understand that camera bodies come and go and are expected to be disposable on a relative time scale.

[SlowBlink]

Well as a Rant I give it 3 out of 5 =

It's hard to say what technology will win out in the market place. Either way it's good reading. I'm looking forward to the day I can get a large format 32mp
for $499.00

Until then my Rebel XT still works as good as the day I bought it. I loved the images then and I still love them. When I do it right that is.

[dbyrd]

Oh c'mon. 4 outa 5, at least, mebbe 6.

You must be the Soviet judge as we used to say
of figure skating in the Olympics
db

[gdl357]

I agree with you 100% except for

A camera is like a Big Mac.

part...

A Big Mac doesn't last more than 5 minutes in my hands.

[seanbaker]

Would someone care to explain to me what about this article indicates any actual change in digital imaging? The article talks about compressive sensing, but what it describes is just a theoretical algorithm for compressing data.

OK, wow, using a (relatively, at least) ridiculous amount of processing power I can store the same image in less space, and ergo transfer it to my computer faster [to say nothing of the increased processor and, perhaps more importantly, power requirements - all in the face of cheap memory]. But all this is moot, too, because the hinge to a 'revolution' as it wants to describe it is the sensor, which is not even considered. I guess they forgot that part.

If you want to completely change the way imaging takes place, we need a completely new way of doing things. I read a suggestion the other day for the 1D Mk IV (in this forum) to include a sensor mode which would allow each pixel to adaptively control its own exposure. WOW, that would be revolutionary. N-th degree HDR, in-camera, and almost independent of shutter speed. ISO would become meaningless. Or how about, since this all assumes massive amounts of processing power (and electrical power), real-time noise monitoring of the sensor? Take what Hasselblad does with a power-cooled back and go to the extreme - real-time monitoring of hot / cold spots, calculation of anticipated RF interference (heck, actively shield it!), and pre-firing of the sensor to capture a null against which to subtract noise just nanoseconds before the shot. I mean, we are talking about changing things, right?

Note: None of this is meant to undermine your actual point, that the latest-and-greatest (SLR) cameras are typically lost on the average consumer, and are in fact far more camera than (s)he will likely ever make use of. P&S are, of course, another story as manufacturers continue to update and add in features to protect the photographer against themselves . Which is where I think the problem actually is - if it made sense to switch P&S's every couple years, then I need to do that with the more advanced cameras as well, right?

[dbyrd]

Originally Posted by seanbaker

Would someone care to explain to me what about this article indicates any actual change in digital imaging? The article talks about compressive sensing, but what it describes is just a theoretical algorithm for compressing data.

Well, if a new algorithm that makes a phone cam produce images comparable to a high-end DSLR isn't new technology, what is? The algorithms, after all, a the real machine.

A computer chip doesn't know what it is. The algorithm tells it be a camera or be a word processor or business ledger.
db

[seanbaker]

Erm... word processor or business ledger, yes, but make it a camera sensor? No. To be clear, I'm positing that how we store and process images may indeed change based on this 'compressive sensing', but that should not be confused with the central role of the camera's sensor, which is what makes it a camera. See the 5Ds slow frame rate (processing) vs. its incredible quality (sensor).

Now were this article about some Tom Clancy-esque technology that would take some test shots through the lens, calculate the refractive error, and through so crazy whiz-bang math be able to regenerate a more pristine image than that which was originally captured, that could change things. Unfortunately, it doesn't appear that this is what that's designed to do, and at that still assumes a ridiculous amount of computational power being available in a camera phone in order to do it.

[dbyrd]

Originally Posted by seanbaker

Erm... word processor or business ledger, yes, but make it a camera sensor? .

The camera sensor is just the keyboard, the input device. Of course, I have nothing against better input devices either. I intended the example of this new technology not as the solution, or even a solution, but only an example of the kind of change to expect.

The point I was making that the first DSLR, designed from the ground up, 2.7 MP, appeared eight years ago and cost $5000. There are now 12.1 MP DSLRs for $800, so that's not quite following Moore's law, but MPs alone are not the best way to judge the improvement in imaging capacity.

Unlike the film cameras, like the classic Leicas and Hasselblads, which were an investment, and you could proudly own these beautiful things, digital cameras, like computers, are consumable items. There is no reason to build digital cameras, with the build quality of a Leica. No one is going to use one of them, doing serious photography for 30years.

[Tapeman]

I justify DSLR bodies by figuring my savings on film & processing. That seems pretty reasonable untill you throw in computer & software costs. I don't count lenses.

My present kit gets me good overall results so I am not motivated to upgrade often.

It's only money!