The Online Photographer

Check out our new site at!

Saturday, April 21, 2007

So You Thought You Had Good Buffer Depth

"...We can comfortably say that in 10 years photojournalists will only be carrying video cameras."

—Dirck Halstead

Many readers have recommended Dirck Halstead's editorial essay "The Coming Earthquake in Photography" from the April 2007 issue of The Digital Journalist. Predicting the future has never been a high-percentage business, and it's possible that some of Dirck's predictions are "mileu-specific"—that is, he's looking at photography through the prism of photojournalism—quite naturally, since that's his bailiwick. You might not read the same predictions in the pages of PDN, for instance—I can't see studio advertising photographers succumbing to a pressing need to shoot video—and of course the camera market as a whole, despite the preturnatural popularity of the deathless marketing word "pro," is almost entirely driven by amateurs, enthusiasts, and consumers. It's a fascinating essay nonetheless.

And check out the Red Digital Cinema Camera Company's Red One, which to me looks for all the world like the over-the-top plastic alien-zapping space guns I used to buy for Zander at the Toys'R'Us when he was a little guy. It'll shoot 60 12-megapixel frames per second. And you thought your Canikon had good buffer depth.

Posted by: MIKE JOHNSTON, thanks to erlik

Featured Comment by Juan Buhler: Regardless of whether still photography dies completely, this is extremely interesting. Imagine what would happen to street, 'decisive moment' photography when there is a device that fits in the palm of your hand and can capture 60 12MP images per second. Will pictures that we would consider great today lose some of their value?

I think this goes right in the heart of what is art and what isn't. Is the difficulty involved in making an image an intrinsic part of the image's value? It is a fascinating issue.

Featured Comment by Matthew Miller: This is just the first step. In the somewhat farther future, but very possibly within our lifetimes, lenses will be obsolete.

Instead, a sensor will capture the unfocused image and transform it digitally into the focused one—you'll pick focus point, focal length, and aperture in post-processing. Combine with the idea of constant recording, and you can choose the instant you want and select "exposure" too. The shutter button just becomes a handy way of tagging interesting points.

Sounds crazy? Consider this: a lens is just an analog computer which does a mathematical transform on the input data (light rays striking the front element). Anything an analog computer can do can be approximated with a digital one with the right algorithms and processing power. We don't have that yet, but it's a very safe prediction to say that we will. At that point, instead of primarily analog cameras with a digital recording element, we'll have real digital cameras.


Blogger Kiliii said...

You know, technological predictions seem to me to often be about a decade or two too optimistic.

That fella predicts the death of paper printing in ten years, to be replaced by electronic paper. I disagree-- about nine tenths of the world and even many in the United States will never be able to afford electronic paper or the accompanying support structure. Reliability will be a huge factor to adoption-- it will in my estimation take a half-dozen years for e-paper and wireless internet to become reliable enough for people to use as a daily paper.

I do agree that news photogs will be shooting video soon, but you're right, Mike, that doesn't mean all that much as far as the art and industry of the still photo.

12:01 PM  
Blogger joolsb said...

What a great design for a spaceship in an SF show!

3:36 PM  
Blogger Liquid Air said...

It seems to me he is missing one thing: lighting. Still photography can be shot with strobes which are easily battery powered. Video lighting requires continuous lights which demand a great deal more power. Still photography will still win when you need to be fast, mobile, and carry your own light. What I think is more likely is that over time professional still cameras will start capturing video as well as processing throughput and memory card sizes increase. The 1DmIII already has live preview; it is just a small hop from there to video capture.

On the topic of aspect ratios, we've been cropping 2x3 to 8x10 for quite some time now. I wouldn't be too surprised if 2x3 remains the standard for sensor size. Any modern DSLR has plenty of resolution to allow cropping to a 16x9 HD video signal.

4:02 PM  
Blogger paul norheim said...

A comment to Juan Buhlers comment about street photography: I`m not sure that anything would change much if you used a video camera.

An interesting scene or potential picture may often show up within a time span of, say, one to three seconds. And you wouldn`think of shooting for a couple of minutes in such situations. With experience (which you obiously have, judging from your wonderful photoblog), you may shoot a few frames during those couple of seconds, and perhaps you are lucky and one of them is a keeper. Or, you may shoot only one frame during that short event, depending on principles or equipment.

With digital cameras, you allready have the oportunity to shoot several frames per second; it is a video camera, sort of (with a lousy buffer size). But my guess is that you wouldn`t obtain a dramatical increase of keepers if your M8 or DSLR was capable of taking 24 frames per second instead of, say, three or five.

If, on the other hand, you happen to shoot something with a video camera, and you discover a great still picture within the rest, then of course you can take it out of the flow of mediocre pictures (jugded as still pictures) and keep it as an isolated frame. But to do this sort of editing from video, is, I would think, just as demanding as going out in the streets to look for good pictures.

7:57 PM  
Blogger paul norheim said...

A comment to Matthew Millers comment: Yeah, and let`s say that this sensor is not only capturing everything that is going on around you, while you`re busy doing something else, but, due to some software, also capable of recognising persons as something different from, say, animals, and even your sister Anna as different from your mother Mary (and your dog Albert as different from the dogs in the neighbourhood..)

Then you could search for pictures of your sister Anna and edit them, trying to find the "keepers".

But isn`that like starting from scratch? (see my comment to Juan Buhler above). You could as well go out in the big world with a camera and a lens, trying to find a fragment in time and space, and klick!

In any case, this future development will require a hell of a lot of editing. Perhaps we will end up EDITING our lives instead of living them. Or perhaps the wealthy ones will pay some professional editors to select and present highlights from their last vacation: three boring weeks on the beach with tiny sensors integrated in their swimming suits and sun glasses.

Interesting times.

8:26 PM  
Blogger Unknown said...

Matthew Miller said -

"Consider this: a lens is just an analog computer which does a mathematical transform on the input data (light rays striking the front element). Anything an analog computer can do can be approximated with a digital one with the right algorithms and processing power. We don't have that yet"

Right, and we won't have it until we have sensors that capture phase information accurately. Right now, our sensors only record intensity. Don't know when but that would be a great day.

9:42 PM  
Blogger Unknown said...

A lens is NOT an analog computer.

9:55 PM  
Blogger Tõnu Tamm said...

I think Matthew Miller likes "The Matrix":)

10:33 PM  
Blogger Unknown said...

Re: Matthew Miller's comment,

For more information, do a google search on "plenoptic camera"


I assume that's what he was referring to.

This is Mitsubishi's take on this same idea:


10:39 PM  
Blogger Paul said...

I already know quite a few freelance photographers who have started shooting video. The reason is plain and simple...its good economics. They can sell the footage easily to agencies as the market for video footage is huge. The market for stills is shrinking.

However, the demise of still photography will probably not come about in our life time. The reasons for this are that to create good video requires learning an awful lot of skills; editing, sound, and lighting. At the moment the amount of pure computing power required makes this difficult for a one man unit to do in the time frames dictated by breaking news deadlines. This will obviously change in the future, but I know from my own experiences working as writer and photographer covering events time just seems to fly by as I edit photographs, caption them , write the accompanying piece , upload and then archive. When I just did the photography side I was time pressured, now with the writing I'm even more so. Throw the needs of video into the mix and will I ever get to sleep?

2:26 AM  
Blogger m3 said...

For Matthew Miller Re. The Future of Lenses -
You mean something like this:
An electrostatic beam focus lens equipped with electron suppression grids (SG) was developed to focus and inject high-current ion beams into a linac from a microwave ion source developed for improving the availability of linac application systems. The SG suppresses electron acceleration by the lens electric field. This contributes to the confinement of electrons in beam plasma in the drift space between the ion source and lens. The confined electrons neutralize ion space charge and prevent the divergence of ion beams by space charge force, which enables an efficient transportation of high-current ion beams through the lens from the microwave ion source with small beam loss. High-current hydrogen beams of around 40 mA with 30 keV beam energy, 400 μs pulse width and 90% proton ratio were focused successfully using the developed lens with high transport efficiency of about 70%; without the SG this efficiency was less than 30%. The 90% beam half-width of the focused beam was around 1 mm, which would be acceptable for injection into a radio-frequency quadrupole (RFQ) linac. The microwave ion source system including a low-energy beam transport can be compact to less than 1 m in length with the developed electrostatic lens in comparison with a solenoid coil lens system which is more than 2 m length. Long-time operation stability (16 h×3 days with 0.8% duty: a condition for a typical application; 110 h with 50% duty: 62.5 times of the condition corresponding to 6875 h operation) was confirmed with a focused beam current of 35 mA and change in the current was less than 2%. ??????
(Extracted from

3:09 AM  
Blogger Eolake Stobblehouse said...

I'm a bit confused as to how the sensor without a lens would tell from where each ray of light is coming.

9:58 AM  
Blogger Mike Johnston said...

"I'm a bit confused as to how the sensor without a lens would tell from where each ray of light is coming."

Me, too.


10:33 AM  
Blogger Angelo Jacinto said...

Video, no matter how good still image quality gets, will shoot at a framerate of 15-30fps. You also have 12gb occupying 1 hour of DV footage. I don't think we'll have editors wanting to go through all that footage and storage space just to pick one decisive frame given the pressing deadlines of newswork. That's what we have photographers for.

9:02 PM  
Blogger Unknown said...

I have to say that lighting is the most direct and difficult challenge portable video faces, though lighting technology has been advancing rapidly. It is not unreasonable to believe that a time sync lighting system which incorporates 24 discrete linear flashes per second will eventually be created.

I think the idea of the glass lens being outdated is a little outlandish though. A photograph is simply a representation of light. Because we (humans) use a lens (our eyes) to view and understand light, we must use a lens to capture images which make sense to our eyes. This does not include ideas for abstract photography. We may eventually use different materials as lenses, but that does not change the fact that the light must be gathered and focused with some form of transparent physical structure.

In the end I think people will choose to use still cameras because of the practicality. Even if a video camera could match a still camera most people would not enjoy flipping through a gagilian frames when the goal is to take one beautiful picture. Unless it was a total 'must not miss' moment and you dont care about your editing time, the video camera is just not the correct tool for the job.

It will be nice to see how these two art forms interact as the technologies advance and cause them to cross paths.

Jesse - CT Photographer

12:18 AM  
Blogger hugh crawford said...

It is not unreasonable to believe that a time sync lighting system which incorporates 24 discrete linear flashes per second will eventually be created.


They have been making them for about 40 years. All those food and soft drink ads on TV that show liquids pouring and bubbling are lit with these.

They are amazing , you turn them on and water looks beautiful.

8:04 AM  
Blogger dasmb said...

Still photography will never die in the way this man's describing because artists will keep it alive.

This weekend I went to the MoMA -- it was a real TOP-inspired trip, I saw the Jeff Wall exhibit, traded Cartier-Bresson trivia with a woman from France, and bought a copy of The Photographer's Eye. I had a real good time, and even though they had rotated out some of my favorite photos they had some really great new ones.

There is something about the simplicity of photography versus the difficulty of GOOD photography that will continue to drive the photographic industry long after photojournalists have been pushed into video (something that I agree is inevitable).

10:46 AM  
Blogger D. Kreithen said...

The array (ccd, or similar device) can tell where each ray comes from if each site is capable of measuring phase information. When this is the case, and each site is converted from analog to digital, it is possible (at radio frequencies this is done all the time today - and at light frequencies it is an area of active research) to discern where each light (or radio) source is coming from using matrix algebra. This is surely what the plenoptic camera is doing, but there are quite a few developments that must take place before it is practical at light frequencies at a consumer level of cost. In our lifetimes? Maybe...who knows?

6:29 AM  
Blogger Unknown said...

The researcher at Stanford who developed the plenoptic camera has graduated and started Refocus Imaging, which is trying to productize his work, by the way.

5:37 PM  

Post a Comment

<< Home