15
\$\begingroup\$

I only want to capture images with the intent of displaying them on a digital screen. I do not want to print the images, ever. Do any specific considerations need to be made upon capturing the image or during post processing beyond these three things:

  • Screen color calibration/profiling
  • Sharpening
  • Resolution and display pixel density

I bring up this question because I feel like film photographers are more keen at producing work for their intended output format, and I would like to know what a strictly digital based workflow would require.

Improve this question
\$\endgroup\$

3 Answers 3

Reset to default
10
+350
\$\begingroup\$

Additive vs Subtractive Color Space

One thing to think about is that an additive color space (RGB) is different from a subtractive color space (CMYK or others). There are colors you can display that you cannot print, and there are colors that you can print that you cannot display.

enter image description here

Notice how some yellows will print, but are not displayable, yet some greens are displayable, but not printable.

Blacks

Black is another really good example of a color that prints very differently that it displays, and displays differently depending on what device it displays on. Black on a CRT is the absence of light. Black on a cheap LCD is just very dark. (This is because an LCD is backlit and the back lighting is always on. Even when the cells are turned opaque some light leaks through.) You would think that this would mean a CRT would better contrast but in practice LCDs can get much brighter than CRTs.

Glow

Another aspect of displayed images is that they are luminous. A printed image needs room lighting reflecting off of it to be viewed. (And of course the color of the room lighting matters!) Consider the same image printed on paper and printed on transparency, a slide. Now illuminate the back of the slide. I suspect the slide will appear more vivid, since the brighter areas will glow.

This is a perceptual change, one that you probably can't account for while shooting, but I don't know. For example, I know this photo of mine has far greater impact displayed rather than printed:

enter image description here

Non-Traditional Inks

Another area is that printed images can have non-traditional inks, say a varnish or a foil. This would be exceptional difficult to capture on a display. (This is also getting way out of the realm of photography and into commercial printing, but it is worth noting in such an interesting question!)

Aspect Ratio

If you sell you work matted and framed it is prudent to restrict yourself to a small number of aspect ratios and sizes. This avoids an annoying, and expensive, inventory control problem. For example, I only sell 11x14, 10x20, and 12x12 images. (And I'm phasing out 12x12!) This restriction sometimes impacts your artistically. Your subject might not fit into either ratio but you are essentially forced to move to one of those ratios. This also changes the way you shoot. I constantly frame the shot and then widen it some, which is a challenge since I almost also shoot wide angle, say in the 20mm range on a crop body, and widening it sometimes isn't possible. I do this so there is some extra image in case I need to crop it just so.

This is not the case with displayed images. You are free to crop each image exactly like your vision tells you to crop it, which means you are free to shoot it exactly like you want to shoot it. However, almost all displays are landscape, not portrait, and if you are shooting for a display you'll probably find yourself avoiding portrait shots.

Future Proofing

From a future proofing standpoint I think you would need to say the image and the post-processing steps in a non-destructive and open format so that future display technologies could re-render your image. That's pretty pie-in-the-sky thinking though.

One thing you can do is shoot in Adobe RGB, not sRGB, just so you have a wider color space.

Improve this answer
\$\endgroup\$
6
  • 1
    \$\begingroup\$ Because of the differences between additive and subtractive color spaces, I always explain that their color space is different, and not necessarily bigger or smaller -- after all, if the color can't be displayed, it doesn't much matter if it's bigger/smaller! \$\endgroup\$
    – Dan Wolfgang
    Commented Apr 17, 2012 at 12:51
  • 3
    \$\begingroup\$ I really like the point about luminous images. As with slides vs. prints, that's a big deal. I'm not sure I agree with the color space parts, though. The difference in gamut is more an implementation detail than an inherent difference. As display technology improves, it's likely to exceed print. On the other hand, for now (mid 2012), the vast, vast majority of digital audiences will be worse served by Adobe RGB, which as you can see from your chart is really designed to better cover print gamuts. \$\endgroup\$
    – mattdm
    Commented Apr 17, 2012 at 16:57
  • \$\begingroup\$ oh, convert to sRGB when displaying on those media which assume sRGB, like Facebook. But you want to capture as much data range as possible. \$\endgroup\$
    – Paul Cezanne
    Commented Apr 17, 2012 at 18:25
  • \$\begingroup\$ @Paul, ultimately, when displaying on a monitor, the image will have to be converted to RGB. The only reason to use another colour space if the pictures are never going to be printed would be for better results in post, but if that's what you're after, shoot raw and edit in 16 bit colour ... \$\endgroup\$
    – SoftMemes
    Commented Apr 18, 2012 at 12:57
  • 1
    \$\begingroup\$ But what RGB Freed? We have regular and wide gamut RGB monitors now, who knows what we'll have in the future. And yes, RAW is almost always the answer. I don't know when I last shot jpg. \$\endgroup\$
    – Paul Cezanne
    Commented Apr 18, 2012 at 13:20
7
\$\begingroup\$

On the photography, nothing has to change. You should still photograph with the intension of capturing all the details and dynamic-range your camera can.

Of course, you could get sloppy and capture less on purpose knowing your display cannot show some amount of DR or some color but you would be restricting yourself to what your present (target) display can do. It would be like shooting a picture with a little camera shake because you want to print on watercolor paper only. It makes little sense.

Where things would differ is in the RAW-conversion stage - which would obviously force you to shoot RAW to get something beyond what a JPEG can hold. Many displays have wider gamut than sRGB or AdobeRGB and you can create a profile for a wide-gamut display and map your images to that color-space, with the display set to native color-space. This would let you show colors which are not printable.

The dynamic range of displays is far greater than that of paper. Again, there is nothing to change at the capture stage if there is no clipping but when you map this you have to be more careful making curve adjustments. Otherwise, artifacts may appear in the stops which are not visible on paper.

The one place where displays are currently less demanding is in terms of resolution. Even the discontinued IBM T221 display only had 204 DPI while the top prints can do better. Yet these things will come back and eventually be surpassed, so there is no reason to not capture as much resolution as the camera allows.

Improve this answer
\$\endgroup\$
2
  • 1
    \$\begingroup\$ On the last point: the iPad 3 is 264ppi, and hopefully that trend will take hold sooner rather than later. \$\endgroup\$
    – mattdm
    Commented Apr 17, 2012 at 17:00
  • \$\begingroup\$ @mattdm - Not only that, but the iPad 3(new) also covers 99% of sRGB-Rec.709 Standard Color Gamut. The iPhone 4/4s has a display with 326ppi too! Once you use the new iPad display for a few hours, it is hard to look at any other display. \$\endgroup\$
    – dpollitt
    Commented Apr 19, 2012 at 20:37
6
\$\begingroup\$

Screen color calibration/profiling

Are you showing the images, or are others viewing them on their devices?

While you should calibrate your workflow, consumer equipment is notoriously uncalibrated and sometimes just terrible. The best stab you have is: make sure its sRGB, and has enough appeal to "pop" on less-ideal displays. Not everyone will have your wide gamut monitor.

Sharpening

I find you can sharpen quite a bit in smaller images without artifacts, which leads to...

Resolution and display pixel density

You can turn some very-unsharp images into good looking ones, since resolution is very limited (unless you plan on allowing users to zoom in and pan about, like in panoramas)

Here are some common resolutions:

  • 1366x768 (terrible laptops)
  • 1920x1080 Full-HD (lots of desktop monitors, slightly less terrible laptops)
  • 1024x768 (iPad)
  • 2048x1536 ("the new" iPad)
  • 2560x1440 (27" iMac or similar monitors)
  • 2560x1600 (the lucky ones)

As you can see, 16:9 dominates, so design your images with that in mind. Your largest dimension could be about 1500 and be very usable.

Patience is also a key for web images - the viewer's that is. If you have nothing but quality 100 3MB JPEGs, they are not going to view many due to loading times. I know it hurts, but even quality 60 is ok for the web. I mean, look at what garbage Facebook churns out of their image hosting.

Improve this answer
\$\endgroup\$
5
  • 2
    \$\begingroup\$ My question was asking about considerations beyond those three things. I already understand those and I was trying to go beyond those basics. \$\endgroup\$
    – dpollitt
    Commented Mar 16, 2012 at 0:41
  • 1
    \$\begingroup\$ @dpollitt: Attention to detail escaped me today. \$\endgroup\$
    – Yann Ramin
    Commented Mar 16, 2012 at 3:23
  • 1
    \$\begingroup\$ I think aspect ratio is a great point — which brings up another. Are you planning your photos for digital display now, or digital display now and also in twenty years? \$\endgroup\$
    – mattdm
    Commented Mar 16, 2012 at 3:29
  • \$\begingroup\$ @mattdm - I was thinking now but if you have a crystal ball and want to give it a shot, I'm all ears :) \$\endgroup\$
    – dpollitt
    Commented Mar 16, 2012 at 3:37
  • \$\begingroup\$ Maybe this is me, but why is it important to work to the resolution of the monitor? eg. you state "Your largest dimension could be about 1500 and be very usable." Why would you not work to 3000 and have it 100% zoomed out? \$\endgroup\$
    – Rob
    Commented Apr 17, 2012 at 11:42

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.