The Bar Also Rises

By Ctein

Before I get started on this column, I have a request from a reader that I'm sure some folks here can help with. Please don't post responses here; e-mail them to me at ctein[at]pobox[dot]com. Thanks very much!

"I would like your advice on purchasing a slide & film scanner. I have thousands of both and value my time, so I'm willing to pay more for a faster and higher-quality one. I'd like to keep it under $2000, but will go higher ($3K) if there is a significant advantage. Any recommendations?"

___________________

Robert Roaldi raised some interesting issues in his comments about the Canon G9 Review. If digital cameras have reached 35mm standards, at what point can we say, "Enough is enough?"

Food for thought. Just what do 35mm standards imply? The picture above is the very first color print I ever made (probably; my record keeping was not impeccable in 1970. It might have been the second or third.) It's an 8x10" print from 35mm Ektacolor, cropped to fit the paper. Below is an enlarged section from that print.

Within months, I pulled all my savings and bought my first Pentax 6x7. Why? Because I couldn't make a  color print from 35mm that came close to my grain and sharpness standards. Keep in mind that my preferred B&W film was Panatomic X. The quality of color negative films of the time was more like Tri-X. If that was the very best I could do, I was not going to be a happy camper.

There was always Kodachrome II, but making really good prints from slides required very advanced printing techniques involving contrast control and color correction masking. In truth, that didn't change until digital printing came along. (With all due respect, anyone who thinks they make good prints from unmasked chromes on Cibachrome or Type-R materials is not using the same definition of "good" as I do).

But, by the late 1980s slow color negative film had surpassed Kodachrome 25 in both fine grain and resolution. By 2000, film of any speed that was only a sharp as Kodachrome 25 would be considered a low-resolution film. One could routinely make 35mm photographs that were as fine-grained and sharp as what I could do with medium format 30 years earlier.

This did not stop photographers from wanting further improvements in films. Medium format quality in 35mm is quite wonderful, but why not hope for large format quality? We're nowhere near the ultimate limits of photographic quality, where any further improvements would simply not be perceiveable. We have surpassed "good enough;" we're far from "can't get any better."

Digital, having reached an affordable parity with film, is taking up this challenge. Some low-priced digital cameras produce less grain (noise) at higher ISO's than the comparable 35mm films, although they may not be better than low-ISO 35 mm films. The midrange and expensive digital cameras better even those. But, medium format is more of a tossup, and we haven't reached an economical large format quality level yet.

What is remarkable and noteworthy is that there's nothing fundamental that prevents us from doing so.

___________________

Ctein

Featured Comment by Jean-Pierre Grüter: "As a professional photographer I compared the Phase One 39MP Back with 4x5 inch Film (Hasselblad Scanner) because I have to produce often huge prints like 1.1m x 1.5m. I did this test one year ago and ended up in buying a Hasselblad Scanner. If you scan at the highest optical resolution of the scanner you get a file with 80MP. I found that I got more information in those scans than in the Phase One files. The huge prints are sharper and hold much better together; it feels more realistic. The digital files when highly enlarged look like watercolor paintings but they keep more details in the shadows and in the highlights. If you shoot with 4x5 inch Provia 100 film you have less dynamic range but more resolution and very crispy images. As I do a lot of architecture photography I need wide angles. Here I was very disappointed about the possibilities of the Phase One Back and the available lenses."

Comments

[the editor]

Ctein,
Your experience over the years brings back fond and not-so-fond memories of results with various films.
Digital will undoubtedly continue to evolve rapidly. The latest generation of Canikon cameras is a case in point, with their low noise, full-format sensors and incredibly high speeds. They can be argued to have gone way beyond film already.
The problem for me is often quite the opposite, though. Shooting my M8 or even lesser Canon bodies, I often find myself adding grain in post.
What would be fun to have are a new generation of high quality digital options that mimic certain classic film looks as an in-camera option.
Perhaps Kodak would license a Tri-X look to the big equipment makers, to take one example.

Glimpse

[The Lazy Aussie]

What's the comparison with the eye? I suppose the bar will go as high as its resolution and the number of shades and colours resolvable by eye and brain for cameras, prints and screens.

[Carl Blesch]

In 1967, I bought a Rolleiflex twin-lens reflex, and marveled how much better my prints were than what I had been able to get from 35mm - especially in Tri-X. Best I seemed to get in 35mm was an 8x10, and it would tolerate little cropping. With my Rollei, I could crop significantly for 8x10 or print high quaity up to 16x20 from the full frame.

Fast forward to 2003. I was shooting Fujichrome 400, hailed by the popular photography press as a breakthrough in fine grain and speed. It was priced to match the hype also - like around US$15 at the non-discount stores. I too thought it was wonderful, until I joined a pack of photographers shooting some of the last Concorde aircraft arrivals and departures from JFK airport in October of that year. I compared an almost identical shot made by someone with a Nikon D100, and the digital image was so much sharper and cleaner than my transparency. I know - could have been the lens, the aperture, the exposure or whatever. But I blame the film. When the just under $1000 D70 came along a half a year later, I made the move, and have never looked back.

[Hitesh Sawlani]

The biggest area for further improvement (in my beginner's opinion) is getting a better dynamic range on digital sensors.

Exposure blending and such should be a thing of the past.

[Robert Harshman]

TO: Robert Roaldi

To answer your question, one needs to understand what you want to do. What one wants to do with film images is critical in the digitial age. Are you looking to produce 8 by 10 prints or ....

Scanning 35mm slides or negatives is awash with so many issues it can make your head spin.

Quality scans take a lot of work and time. The prep, scan and post production all require skill and time.

Hence, what's you final desination? How big of prints or ...? And if you have thousands as noted are you retired or just have complete freedom to spent on this project?

In general, scanning takes a lot of time, quality scanning of 35mm format take a lot more time and expense.

Define your end needs and then edit,and edit and edit again. A perhaps useful benchmark at this stage is figure about one hour per image -ouch!

You might also consider finding a quality scanning service that meets your needs, but if done right it will not be cheap, $35.00 per images is close to the bottom end for this well done.

Edit :)

Robert

[John Camp]

I'm sure there are technical improvements both possible and on the way, but there are some limitations that don't have so much to do with equipment but with "need" and "money." People have said for years (I won't vouch for the truth of it) that the consumer-computer-chip-speed race was driven by gamers. That development slowed down two or three years ago (or maybe five years?) when we began approaching real-time speeds for figure generation.

The same could happen with cameras -- once we have cameras that can produce high quality color and detail in a print size that coordinates well with a typical home art-display space, what large group of consumers will pay more for prints that will be, say, twice that big? The sales of small digital cameras are already slowing, not just because of saturation but because of satisfaction: a G9 already makes prints good enough for the large mass of consumers.

With film, a huge semi-monopolistic company like Kodak or Fuji could put a small group to work on improvements, and space the improvements out every few years, and the customers went for it because their incremental cost was almost nothing -- an additional buck for a whole new higher-quality sensor, usable in their old camera. Now that you need a whole new camera to get a whole new sensor, and you have to pay for all your "film" use up-front, the cost incentives are reversed: if you're satisfied, it pays you *not* to buy for a small increment of increased satisfaction...

It could be that the high end improvements will increasingly be niche products, built at great individual cost and sold to relatively few professionals for very specific purposes...Or that while possible, they will simply be seen as unnecessary, and not be built at all.

JC

[Bill Mitchell]

"But, by the late 1980s slow color negative film had surpassed Kodachrome 25 in both fine grain and resolution. By 2000, film of any speed that was only a sharp as Kodachrome 25 would be considered a low-resolution film. One could routinely make 35mm photographs that were as fine-grained and sharp as what I could do with medium format 30 years earlier."

I have to disagree with this statement. I have Cibas and Dyes from Kodachrome II and PKM far better than anything I've gotten from any color negative material (including Ektar 25).

[Stephen Best]

While digital is still advancing in leaps and bounds (mostly in accessibility, convenience and amazing high ISO performance) film hasn't been sitting still. Current B&W and transparency films are the best they've ever been. It holds true for both digital and film that you still need to move up in format (larger sensor or film area) for a real quality jump.

One could argue that the real revolution has been in digital printing, and the accessibility of the technology by one and all. For the print on the wall it doesn't matter what the original is so in this respect digital capture hasn't really changed much. Real image quality comes at a cost and/or greater rigor. Waiting for digital to somehow repeal this is counterproductive.

[Eolake Stobblehouse]

"But, medium format is more of a tossup, and we haven't reached an economical large format quality level yet."

Are you sure? According to the testing by Mike Reichman and his friends, the 39MP cameras are neck and neck with their results from 4x5 cameras. (Of course you may be talking about 8x10. In which case a scanning back is required, I guess.)

[Martin Doonan]

There's an interesting link between your initial request and the meat of the post: scanning and the mixed workflow.

My observation is that, in terms of quality, digital only is being compared to print-from-film, rather than print-from-scan: I think the latter produces the higher quality.

The last 2 cameras I bought are both film cameras (4x5 & 6x7) to enable large prints. I'm only running enlargement factors of about 6-8:1 (30-40" prints) but the results from the scanner suggest I could easily go to 12:1 (but I run out of wall space). For an MF digital you'd need enlargements of 30:1 to match. Not sure they'd stand up to that.

Comparisons will always be determined by the print size you want.

The scanner is getting detail at the limit of the film & digital processing helps tidy up imperfections.

Of course, scanning is time consuing but works out OK for my low volumes. I'm a long way from abandoning film.

[Alan Little]

Eolake: how do the (doubtless very very good) cameras Michael R and his friends are paying tens of thousands for compete "economically" with a decent basic view camera rig that one can acquire for a couple of grand?

(Probably quite easily if you're a pro chewing through lots of expensive large format film & processing, especially as you can amortize the cost of the hardware. So there is definitely an economic case for e.g. advertising & fashion shooters, despite the shocking startup costs. Outside that niche ... ?)

Those interested in a "TriX look" for digital might want to check out Grubba Software's TrueGrain: http://grubbasoftware.com/

(Disclaimer: I have no connection with Grubba Software and haven't used the product. It's on my "somewhat interesting things to maybe check out one day" list. I notice TriX isn't actually on their list of emulated films just yet; I assume it must be on their to-do list)

[Ctein]

Dear folks,

So many seriously good comments! I'm going to have to reply briefly, en masse.

Editor: There are several software products that do a very good job of emulating the grain and tonality/color rendition of conventional films. Rather than degrade the in-camera image, you can do this better in post-processing.

Aussie: The human eye can only resolve about 800 distinct grey levels (less than 10-bits), and far fewer than 24 bits worth of colors, but they span an insanely long luminance range and wide color gamut. Also, the human eye is very good at enhancing local differences that improve tonal, color and detail discrimination; it's a nonlinear system.

Hitesh: Dynamic range on some under-$500 digitals is already as good or better than slide film. $1K cameras match or better B&W neg film. High-end cameras are just about where color neg film is. I have no doubt we'll continue to see improvements in this.

John C: I expect the rate of progress to further slow down (it already has, considerably). It's not likely to stop so long as camera makers need to sell new cameras to stay in business. And the cost/performance ratio still has LOTS of room for improvement.

Bill M: I can't speak to your print quality, but I can state as fact that color neg (including Ektar 25) can blow away the sharpness of KII prints (remember, I'm the guy who comparison-tests all this stuff). The sharpness of dye transfer prints isn't even in the same ballpark as Ilofchrome/chromogenic prints, not even close. Dyes simply are not inherently capable of it.

Eolake: You missed the word "economical." A 39 MP camera isn't close to economical for mere mortals like me. Even a $5-8K camera isn't. "Economical" means "what a 35mm camera costs" (and I ain't talking Leicas!). If it ain't about the bucks, digital surpassed medium format film 20+ years ago... albeit at breathtaking prices.

pax / Ctein

[Peter Robinson]

I think Robert Harshman should have addressed his comment towards Ctein's reader, not Robert Roaldi, however, it's true, scanning is just one step of a process towards an end goal. Without knowing what that end goal is, it's not possible to help. It would be interesting to know what this reader is planning to do with the 'thousands' of scans that he will end up with because that could take even longer, and cost more, than the scanning process.

[Player]

Maybe I'm reading into this too much, but a comment by the first poster, "the editor," struck me as insulting (though not personally):

"Shooting my M8 or even lesser Canon bodies,"

Even lesser Canon bodies!? Why are Canon bodies "lesser"? Do you mean an SLR is "lesser" than a rangefinder? Or that Canon is "lesser" because it's not Leica?

These are entirely different tools.

[Peter Hovmand]

Well, "dont take my Kodachrome away" - as Paul Simon sang in 1973, here meaning: Be careful with your critique. Kodachrome was in fact the sharpest film until Velvia finally came in the '90s. Talking about slides and scanning, here are some examples of Kodachrome scans on a Nikon Coolscan, a highly recommended scanner, pretty good at handling Kodachromes:
http://landscapes.dk/alps

[Tim Gray]

I think the reference to 35mm as the "standard" is more about nostalgia than objective quality. Since the previous poster referenced Michael Reichmann, I guess I will as well but in a different context. (Isn't the appeal to authority a wonderful tool :) When the 3mpx D30 was first introduced he said something like (and I'm going from memory) "At prints up to 8x10 the D30 surpasses film." He initially took a lot of flak, but in the end there really wasn't much to argue about.

[Claire Senft]

I intend to follow an alternate route. I intend to stick with film and the darkroom. It is I believe the best choice for me. It may well not be the best choice for others.

[Max]

Nobody believes human limitations are the other serious issue here? In terms of image quality, noise, resolution, even if the gear is capable of more, squeezing out that extra drop means a lot of effort. With the last generation of 35mm film very few users had the gear and the knowledge to get it. At least to me, depth of field distribution was very elusive in medium format, for example, and when you have a mediocre setup, that's ok because everything looks the same. When you have a better lens, a better film/sensor, a good tripod, you start to push the quality boundaries, and suddenly your incompetence is the bigissue. That's why I still think that there's a plateau in the amount of quality needed to satisfy the bigger part of the users, and that would define how much is enough money spent in manufacturing better quality optics and sensors. For most people, 35mm was already there, I think. For most of them, digital has only brought comfort. Niche markets will be there always, and those would continue a little further, but for most of the population I think the only real step ahead would be faster sensors that would improve the whole experience without most people even knowing why.
Apart from that, it would be more of the face detection kind of improvement.
An idea for a tecnological move:
Would it be feasible for a camera to scan an image through the whole focus range, keeping hundreds of images (the information of a tridimensional image, in fact) and then using software the user could choose a virtual focus plane, in fact it wouldn't have to be a plane either, and create an image choosing focus distribution afterwards? Since the image has the distance info, that could also be used for selecting planes and stuff like that.
I think the tools are available, probably a lot of resources must be polished, but it would be feasible. And the possibilities...

[Mike]

"Even lesser Canon bodies!? Why are Canon bodies 'lesser'?"

I read that as "lesser Canon bodies" meaning relative to Canon--lesser bodies being Xti, 40D, "greater" Canon bodies would be 1D, 1Ds...no?

Mike

[the editor]

Thank you, Mike. That's exactly what I intended to say by "lesser" Canon bodies.
I'd be very happy to have a 5D, or one of the newer full-frame bodies, except that the M8 has eaten my budget (more or less contentedly) for the next year or two. My "lesser" body is a 20D, and though the camera has its uses, it is definitely a second-tier offering from Canon.
The other thing I'd like to clarify involves my suggestion of in-camera film looks. This need not degrade the file. You could conceivably devise a system whereby you preserve the RAW file AND get the patented look of a variety of films, straight out of the camera.

Glimpse

[aizan]

i wonder if it'll ever be possible to take a photo and enlarge any portion of it to life-size with perfect clarity.

[Mike]

Analyzing this historically, what usually happens is that the push toward improvement is spirited until a point of sufficiency is reached, at which time it calms down and fades into the background. Right now we're still very concerned with image size, because we're still very close to the time when digital sensors didn't have very many pixels. But that won't go on forever, or at least it's unlikely to if history is any guide. More likely, as soon as that problem is agreed to be overmastered--it can't be long now, although I don't know at what point it will occur--then it will probably recede somewhat as an issue, and the competition will calm down.

Mike

[Peter Robinson]

"i wonder if it'll ever be possible to take a photo and enlarge any portion of it to life-size with perfect clarity."

Easy peasy, they do that sort of thing on CSI all the time...

[Paul De Zan]

I think we are nearing the end of the line for still cameras for at least some applications. A lot of work is going to be shot with super hi-def video and prints will be made from carefully selected stills. With digital there is absolutely no reason not to do this. I wonder if there are not some wedding photographers already moving in this direction.

[Lloyd Chambers]

Really large prints are a challenge in any format, but I do have several 4' X 5' prints on my wall, taken with my 4X5 camera on Fuji Velvia or Provia, printed from drum scans (and/or Imacon).

But these days, stitching is a quite viable alternative for some work, and can certainly exceed 4X5 quality if excellent technique is used—I’ve done it. I'm personally not very impressed with some aspects of medium format digital (see my medium format reviews at diglloyd.com). Especially the Hasselblad H3D-39--it's pretty poor on a pixel quality basis IMO.

But I've also seen Bill Atkinson's prints from his PhaseOne P45+ and they are extremely impressive (billatkinson.com). But even Bill lusts after "more"! And then there's Charles Cramer (charlescramer.com), a veteran 4X5 shooter and master printer who now also uses a PhaseOne P45. These guys have a keen eye for quality, so I think it's fair to say that the decision of 4X5 film vs medium format digital is one that each photographer will have to make a personal decision on.

[Peter]

Max:

http://graphics.stanford.edu/papers/lfcamera/

[Ctein]

Wow, I thought people would be kinda bored by this column. Shows what I know.

Anyway, new round'o'comments...

First, a clarification-- my earlier comment to hitesh about dynamic range comparisons is about film used normally, not heroic experiments or exotic technique (Yeah, I know how to put 20 stops on BW neg film. It ain't normal nor useful 99.99% of the time, really).

Peter H: "Sharpness" is hard to measure; it's a subjective impression that combines resolution, acutance, contrast, grain and color and tonal differentiation. That's why I talked just about grain and resolution. Anyway... Ektar 25, 1989. In-camera resolution of 150 lp/mm (I've got the negs) which is 50% higher than the theoretical limit of KII. Also finer grain. What you said about Velvia (1990). Quickly followed by Lumiere, Provia, and Konica Impressa 50. All substantially higher res than KII, most finer-grained. By 2000, ISO 800 and 1600 films were routinely resolving 125+ lp/mm (although the grain was certainly not even close to KII).

I ain't saying you shouldn't like KII. I'm saying it stopped being the objective standard of maximum image quality almost 20 years ago. Dem's da facts.

Tim: Why 35mm as the "Standard"? Because for 3-4 decades, 90% of pros and 99.9% of everyone else's photography has been 35mm. "Standard" doesn't mean "best" (I still use medium format), but it's the meaningful reference.

Max: Volumetric capture already exists. Just not affordable. Yet (kinda like medium-format digital quality in the 80's: you could get it... for megabucks).

Aizan: How big is "life-size?" I routinely do macro work in the 1:2-2:1 range where that's no problem. It'd be a tad tougher if your subject were a blue whale. [grin]

Mike: yeah, the insane pixel race finally slowed down (tho' it still runs faster than it should).

pax / Ctein

[Max]

So, Pete, that's exactly what I was talking about! I wonder if you can choose different focus plane angles through software. I got thinking of this because talking about image quality improvements, focus is one elusive issue, and one incredible way of improving things for those that know nothing about it would be a touch screen were you pointed your finger at the area you wanted in focus (or at several of them), in-camera, but then it became obvious that would work great in post processing if all the depth information could be recorded. That would be having tilt capabilities during post processing, would't it? And also you could for example use distance info for selecting image areas, edit according to image planes, etc.
You could choose bokeh qualities for example.
I was thinking of sequence of images shot with a traditional sensor and lens, from closest focusing point to infinity, analized through software allowing choice of focus plane. But yes, it looks pretty much like that link you pointed at!

[J Alan]

How come we (as pretty much the world wide web)endlessly talk about technical feautures and hardly ever about creating visionary photographs?

[Mike]

Well, I don't know what a "visionary" photograph is, exactly, as I'm not sure how you'd take a picture of a vision, but we talk about excellent photographs of various kinds a lot around here, all things considered. Want some links?

Mike

[MikeP]

All of this technical babble is totally irrelevant if the images are boring. If vision eludes anyone it is probably because they have not spent enough time actually looking at and truly absorbing images and their emotional impact. There are already way too many excellent images that are technically perfect in every way, but completely lacking a point of view from the photographer. A through knowledge of craft is essential to any artist who wishes to express him or her self. However one must get to the point where craft is the vehicle that one uses to express a point of view, or vision. Craft for craft's sake is a total bore. I'll take a Winogrand over an Adams any day.

[Max Hertelendy]

As pretty as the world is, I'd rather discuss technical details than visions in some cisrcumstances and places (when nobody openly asked for your visions, for example). Technicalities are useful, verifiable, they always work the same, and you can learn them from other people willing to explain them with just a little logic.
Visions... well, just go to any critique site and look what that mess looks like. I have my visions, would you like to hear about them? I'm not so sure (i'm not so sure I wanna share them, either, for a lot of reasons, and not all of them selfish). Is vision more important? yes, sure. Is it more easily and productively discussed in such a place? We can disagree wildly on this one.

[MikeP]

Point taken, technicalities are very useful and I am very much a tech geek myself, however not to the exclusion of vision. A balance is essential. I spent way too much of my life caught up in technique, now after 30 years of learning my craft I can finally make some interesting images. Or at least I find them interesting, I realize that's a point of conjecture however now I only shoot to satisfy myself when doing personal work. My commercial work still has a job to do, but I'm able to separate the two visually, technically and emotionally.

My response was to a seeming dismissal of vision as irrelevant. Trust me, I abhore those critique sites also. If I want an opinion I will seek out those who I respect and admire and get a literate and thoughtful response based on their experience, visual sophistication and knowledge of art and photography. For too many years I listened to the opinions of too many people who's motivations and qualifications were unknown, no more.

[Gingerbaker]

Todays DSLR's can already address many of the desires communicated by previous posters in this thread.

Want large format resolution? Take multiple shots and stitch them. I remember someone on the net posted a cathedral interior shot that was a pastiche of something like 250 individual frames. One could likely print the result at 10' x 10' with better resolution than any film format.

Want extended DOF, selective areas of focus a/or bokeh? Easy to do with a tripod, a light touch on the focus ring a/or f-stop, and a simple manipulation in Photoshop.

Same for expanding the dynamic range of a photo - combine multiple exposures using an automated luminosity mask in PS. Easy! [sort of! :D ]

Granted, not *every* photographic situation avails itself to these sorts of approaches. But any photo worth taking a little time to set up can use these techniques.

Any one of which has to be lot easier than manipulating a large frame film system, and then developing, scanning, and printing the result.

I think the fact that quite a few posters to this thread have asked for improvements which are already available, shows that it is difficult to keep up with the benefits of todays "digital craft".

Photoshop is a Frankenstein Monster - even its creators do not know the ends to which it is capable. :)

[Robert Roaldi]

What partly prompted me to write about equipment standards in the previous comments was something that I have seen here, in Luminous Landscape and other places. Someone writes comments about pixel level noise in a camera model or other, then adds that that noise would not be seen in any final print so it's not worth worrying about. Then forum contributors consume megabytes of disk space doing precisely that, worrying about it. In the end, do I know anymore than I did before? I don't know.

I fully understand the need to pursue excellence. Pushing manufacturers to do better for less money is perfectly fine, so far as it goes. I guess I am just a little sick of trying to decify what is meaningful and what isn't. I can't buy one of each model and test for myself.

Although we supposedly have great consumer choice, in my experience it is mostly a shelf space illusion. I recently decided to replace my 6-year old inkjet. There are dozens of models avalable, with relatively small differences in features that distinguish them. Would I be able to tell the difference between Canon's 5-colour model for $99CDN vs one of their 6-colour models for $199CDN. Who knows. I know that no store near me would let me bring both home for a week for me to test. So I bought the 6-colour one because of a test site that is linked from TOP (or was anyway). There are hundreds of equipment tests on the web and my final choice was made almost by flipping a coin. Would I have been equally served by the 5-colour $99CDN model? Maybe, but I will never know.

Thanks, Ctein, for the illuminating column. You said that my comments inspired you to write it; I guess my whining accomplished something then. :)