Digital Imaging Day Seybold San Francisco 2000
Moderator: John Henshall

Digital Imaging: State of the Art

Ladies and gentlemen, I don't know how to introduce him. Before, I said he's the digital incarnation of Ansel Adams. He's a perfectionist and tests things right to the limits. You need only look at that print from wherever you are in the room to realize that he took that right up that mountainside. That is absolutely phenomenal. Ladies and gentlemen, Stephen Johnson. [Applause]

Stephen Johnson:  I couldn't resist putting up the photograph of all of us up on top of Mt. St. Helens, which, I think, might also go to illustrate that something Carl was talking about all of the equipment that I am carrying. Although, in this particular case- Only the digital technology is just the head inside the view camera, the core going back and inside is a Dicomed digital camera with a PowerBook 540C on top. All the rest of this was just so we'd have food and water for the climb up to the top of St. Helens.

I want to try and do a few things that I don't normally try and do in one of these talks. That's show relatively few images, but talk a lot about where we've been and what I've been through and where I think we're going. Why don't you go ahead and put the light back on. Can we have the room lights up a little bit so that I can see you as I'm talking to you at this point?

I've spent the last eight months printing, and I've spent a lot of it printing work that I've made over the last six years. I'm not sure where this particular group is on what I've been up to, but in 1994, I quit using film and started using a digital scanning camera made by Mike Collette. That later became the Dicomed camera, and now I'm using his BetterLight Model 6000, actually the Super 6k is what I'm using these days, on a Sinar-X camera, as Carl referred to earlier.

The reason I've been so concentrating on printing these last few months is because I've been testing the Epson pigment printers, the Epson PhotoStylus 9500. The beauty of the introduction of the Epson 9000 last year, in 1999, was that we suddenly had a very large printer with extraordinary image quality. The huge downside is, we could look at it for awhile and we could start to watch it fade away, and that was very frustrating and has made the process of seeing the full value of this kind of printer be very limited.

When Epson told me late last year that they would be sending me a pigmented version of this printer, I was quite interested and excited. It arrived in February and we immediately started trying to profile it and make some prints, and I found out, as is true of most of these things in life,  it's a great, big overall story. It's just that; it's great, but the devil is in all of the details of getting some of these things to work.

It turns out that Epson has done quite a big thing making these pigments work, things that Roland had not done with their Hi-Fi Jet pigmented printer that had been available for a couple of years, or HP had not done with their pigmented inks that they had developed largely for signage prior to that.

Of course, the reason for the pigments is because, as image makers, we don't want to go to the trouble of making a beautiful print and then not be able to sell it because it's going to fade, or not be able to continue to enjoy it or put it up on a gallery wall because it's going to fade. The pigments that I'm using now, the testing has already indicated with great confidence, by Henry Wilhelm-- Henry, are you there? I'm told I can say without any doubt over 100 years on most any material, and on some materials, it's starting to look quite a bit longer than that. Is that a fair statement, Henry? OK.

Mr. Henshall:  He won't be around, will he, when we...

Mr. Johnson:  When we end, I think.

Mr. Henshall:  No, I mean in a hundred years.

Mr. Johnson:  Well, [let's have him] pickled in the meantime, John.

What that does is that takes me back to when I was working in the dark room with the gelatin silver print. I made a print that I knew, if I processed it and washed it correctly, it would literally last hundreds of years if anybody took the time to care for it. That was never true of any of our color work anywhere, anytime, any place, with any media. The only thing that even began to approach it was color carbon prints from the late '40s that, as it turned out, had varying severe surface-cracking problems over time, so that they didn't last either.

We were under the illusion for a long time that dye transfer materials were long-lived. We are now, shall we say, relieved of that illusion.

We were under the impression for a long time that Cibachromes were a long-lived material. We now find, through Henry's testing and others, that 29 years is about the most that we can expect these things to last.

This is not just a matter for the arts. There is an ethic riding through this entire photographic world that if you go to the time and trouble of making a print, it needs to stay around, not only in the fine arts, but in people's family histories. They're investing their whole family histories on color film-weddings, books, other family events-and then they're seeing these things fade over time.

How many of you have family histories from the '60s that kind of look psychedelic without having started out that way? [Laughter]  It's a shame; in fact, it's a crime. The fact that people bought the advertising lines that they should move from black and white, which was still largely the medium in the '50s, into color, when color was known not to last, is unforgivable. I've had people approach me in great sadness about things that aren't around any longer that they really wish were.

Now we have access to pigment printers, and Epson is certainly not the only pigments on the market. You go out and look for "pigment" and "inkjet" in Web search engines, and you'll find a lot of returns back. So far, I think, from my own experience and what I've seen and when I've talked to people, the Epson pigments are the ones that are getting the most attention, because they made them with the printer, a very high-quality, and they're designed to really go that printer.

That's the upside. The upside is we've got the permanence. I can put it on rag paper like the print of Mr. St. Helens in the back, and I've got a great deal of confidence from that. Now the real story: They're not very easy to use, at least in my experience. There are some strange qualities to some of the pigments.

First of all, we tried making a custom ColorSync profile, the first thing we do. Whenever I start with new ink and new paper, I print a profile target, I read that target with a spectrophotometer, I make a profile, and then I proceed to make a print. Well, for two solid months, I was finding that those profiles were just not very good and, in order to print, I was having to run color adjustment layers on top of my file just to skew the profile. I didn't want to go back and tune the profile, because I wanted to be working from the raw created profiles.

We then, through Mac Hobart at Nash Editions, Mark Radogna at Epson, and me and my good friend, Bruce Fraser, determined that first of all, my Gretag SpectraScan was not reading an error. Bruce and I did extensive tests, because we thought there was something wrong with my spectrophotometer. Then, we determined that if we take the ultraviolet filter off the spectrophotometer and put the polarizing filter on instead, suddenly we're getting a very different kind of reading that is more accurate of what the eye is seeing.

These pigments are encoated in resin for two reasons apparently, and Epson could certainly talk about it in more detail. They're coated in resin to apparently increase the apparent reflectivity of the pigment-because pigments have a history of having a very dull-looking surface in terms of surface reflection-and also to keep the pigments from sinking way down into the paper fiber.

Well, we're beginning to suspect that this resin coating of this ground-up planet earth that the pigments are made from is actually causing the spectrophotometers to have problems. We're getting a little big better pigment, or a little bit better profiles now with the polarizing filter, but there's another issue that comes up real quick, and this is starting to show up on a lot of the color management forms as well.

There is always a difference as to what a color print will look like, depending on the kind of light you're looking at it in. So you have to decide how you want to balance your print. I tend to try and balance my print for about 5,000 Kelvin, knowing that when people put artificial light on it, it's going to look warm, but knowing that when ambient daylight is coming through, it's going to look pretty good or maybe slightly cool, since daylight is really a higher color temperature than 5,000.

As it turns out, what people are starting to discover is that there is a more dramatic shift in the apparent color balance of these pigments than we're used to with dyes or traditional photographic materials. So, as I look in the back of the room now at that print of Mt. St. Helens, my eyes have so adjusted over these last few hours to the warm light of the tungsten lights that are in here, so that it doesn't look quite so pink to me as it did when I came in here. I balanced that print from 4,700 to 5,000 Kelvin, and I was going to use it as a dramatic example today of how it looks too pink under these lights but looks really good under daylight, and the fact that that difference is now exaggerated-greater than it's ever been before-is one of the other downsides that we're starting to look at and talk to Epson about in some seriousness to say, "Look, we've got an issue here, and we really need to figure it out."

We've got a lot of hype going on about the pigment printers, like the Stylus 2000, which is their desktop printer; their 9500, their very big 44-inch printer; their 7500 which is, I think, 24 inches. Don't think that for all of the upsides that are there-and they are considerable, they are real, and I'm delighted to have them-don't think that there aren't some downsides linked in with that too.

There are challenges that come with everyone of these new technologies. As we get a higher-res sensor, as I think Brian was pointing out, we've got to look at what the camera manufacturer actually does with that sensor. As Sinar Braun is moving this sensor around to get some higher resolution, we need to look at what is going on with the image quality in that. Certainly, it gives us the potential for a much more artifact-free kind of result, and that's good.

As Foveon builds their 3-chip camera with the prism, and we get three discreet images-one each for red, green, and blue-yes, we get rid of almost all of the traditional color artifacts that we have with a [inaudible] color camera that every other manufacturer is using for instant color capture, but they end up having some technological challenges that they have to get through in order to make that camera the best it can be.

Look at all these developments with a fairly high degree of skepticism, tempering that extraordinary enthusiasm that you're probably feeling. I don't have much negative to say about Photoshop 6.0. However, we've been very suspicious of one of the new features in it that, as you get the software and try it, I want you to look at it real hard. We think the problems I've been having with their new RGB soft-proof feature is directly related back to my use of the pigments. One of the features that Karen didn't get a chance to talk about is potentially extraordinarily useful, because if you have a profile of a printer, you can actually load that in now-and I'm talking about an RGB printer like the Epsons masquerade at being-that profile can be loaded into your screen preview, and essentially you have a soft proof on your screen predicting, or attempting to predict, how that image will print on your printer. If that works for you as it hasn't quite worked for me, that could be an extraordinary step forward for what Photoshop 6.0 brings to the market. So, do look at that as you're looking at Photoshop 6.0. It's one of the items in the View menu.

I think I've said everything I want to say about the pigment prints, although then a lot of that dullness that is still somewhat there in the reflectivity, disappears the minute you put the print under glass or Plexiglas, so it's less of a problem that I would have thought it was.

What I really wanted to talk about today and what I don't have a lot of time for, but I'm going to try, is where we're at in terms of starting to think about our sense of beauty and our sense of what is contained in a photographic print as we're moving now many years into the digital age in photography.

I've said many times that I would call this the first revolution in photography, the transition from silver to silicon. There obviously will be other sorts of materials other than silicon eventually as well. Most of what has been talked about and felt, as the impact of that, have to do with convenience kinds of features-the ability to see your photograph and re-purpose it and have it immediately. Those are all things that are great adjuncts of what this first revolution in photography has brought us. But I would argue that there is a much greater and more fundamental and broad implication of this revolution because of the kinds of image quality that is possible. What does digital imaging at its very best actually allow us to see and record?

We've talked a lot, even in these last few hours that I've been here today, about image quality, about color fidelity, a little bit about color management. I have to say that I think that, when we look back at the best of what film could do, one would have to admit that at some level film is in fact a reality-distortion field. It does not portray light the way your eyes see it. It makes shadows black and has all sorts of problems with many, many colors, dark greens especially. It's hyper-sensitive to blue. It introduces a granularity into everything you see. It is funky color, by any honest definition of it, despite the fact that we've grown to accept it.

In fact, we've grown to accept it-that's me going to sleep up here, so. As I said, I'm going to show less today and talk more-We've come to accept film's view of the world as being what the world looks like. All you have to do is take any photographic print you've ever made, hold it in your hand, and go back to the site that you shot it in. You will see that what you're holding in your hand is dramatically different from what was there: too contrasty, not enough dynamic range, inaccurate color. To the point now that Kodak's chief competitor in the film world, Fuji, has made a film that is so purposefully distortive, that people are buying it because they like the kind of surrealistic color that comes out of it.

I don't know what your own tastes go to, but I would contend that the best of what digital photography can teach us is that the way we've looked at the world is not only not the way the world is, but it's not the way photography has to be. We can, in fact, expand our dynamic range considerably so that the cameras actually see perhaps even more than our eyes can see. We can record color accurately and see color and hold it in subtle, beautiful, pastel ways. Not just this over-saturated reality-on-steroids kind of view of the world that we have been shown from film, especially in the last decade or two.

We can also record things beyond what our eyes can see. One of the images I am going to show you today real quick is is an example of an image that, in fact, records much more light than our eyes can see. I wonder if in fact I should try and lighten that a little bit? Well, I think, given time, I might not.

This is an experimental shot that I did with the BetterLight Model 6000 with all black and white sensors on them. This is an indoor scene up in here in the rafters of my studio, in bright sunlight, and on white cars. We actually shot this with three sensors simultaneously, with different gain on the three sensors so that we simulated a dynamic range of the sensor itself, much greater than the sensor was actually capable of holding.

This is holding 18 stops of information in this single photograph. I would like to say, Brian, that I'd really love to see a sensor that could do that, but I have to say at the same time, as we start to push the parameters of what the aesthetics of this new revolution in photography can bring, we're actually going to start looking at photographs differently. We won't initially. We'll keep trying to make digital images look like film, and in fact that's what almost all of the little cameras do. They over-saturate, they over-contrast, and they over-sharpen to try and compensate for things, and they deliver out something that looks as bad as a piece of film would or, because of the resolution, worse.

What do we need to do to get better functionality on the low end? At least lobby these camera manufacturers to develop an archive format for their files so that they don't step on the image quality that the sensor could produce by going ahead and processing them into some sort of final file with sharpening, color processing, contrast enhancement, saturation enhancement, and making the photograph distortive of what their sensor could be. Lobby them for that. Ask them for that. Kodak has done that for years with their DCS format. Leaf has done that for years. It's time that these smaller cameras do that as well.

This is especially true now that they're finally listening and giving us a TIFF Save option in the camera, but then what do they turn around and do on these 3-million pixel cameras? They completely process the 3-megabytes of data into almost a 10-megabyte file and then they save that and they take up almost your entire storage for the little camera and, in the process, you're very discouraged from using that kind of artifactless save.

If they saved it as an archive format, you would be able to hold it and take at least more advantage of what the sensor was able to see as their software evolved, rather than just their hardware and software combination.

I'm contending that we're making or moving into an era where we can see photography differently than we've ever seen it before. Now, if you go back and look at the print of St. Helens during our break, you'll see that there's both some saturation and some pastels. I think that one of the things that we've done over time is we've so skewed our vision of what a photographic print ought to look like toward what film can deliver, we're now in a very curious state of time where we're looking at the potential of what high-end digital imaging can deliver and we're trying to truncate it back to what film could look like.

That's what many of these camera manufacturers, using the 2 x 3-million sensor from Phillips, are designed to try and do. Make it look like film--over-saturate it a little bit, make it contrasty. So, don't think that the photographs that you make, as this age dies, have to look like prints that you've seen before. Look for what the potential is to grow in your vision as the capability of these sensors and cameras also grow. I think that there is something very exciting about that.

I think there is something more fundamental to the revolution of what digital imaging is than any of the convenience things, because it will literally change the way we see and record the world. It's clear it does with regards to showing us the image quickly. I think it's far less clear what it does in showing us the world in terms of the intrinsic beauty that is there and our new-found ability to capture it,. We really need to take that in, take a deep breath, look at what you think the camera can do, and then look and see what, in your heart, you want that photograph to be. Don't try and imitate the past. Look with your eyes and see what it is you really want to record.

In that, I think it's time for me to shut up and give the floor back to John, who's running overtime, I know, but give some thought to that and talk to me afterwards if you would like. I'm sorry I didn't show my images, but most of us wouldn't translate through the OCD projector anyway. Thanks a lot. [Applause]

Mr. Henshall:  Thank you, Steve. It's always good to have you aboard and really making you think, really making you look. Any questions for Steve? Coffee is waiting. Look, Fred's running.

Audience:  Steve, just a comment. The new Olympus camera that was shown on Thursday, the 4-megapixel camera that lets you store it as a raw file, off the sensor, so that's a first. The new Olympus 4-megapixel camera that I saw last Thursday...

Mr. Johnson:  Well, maybe they listened. I certainly complained loudly about getting the TIFF format and then not being able to save the archive. That's great news, Fred.

Audience:  I wish we had the Ofoto CEO here, but from your experience, what do you think these companies like Ofoto and others are going to do? Give us the saturated kind of look, or can we as consumers say, "Hey, we'd like a more pastel kind of look, a more... ." What was their kind of look? Are they going to be capable of doing that?

Mr. Johnson:  The capability is probably there if you manage to get something on the sensor or on the film that reflected that. They won't make those kinds of adjustments unless you start asking for them. Right now, most of those companies don't believe you want this.

A few years ago, I was told by Kodak film engineers that they made a color film that was much more accurate than many of the other films, and it was a bomb in the marketplace because it didn't distort things. I'm not sure that I'm going to be able to rally people to the cause of a little bit more reality in photography, but that's what I think I'm trying to accomplish and what I think I'm trying to say.

Ask for it if that's what you want. I know that the processing schemes and scripts that get called upon from most of these automated processing labs are definitely not skewed toward what I've been saying. They've been skewed to try and make it look like film, to [inaudible] the fact that the originals were digital, rather than take advantage of it.

Mr. Henshall:  Steve, thank you very much.