Status of current research on digital and analog photography

[Ed Sukach]

I've heard that they actually produced one of those "super lenses"... capable of producing invisibility ... but they dropped in on the floor - and - they couldn't FIND the floor!

That is the explanation given by a couple of the Engineers at the last "after work" Lab party....

 

[livemoa]

Say whut?

Its called science, or more to the point physics.

 

[Ed Sukach]

Its called science, or more to the point physics.

Really?

I leave this discussion open until I see more information about the "science", or more properly "physics" involved here. It sure seems to contradict what I have learned about optics ... but that is really not relevant ... I DON'T know everything about optics.

.... Or - come to think of it - is it possible that it might be what is produced by someone GIVEN a "physic"?

 

[Kino]

I take piktures. Theay are fun.

 

[Eric Leppanen]

1. New sensors are being designed in which the R/G/B sensor array is not side-by-side, but rather stacked in a manner similar to the layered structure of analog film. This may reduce "grain" and improve sharpness of digital imaging. It will eliminate aliasing...

2. The invention of the superlens...allows materials to have a negative index of refraction. This means that a sensor array element can be constructed with a size smaller than the wavelength of light used to create it. This is not theory, it has been reduced to practice and is coming.PE

"A slab of negative-index material could act as a superlens, able to outperform today's lenses, which have a positive index. Such a superlens could create images that include finer detail than that allowed by the diffraction limit, which constrains the performance of all positive-index optical elements....

The hurdle of translating the wizardry of...negative-index materials into usable technology remains. That step will involve perfecting the design of matamaterials and manufacturing them to a price. The numerous groups now working in this field are vigorously tackling these challenges."

Dear PE,

Thanks for posting your thoughts. My interest in generally tracking these developments is to get a bead on when digital capture will cost-effectively do what my LF film gear currently does. My general impression is not to hold my breath.

While the current Phase One P45 digital back (39 megapixel) apparently yields image quality approaching that of 4x5 film, it can do so only when stringent limitations on aperture are applied (e.g., any f-stop smaller than f/11 results in significant resolution loss due to diffraction, see http://www.luminous-landscape.com/tutorials/understanding-series/u-diffraction.shtml). Thus, f/11 in MF digital is roughly equivalent to f/22 in the LF world (i.e., the aperture beyond which significant diffraction loss occurs), a difference of two stops. If one normally presumes a three stop depth-of-field difference between 645 and 4x5 (one stop between 645 and 6x7; two stops between 6x7 and 4x5), then one gains a total of 3 stops minus 2 stops equals one stop of depth-of-field by moving from 4x5 film to 645 digital. Spending $40K+ on an MF digital system just to gain one stop of depth-of-field certainly does not seem very compelling, and when higher resolution MF digital backs come out the problem will get even worse (presumably any aperture smaller than f/8 will result in significant diffraction loss). In a way, MF digital technology is chasing its own tail: it can continue reducing pixel pitch, but only at the cost of reducing lens aperture requirements to a point where depth-of-field is too limited to be of use.

The alternative approach to achieving higher pixel counts is to enlarge sensor size. But this will increase image sensor cost significantly (the larger the sensor, the lower the die yield per silicon wafer) in an industry where costs are already sky high. Enlarging sensor size does not seem to be a very economically promising approach from what I can tell.

Hence my presumption that the two technologies you mention (small sensors with R/G/B pixels, and superlens-based lenses) are collectively considered necessary to take digital technology a quantum leap forward (no pun intended). With a superlens, a small sensor would no longer be diffraction limited at useful apertures, opening the way to significant resolution gains. But while the concept of a superlens may have been theoretically validated, it is very much a science project at this point. We are clearly many years away from ordering a superlens for one's Hasselblad or Technika.

So my take on all this is that, until R/G/B plus superlens occurs, MF digital will largely be technologically stuck in neutral, and film will continue to rule for those who don't need fast image turn-around times.

35mm-derived digital products (such as the rumoured 20+ MP Canon and Nikon FF cameras) will continue to be interesting, but of course these products are not of the same image caliber as LF (4x5 and beyond).

Does all this make some sort of sense, or have I been sniffing the XTOL again?

Best regards,
Eric

 

[Photo Engineer]

Eric;

Your post makes sense to me, but the SA article shows that the superlens and metamaterials are fact. Practical application will need cost reduction, but at the exponential rate of digital growth, this application may come sooner than anyone could predict.

Back in 1980, the predicted growth rate and costs of digital ruled it out as a viable product line. I have seen the figures. With foresight, but much too slowly in the opinion of some, Kodak began the shift to digital based on a positive exponential curve in development of technology and a negative exponential curve in prices.

Unfortunately for Kodak, Ilford and Agfa, the curves were too conservative compared to the loss of sales in analog product lines, and you see what took place in the B&W market. Again, that is not a comment for or against any digital or analog technology, it is a 'report' on the market facts and based on the development and price curves that actually took place.

PE

 

[Ed Sukach]

Eric Lappanen wrote (sorry - I couldn't get the "quoting system" to reproduce this, so I'm re-entering it by hand):

Originally posted by Scientific American, July 2006 (? Is that date correct? - the current issue?)

"A slab of negative-index material could act as a superlens, able to out perform todays lenses, which have a positive index. Such a superlens could create images that include finer detail than that allowed by the diffraction limit which constrains the positive-index optical elements.

The hurdle of translating the wizardry of ... negative-index materials remains. That step still will involve perfecting the design of materials and manufacturing them to a price. The numerous groups now working in this field are vigorously taking these challenges"

Bold type mine, for emphasis.


I would say, a decidedly different tone than "these materials and (inferring) these lenses now exist." From reading this, I would conclude that the THEORY now exists... and it is indefinite how difficult the "hurdles of translating" will be.
Whenever I read the word "could" in an article like this, I always add one other word, "maybe". " .... these materials could - maybe - act as a superlens..."

This all sounds fascinating - not as a proven new product, but at a theoretical level. I'll have to try to obtain a copy of that issue of Scientific American, to really wring that information out - and see what is written there.

The beginning sentence of the last paragraph bothers me, as far as interpretation: "The hurdle of ...., and, "that step will... These are both singular; one hurdle, one step. Continuing... "... perfecting the design of materials and manufacturing them to a price". That indicates two components of ONE hurdle/ step: "perfecting the materials", and, "manufacturing them to a price".

Finally, "numerous groups ... are now vigorously taking these challenges".

Some how .. this doesn't exactly fill my heart with a great deal of confidence - certainly not enough to invest in any enterprise narrowly dedicated to "taking these challenges". We could (maybe) be witnessing a MAJOR breakthrough in optical technology - and I could (maybe) have inherited
$25 million as a result of the death of an unknown uncle in Nigeria. I could, vigorously, take the challenge of getting those funds...

 

[Sean]

So the government is going to allow public mass production of "invisibility cloaks" and spy satellite grade "super lenses"? This should be interesting, guess I'll believe it when I see it. This is Area51 technology not public sector tech..

 

[Photo Engineer]

OMG!

People, please re-read the article. These metamaterials, especially the superlens with negative index of refraction was produced for the first time in 2000. Examples are shown in the article along with drawings.

I'm really surprised at the 'selective' reading you are doing and posting. Anyone, it seems, can use something to prove or disprove what they want. These things were postulated over 40 years ago, and actually implemented in 2000 and now the race is on to use them in digital fabrication. That is my point and is the gist of the article. Please read it for yourselves. Otherwise, stick your heads in the sand!

The future is coming. The rate of approach is unknown and unknowable, but faster than you think regardless of how selectively you quote something.

PE

 

[Donald Qualls]

Okay, let me say that I now stand corrected.

What I was thinking about negative index of refraction should in fact have been a positive index of value less than 1 (i.e. implying that lightspeed is faster in the medium than in vacuum). These negative index metamaterials (which, in visible wavelengths, would require nanomachining to produce) have a *negative* index -- which is to say that the velocity of the light becomes, in some sense, *negative*. That is, while the pulse travels forward (at a speed less than the vacuum speed of light, violating no physical laws), the actual *waves* (electric and magnetic field fluctuations) of light travel backward in the material.

Or so claim the experimenters. They also claim, and seem to demonstrate, that a simple flat slab of the stuff (potentially, for visible light, as simple as a thin layer of silver metal) can act as a "superlens" exhibiting negative index and result in focusing an image with resolution smaller than the wavelength; further that it's possible to uncouple the energy and information carried by a light pulse from the EM waves that, in a vacuum, are inextricable.

Does *not* require rewriting special relativity, but *will* require rewriting optics texts, starting about where high school physics leaves off -- and promises to be the instrument of the next generation of Moore's Law in operation, allowing existing fabrication methods to produce another increment of smaller features in microchips, making denser memory, faster CPUs with lower power consumption, and yes, finer CMOS sensors. Of course, finer sensors aren't really an advantage unless the main goal is to make the camera smaller than the already tiny ones we have, but that's a discussion well outside the realm of APUG.

Apropros of *analog* photography, it appears that metamaterial lenses could, in theory, allow a 35 mm camera with a microfilm based emulsion to genuinely record levels of information comparable to what now requires a 4x5 or larger film (though existing lenses and microfilm are capable of approaching that pretty closely). These advances could also lend some additional life to film, or at least some forms of film, allowing a new generation of ultradense micorforms that could continue to compete with electronic storage (especially given the already-proven longevity of archivally processed silver-image microfilms, and the avoidance of the ongoing costs of media migration that creates such a headache for archivists).

Interesting times these continue to be....

 

[Eric Leppanen]

I agree, be sure to read the article (Scientific American, July 2006 issue, page 60) before drawing any conclusions. The article also provides a variety of weblinks with additional information.

As a scientific layperson, I still have no clue after reading the article as to how soon any of this will become practical reality. Clearly the development of a practical superlens would be a revolutionary development, and historically most revolutionary developments have had their fair share of hiccups before being successfully brought to market. I certainly hope PE is right in that this particular development will yield practical results relatively quickly. I just hope that any resulting field superlenses will be considerably lighter than my 8x10 plasmat glass!

And a cloaking device to keep tourists at bay would be nice too; see http://www.ee.duke.edu/~drsmith/cloaking.html. I always thought the Predator was sorta cool...

 

[Ed Sukach]

OMG!

People, please re-read the article. These metamaterials, especially the superlens with negative index of refraction was produced for the first time in 2000. Examples are shown in the article along with drawings.

"Re-read"? I'm trying to read the article for THE FIRST TIME. I think I only have two out-of-context excerpts at the present.

I'm really surprised at the 'selective' reading you are doing and posting. Anyone, it seems, can use something to prove or disprove what they want. These things were postulated over 40 years ago, and actually implemented in 2000 and now the race is on to use them in digital fabrication. That is my point and is the gist of the article. Please read it for yourselves. Otherwise, stick your heads in the sand!

Chill out a tad! I'm not trying to prove or disprove anything. I am only saying that I can't see definitive proof of anything from the information I have at hand!
A lot SOUNDS like PR "hype"- from the word formations that do not contain untrue information when read carefully - are not false, but ARE constructed to lead a casual reader to "jump to a conclusion" or two. An example might be ... and I'll qualify this by saying that I realize that the way I "read" it might be FAR from what was intended ... "The existence of these materials and superlenses is a FACT." Do they actually exist NOW? ... or are we assuming that after all this time they must exist, somewhere?

The future is coming. The rate of approach is unknown and unknowable, but faster than you think regardless of how selectively you quote something

Wow ... the stress contained in that statement! "Unknowable ... but faster than I think?"
How "fast" do you think I think it is coming?

Please - let's not be this confrontational. I've become deeply interested in the subject - as I am with everything that seems to contradict long established understanding of something so dearly learned and integrated.

Is Eric's information about the Scientific American issue; July, 2006, correct?

 

[Ed Sukach]

I agree, be sure to read the article (Scientific American, July 2006 issue, page 60) before drawing any conclusions. The article also provides a variety of weblinks with additional information.

Will do!! and thanks for the information! Kiitos!!

And a cloaking device to keep tourists at bay would be nice too; see http://www.ee.duke.edu/~drsmith/cloaking.html. I always thought the Predator was sorta cool...

I agree! This was more about the relation between Science Fiction and Actual Science. Interesting!!

Next, I 'll have to struggle with the "security" settings on this PC, to see the specific articles listed. Nothing from the "archives" now works.

BTW ... THis is the best, clearest, instructional definitons of Snell's Law I've ever seen. Highly recommended!!

One other comment: A lot of the theory of "metamaterials" was dependent on the ablity to create a "void" in space (with the recognition of, "It will not be `easy' ").
Huh??? A "deeper" level of "nothingness"??? How will we be able to tell where one starts and the other ends??

 

[Photo Engineer]

Ed;

Eric's posts are indeed substantially correct. The use of the word 'could' is the hangup. They indeed could produce these effects, but the actual fact is that they have produced these effects and the fabrication of such devices has taken place. A diagram of one is shown on page 60, and on page 62 they describe the actual experiment which proved that the negative index of refraction was possible.

As for confrontational, that was not my intent. It just seemed to me that a number of people were denying the possibility of this discovery regardless of the fact that it was published in two respected places. In fact, as I noted above, the word 'could' should be replaced by 'did' as it has been in actual use since 2000 and we really have no idea where it is now.

That brings me to the last comment regarding the future approaching faster than you can imagine. It is true. Unless one is a member of a cutting edge research team like this, it is impossible to predict the slope of the exponential curve that we are on regarding the development of any electronic or digital device. You might want to read "The Singularity Approaches" by Kurzweil. It describes this phenomenon. The 'toe' of a development curve of a material in its early stages of development does not give us a hint of the slope of the exponential rate at which it will be adopted in industry or by the average individual.

As for where the superlens is right now? They made one, but can't find it. It is invisible.

PE

 

[Ed Sukach]

They indeed could produce these effects, but the actual fact is that they have produced these effects and the fabrication of such devices has taken place. A diagram of one is shown on page 60, and on page 62 they describe the actual experiment which proved that the negative index of refraction was possible.

I'll get that copy of Scientific American as soon as I can.

 

[roteague]

Eric's posts are indeed substantially correct. The use of the word 'could' is the hangup. They indeed could produce these effects, but the actual fact is that they have produced these effects and the fabrication of such devices has taken place. A diagram of one is shown on page 60, and on page 62 they describe the actual experiment which proved that the negative index of refraction was possible.

But, going from this level to manufacturing is quite another thing. The last I heard, Canon was still getting an almost 90% reject rates on the sensors it makes for its MkDsII camera.

 

[Photo Engineer]

Bob;

Again, unless you are part of such a cutting edge team, you do not know the level of research, or the level of rejects. Kodak had a 90% reject rate on Ektar 25, but the customer base here still criticizes them for cancelling the product.

Think on that!

PE

 

[Petzi]

I'm trying to imagine the amount of criticism Canon would get if they discontinued the full frame digital cameras.

 

[Justin Cormack]

What is interesting about potentially much higher resolution digital is that it might go from being analogue to digital.

At the moment CCDs and other digital capture is an analogue device (with ISO being an amplifier), digitised for readout. Analogue film on the other hand is "digital", with grains being more or less black or white. A process that makes much higher resolution digital sensors would probably end up being more like film but with a digital readout.

I havent read the article, but suspect this stuff is a long way away from making a difference to any sort of photography, analogue or digital. Its about chip etching really isnt it.

 

[Photo Engineer]

What is interesting about potentially much higher resolution digital is that it might go from being analogue to digital.

At the moment CCDs and other digital capture is an analogue device (with ISO being an amplifier), digitised for readout. Analogue film on the other hand is "digital", with grains being more or less black or white. A process that makes much higher resolution digital sensors would probably end up being more like film but with a digital readout.

I havent read the article, but suspect this stuff is a long way away from making a difference to any sort of photography, analogue or digital. Its about chip etching really isnt it.

What?

Digital is digital from the fact that each level of light is a discrete step just as on a digital screen. It is countable and predetermined. Density can be 256, 1024 and etc. Colors are created the same way from a set of 256 or more colors as defined by the number of color pixels. In effect it is like a quantum step to change from one state to another. (gross oversimplification here)

Analog is a continuous change in level and color with no discrete steps.

So, with analog, with good instrumentation, you can have densities of 2, 2.1, 2.271, 2.384, etc but with digital you would have 2, 2.1, 2.3 and 2.4 regardless of the quality of your instrument. In analog (silver) imaging there are theoretically an infinite number of steps, but with digital, as stated above, the number of steps are preordained when the sensor is manufactured.

In analog, you can change the number of 'pixels' by changing development, but in digital, although you can effectively decrease the number of pixels, you cannot go above the limit built in by the chip maker.

PE

 

[roteague]

Hi Ron,

I saw the article you were referring to in the July Scientific American, I believe. Writting in a little more layman's terms. I haven't read it all, but just glanced through it.

I think what Justin is referring to is that the actual CCD itself is an analog device; a sensor is more like a transistor (which is an analog device). In this case, the "sensor" changes output voltage based upon the strength of the light striking it. One problem is that when the signal gets weaker, it becomes harder to distinguish the real signal (light) from the noise of the sensor itself. That signal is fed through a low-pass band filter (which is also an analog device). The output from all this is fed into an A/D converter - where the digital conversion takes place. This may be a bit fuzzy, it's been a lot of years since I got my Electrical Engineering degree.

 

[Photo Engineer]

Bob;

Having done process control for years, I am familiar with A to D and D to A conversion and all you imply.

I agree with you, but feel that there is no need to confuse the 'unwashed masses' if you all forgive me for that phrase. I mean nothing derogatory about it.

I felt that his explanation would be cnfusing to most.

Sorry if it made things worse.

PE

 

[roteague]

I agree with you, but feel that there is no need to confuse the 'unwashed masses' if you all forgive me for that phrase. I mean nothing derogatory about it.

I felt that his explanation would be cnfusing to most.

Sorry if it made things worse.

PE

Ron,

I understand. You are correct, of course. My knowledge of these things is pretty fuzzy these days - I've been a programmer for the past 15+ years or so.

 

[jd callow]

This may be a bit fuzzy, it's been a lot of years since I got my Electrical Engineering degree.

I once worked with a programmer who we referred to as super geek (no insult, in fact a high complement). He had his degree in electrical engineering. The grace (grace is the only way to describe it) and logic of his systems were simply amazing. I wonder if it is the discipline...

 

[RobertP]

Whew.....This reminds me of the experiments of a cloaking device that Einstien was supposingly working on. It was something like using magnetic fields to bend light waves or something along that line. I think it was called " The Philadelphia Experiment" or something like that. Then again maybe it was just a movie i saw.