'In a Printer Mindset, the Screen is an Intermediate Tool'

Christer Almqvist wrote: "I plowed thru my computer files and found this: Ctein on digital printing:

Optimizing for the screen and optimizing for the print are two very different processes. The color spaces of a slide and of a print are quite different. What you see on the screen is not what you will see on your print, even using soft proofing. You need to process your files in a way that doesn’t always look good on screen but will be optimal in the print. In a printer mindset, the screen is an intermediate tool. Never get invested in what you see on the screen. I recommend you simply pull up an image you like on your screen, straight out of camera, and make basic edits and corrections. Just get the tone and color roughly correct; don’t try make it look perfect on screen. Print that photograph in A4/letter and let it lie a little. When you come back to it, note what you don’t like about it. Is the contrast wrong? Too dark? Too light? Does it need dodging or burning-in? Write notes on the print; use it as a learning tool. Go back to the computer and correct what doesn’t feel right on the print. Then print again, let it lie and re-evaluate it. After you’ve printed three or four of these, the picture won’t look good on screen but the print should be much better than the first version. As I said, with time and experience, you will be able to make perfect adjustments in just one or two iterations.

Mike adds: It's good to remind ourselves of this in the discussion about monitors and calibration. I'd venture to say that there comes a time in every printing process that you have to let go of any by-the-numbers approach that is only meant to get you close, and look at (and "feel into," as my ex-gf would say) the picture in front of you, and then give it what your best judgement says it needs—regardless of what the numbers say. I have a print I love of the "green light" before a thunderstorm, and if I "correct" it according to calibrated neutrality, it robs it of any sense of that greenish light. I know the green in the print is technically overdone, but in that particular image it captures the magical and elusive feel of that particular light very well. It looks and feels like the way I experience that light.

I think this happens to figurative artists too. They paint according to the reality of what they are looking at, or whatever their visual source is, but at some point they have to let go of the real world, or the pretext, and start looking at the painting and letting their sense of it dictate how they proceed. I know I personally use my well-developed sense of the real light in real scenes as a strong guide of where to take the print, but there often comes that time when I need to let go of my memory of the reality of the scene and look carefully and well at the print, take it as a thing in itself, and give it what I think it needs, or heighten the effect that I want to emphasize. That might entail a departure from the strict reality of the light or a literalistic description of a scene.

Mike

P.S. Just as an aside: Don't you wish Ctein would write a textbook on digital printing? I do. He demurred years ago because the hard- and software was too volatile and the products that would have to be discussed got out of date too quickly. But it seems like things have settled down sufficiently that perhaps a book might work. One solution to the problem of evolving products could be to get Epson or Canon to subsidize the book, which I visualize as being about the size of Canon's EF Lens Work III. Of course then the examples would be limited to that company's own products—but the upside of that is that then it could be up to the company to keep it updated in successive printings. Are you listening, Epson? (Ctein uses Epson printers, inks, and papers, so that would be the natural pairing.)

Featured Comments:

Keith B.: "There is no further elaboration of digital printing needed beyond what's in the Ctein quote. (Ctein was right that the world of software and hardware is too fugitive to make a book practical.) The calibrated monitor gets you closer than a junk monitor, but the operator almost always has to tweak and adjust to get the print to look the way he or she wants."

Rob de Loe: "I would say, 'Yes, but....' I think Ctein's advice is solid, but it's easy to go off the rails because of errors in one's printing process. One wrong setting in the journey to clicking the 'print' button can send you back to making all kinds of unneeded or inappropriate corrections to what you're seeing on the monitor. A case in point is a lot of fiddling to 'correct' what I saw in the print, when most of what I saw in the print was due to me forgetting to apply the right ICC for that paper. My friendly amendment to Ctein's advice is to add, 'Assuming that you're sure the print is set up properly....'"

Geoff Wittig: "I have never made a darkroom print, never personally developed a roll of film. My first printing experience involved using Hewlett-Packard's first generation slide scanner and Epson's first serious pigment printer, the 2000P, which had truly ghastly metamerism issues. Yet I am very fond of Ctein's book Post Exposure, which details his incredible darkroom chops and his philosophical approach to printing. His logical, methodical approach is a revelation. I would consider sacrificing a limb for the digital printing equivalent."

Comments

[Bob Johnston]

To make a meaningful photograph you have to show what you feel, not necessarily what you see.

[Rick Popham]

I would be the first in line to buy one, because I'm always interested in Ctein's take. I have some of his prints so I can see he knows what he's talking about.

A few years back, Jeff Schewe published two excellent books on the subject: "The Digital Negative" and "The Digital Print". "The Digital Negative" required a second edition because of updates to LR/Photoshop, etc., but not much has changed once you send that file to the printer.

[Rob Campbell]

But Mike, this is nothing new: theory, science and mathematics operate on a level removed from the artistic.

My printing interest (digital) was almost exclusively in the field of black and white. I bought an HP B 9180 A3+ machine a long time ago because it received a lot of favourable mention on behalf of its b/w printing capabilities. In the event, it proved to be fairly good on matt paper, and those prints improved considerably once inserted in transparent, archival sleeves, the applied gloss lifting (or at least lending the impression of that) the deadness of matt paper to something more closely resembling the benchmark WSG papers of old.

Sadly, just as the machine began to emit an unusual soundtrack, HP stopped its support. As bad, the genuine inks vanished, though as the machine was on the blink, ink supply became academic.

However, even when the machine was functioning flawlessly, the monitor vision and the printed version were quite different, and not, I believe, simply due to the basic difference between transmitted or reflected light viewing. I always had to tweak. But hey, it was the same in the darkroom: testing was always necessary with b/white, and if any of you have experienced printing colour, that was far more complex because there were more variables to understand. Eventually, much of that diagnosis became second-nature and quicker to handle: one knew the pathways a bit better.

Of course, this all adds to the futility of having third parties do the printing: you end up with either their version, or a mix of several inputs, neither pure. This isn’t only an amateur’s problem: for a while, I used to handle colour for the industrial unit where I began my career; I knew perfectly well the subtle difference that even a minimum of filtration change could create. Once I went out solo, I could not justify running a colour darkroom, so I was obliged to use commercial labs. I hated getting colour print shoots: I knew before I handed over the films that it would end up in the desperate position of the lab’s prints being “commercially acceptable”, which simply meant they would stop testing very quickly, too quickly. Of course, if you work to published price lists, it’s what you have to do, or lose money. Knowing that just one further filtration step would have fixed things was knowledge I might have been better off not having.

I guess printing is a route that I will never undertake again.

Wouldn’t it be nice if we could all produce 8”x10” colour transparencies instead…

[John Camp]

Ctein could also write the book as an e-book and put it up on Amazon. Since we're talking about complicated computer use, most users would intimately familiar with e-books. That would allow Ctein to update the electronic version over the years, without having to rewrite the whole book each time. Make some changes, withdraw the old book, release the new edition.

[Jeff]

I don’t disagree with any of this, and go through my own similarly iterative process with all my digital prints (and wet prints before that), but I still appreciate many of the benefits of ImagePrint software that I listed in another recent comment. The soft-proofing aspect, after using IP for a long time, just gets me in the initial ballpark more reliably. It’s the other benefits of IP… custom paper profiles, printer setting optimization, avoidance of Epson/Adobe/Apple interface issues, adjustment tools for final print rendering tweaks, etc…that make it worth the significant expense for me.

[Speed]

As photographers and printers we are not duplicating ... we are interpreting.

[MikeR]

I'm keen to see how Ctein works with your monochrome images. Will there be some exposition regarding paper stock, 3rd party printer drivers VS Epson's native one, or Epson's monochromatic one, which BTW I found a bit clunky to use? And of course, what he felt he needed to do with the digital original, such as global curve adjustments, local adjustments, things like that.

[John Holmes]

I find the same thing applies to making B&W images from raw files. Once I get the B&W image to look like I think it should, and convert it back to color, I often find the color image is terrible and totally unrealistic.

Hmmm, what does this say about B&W images ... that they are all fake, not real?

[So are you saying you think there are some photographs that are actual and real? --Mike]

[psu]

Back in 2009 there was a series of videos by Richard Benson that went with this exhibition

https://www.moma.org/calendar/exhibitions/309

And in one of them Benson talks about various issues related to digital controls in color and what not, and made the point that while they were all fine and good for certain applications, ultimately what you are after is to make a print that looks like you want it to. And as he says in the last video in the series: "f--- the controls."

I think this is generally a wise attitude.

[David Miller]

Regarding the prospect of a digital printing textbook by Ctein: I am standing by my Epson printer with hope in my heart and cash in my hand…

[Stan B.]

I wish I had the money and patience to invest in a printer, and yes, I realize that with regular usage, home printing is more economical (as well as satisfying to master)- but I literally don't have the space in a mini studio for a printer. So I've learned to live with a (very) small handful of commercial prints every year. BTW, as per a previous comment I made, "may" should have read... "will:"

"Granted, that's not a completely seamless transaction and some adjustment may be necessary, but once the technical parameters have been set to achieve the result you submitted, the resulting digital print is very much your own vision."

[John Holmes]

I think what we are saying is that pictures lie. I'm wondering in more depth as to what we can trust with images. With all of the capabilities with processing and use of AI, etc., I'm reaching the point of understanding that no image can be utilized to show what actually happened. I don't think I will believe in any picture or video anymore. Documentary photography is dead or dying.

[Andre Y]

Beyond color, the impression of resolution is very different between a print and what you see on screen. For one, a print uses every pixel that is sent to it, whereas a screen rendering is subject to whatever Microsoft, Apple, or Adobe feels is OK to reduce the resolution of a photo to fit on your screen. Often they prioritize speed over quality for downsizing and toss most of the photo away, and that makes sense for a computer since you're often browsing dozens if not hundreds or thousands of photos.

In other words, a photo on a screen is often a rough approximation of the data that's actually in the photo. Yes, you can zoom in to 1:1 and see individual pixels, but then the overall sense of the photo is lost and you never get to see the total effect of all those pixels on the entire photo.

Is that last point relevant given that we don't print very large and some photo data may not be visible on a print? I would argue that it is: I didn't fully appreciate how good the Nikkor 50/1.2S for their Z mount was until I printed a photo I made with the lens at 13x19. It looked good on screen: details were sharp and clear, but it looked mostly like any other good lens out there in terms of pixel quality. But when printed (on a Canon Pro-10 on Red River Palo Duro Soft Gloss, which is a remarkable clone of Canson's Platine), the photo took on a quality that was almost invisible on the screen. There was a sense of transparency to it, like it was a window onto the scene full of natural, unforced detail, which is not a quality that comes across almost any other lens I've shot.

I tell people now that they can't fully appreciate the strengths of the 50/1.2S unless they print, and I'm sure there are other optical treasures to be discovered if only people printed.

[Chris Kern]

I’m aware that Ctein is not an advocate of soft-proofing, but an alternative to his successive approximation approach is to calibrate a wide-gamut monitor carefully, use an ICC profile for your target paper stock (the ones supplied by the paper manufacturers have always worked well for me), and adjust the colors and tonality of the soft proof until they are as close as possible to your on-screen edits. Of course, an image displayed on a reflective medium will never look the same as it will on a transmissive medium since even a consumer-grade computer monitor has a considerably greater dynamic range than the highest quality photographic paper, but I’m rarely disappointed with the results when I print from my soft proofs and I almost never need to make a second print to get the effect I want.

Mike: Don't you wish Ctein would write a textbook on digital printing?

Jeff Schewe’s The Digital Print (Peachpit Press, 2014) is an excellent resource. While it is keyed to the use of Adobe products (Lightroom and Photoshop), it provides generalized information on printing technique that no doubt applies to other application software used for post-processing.

[Peter Conway]

I recently experienced what you're discussing, but kind of in reverse. Years ago I had taken a workshop with Charlie Cramer and Bill Atkinson, the whole focus of which was to produce a final digital print. I came out of the workshop with a print I was very happy with, and it has hung on a wall in our house ever since.

A few months back I decided to include the image in a web site I was putting together, and figured it would be easy since I had finished it in the workshop. I opened the workshop file on-screen, and was surprised by how terrible it looked. Looking at the print virtually every day for 15+ years, that is what was in my mind's eye, and the on-screen image really wasn't close. It took a fair amount of tweaking before I felt I had a JPEG which was reasonably faithful to the print.

[Darryl]

The talks by Richard Benson are still available online, the relevant one is the last in the series - The Future of Photography, available here: https://www.benson.readandnote.com/allvideo

[Rand Scott Adams]

Screen to print . . . a fascinating subject. Since color vision is very adaptive, a lot of us “don’t see” color casts and other weird things on screen. When we print we wonder, “Where the heck did that come from?” In teaching color management and fine art printing, I spend a lot of time teaching my students how to “See what it is they are seeing.”

And while I agree with you 100% that “correct by the numbers” is very often “incorrect for the vision of the image,” it’s always a good place to start. Use L a b values in Lightroom or Photoshop and find something that “should be neutral” in your image. Look at the a and b channels and bring them both to zero for that sampled area. “Then,” you know that you’re looking at neutral, neutrals! “From there” you can ask yourself, “Where does this need to go to realize my vision for this image."

Soft proofing will get you a long way down the road in terms of getting something you “like” on the screen onto paper. But soft proofing is as much art as it is science and learning what to do, while there are some general principles, takes time and patience and many iterations until one begins to understand how to get where you want to go.

Rand

[So very true. It almost seems to happen by magic...you just keep making prints, doing your best each time, and not only does it gradually get easier and easier, but the prints get better and better! --Mike]

[Mark Sampson]

Just like in the darkroom, it's an iterative process. The numbers may get you close... but in the end you must satisfy your eyes, however many attempts it takes.

[Jeff]

To Rob de Loe… you’ll never have that problem with ImagePrint (Black): pick a paper/size and ink choice, and the custom profile is automatically loaded and all printer settings are automatically optimized. Period. The rest is left up to you.

[Darin Boville]

I've never believed the soft proof looked like the print--how could it? Useful, but not the final point in the process.

On the other hand, Ctein's mention of stepping away should be given more emphasis: A few years back I was making a print--got as close as I could with the soft proof then went on to fine tuning the print. I'd make a version, not quite right, throw it on the floor, tweak the image, print again, not quite right, throw it on the floor. Round and round. Must have had twenty prints on the floor when I finally decided I was too frustrated to think straight.

Came back the next morning and scrutinized those rejects. I scrutinized the problem areas. I couldn't see the slightest difference.

[Chris Nicholls]

A book by Ctein? Well there was "Post Exposure - Advanced Techniques for the Photographic Printer" copyright 2011, by Ctein. I have a PDF file of that book which I must have found through TOP, back in 2011 or thereabouts. That pre-dates Ctein's shift to digital of course.

[Stephen Jenner]

I long ago decided that Photoshop and Lightroom were not only overpriced, they were more than I could afford. I am retired for more than twenty years due to the wicked inconvenience known as Crohn's disease.
I seem to remember that it was an Irishman called William Fagan, who alerted me via the excellent Macfilos (Mike Evans), to a program called "Iridient Developer"), .
I have never bothered myself with PS or Lightroom since.
Iridient is still under constant development by one man with his own distinctive vision, Brian Griffith. If one needs clarification on any aspect, an email and even a phone call connect with the man himself, rather than a help desk.
Since I and those who see my prints are happy with the results, I have no need to even gasp at the prices for what are essentially overpriced legacy programs.

[Christer Almqvist]

As we all know thanks to Ctein et al, it is normal that the print coming out of your printer will not look exactly like the image looked on your monitor. Taking a few very simple precautions will reduce the discrepancy problem greatly.

(One) Make sure the ambient light around your computer is more or less the same every time you process a file. Bright sunlight now shining into your study? Pull the curtains. &c.

(Two) Alway have the computer screen brightness set to the same value when processing image files.

Experiment a bit with these setting until you get a print that satisfies you for general viewing and note those setting for future use. Of course, you will still need to work on contrast and all the other variables, but you will have a firm basis to work from.

[Larry Gebhardt]

I agree with Ctein that you need to adjust based on prints. I don't see any way around that, unless we get reflective monitors. However I think monitor calibration can get you closer to right on your first print. I use an NEC PA301W and the most valuable feature to me is the SpectraView II software. I make a target for printing with a warmer white point (5000K works well in my house which has lighting on the warm side), a low intensity (80 cd/m^2), and most importantly a vastly reduced contrast ratio over the native monitor (200:1). That makes all images on screen kind of bland, but combined with soft proofing gives good color and more importantly contrast that's approximately right.

The reason the SpectraView software is so good for this is it will allow you to switch the monitor settings for each target (purpose) with just a change on screen. On my LG I can calibrate it like above, but if I want to edit for screen again it's a painful process with recalibration. Most monitors also require you to manually change the brightness, contrast, and other settings to get a display setup that works for printing. I don't know if any of the alternative monitors you found support a software like SpectrView II but I sure hope so.