UNDERSTANDING DIGITAL EXPOSURE:

Part 1: Exposure Reduction for Highlight Retention
Part 2: Digital Exposure & Noise
Part 3: When Blocked-up Shadows Aren’t Really
Part 4: Take a Balanced Approach to White Balance
Part 5: Dynamic Range
Part 6: Extending the Tonal Range
Part 7: What’s the Real Difference Between RAW and JPEG?

GENERAL:

1) COST-EFFECTIVE PHOTOGRAPHY
2) CONTROL THE RANGE OF FOCUS
3) IMAGE INTERPOLATION
4) LOSE THE DEAD SHOTS
5) SCANNING 35MM FILM
6) THE RAW vs. JPEG DEBATE…
7) IS PHOTOGRAPHY EASY?
8) MUST EVERYTHING HAVE AN ADOBE SLANT?
9) A CAMERA TO PAINT WITH
10) WHAT'S THIS COMPOSITION THING ABOUT?

Understanding Digital Exposure

3: WHEN BLOCKED-UP SHADOWS AREN'T REALLY

Digital is different to film. Thankfully!

There will always be different approaches to exposure. It could be you’re walking along shooting as you go – street photography perhaps. Or maybe you’re working with the camera on a tripod.

Talk to different photographers about digital exposure and you’ll probably hear several approaches to technique. But I think we can safely generalise. Each time the shutter fires we’d like to capture as wide a range of tones as possible. We want the histogram to look pretty good. We don't want to drastically overexpose shadow detail and mid-tones and we'd like to hold the brighter tones.

However, digital sensors are none too impressive at holding highlight detail. Hmmm… Here’s a problem straight away, especially if we’re casually handholding the camera and finding few opportunities to check the histogram and take the shot again.

Camera manufacturers have plenty to say about modern in-camera processing routines that extend dynamic range, notably in the highlights. But are they overly optimistic about their "14-bit A/D converters" and 16-bit processing? Users' real-world experiences tend to be mixed. We shouldn't allow ourselves to be soft-soaped by the manufacturers' marketing spiel.

Have you bought one of the lastest DSLRs? If so, what are your images telling you? How much tonal range is your new camera delivering in RAW and JPEG files?

If you leave your new (or not so new) SLR on auto there’s a good chance you’ll probably blow significant highlights sooner or later or plunge mid-tones into the shadows. If you’ve used film for years you may choose to lock onto a mid-tone and recompose. But, again, digital’s sensitivity to highlights means you may damage some of the brighter tones your image needs.

Like it or not, it’s the brightest parts of the image/print that first catch the viewer’s eye. If those zones are significant and have lost all or most of the detail, then the image is a lot less than it could be. Really, when all is said and done, an image that’s lost too much highlight detail is usually unacceptable.

Is there a solution? Well, I’d recommend exposure reduction so that the significant highlights will usually hold. What’s the best way to go about this?

When handholding we may lock the exposure on part of the scene that’s obviously a bit brighter than a mid-tone. This will take practice. Then we compose the shot with the preferred exposure locked in. The camera may be set to underexpose slightly all the time so we know that those quicker grab shots will probably favour the significant highlights. It’s just insurance. My digital cameras are now set to underexpose everything very slightly – 1/3 of a stop. With image-editing in mind, I’ve found that’s helpful in many situations.

If the camera’s on a tripod, I’ll usually make sure the lowest ISO is selected, compose, focus, shoot and then check the histogram to get a good idea of what the best exposure should be.

But if we’re doing all we can to hold on to the highlights, we need to keep an eye on the shadow details. We don’t want to lose them. Of course, there are times when we can use several exposures, graduated filters and polarisers to help us capture a much wider range of tones.

If we accommodate the important highlights, as we should, it could be that shadow detail slips too close to the edge of the histogram, or possibly just off it altogether. But the good news is, an image straight from a digital camera can be significantly enhanced to open up shadow tones and so produce a more acceptable and realistic result.

This becomes particularly important when we realise that a print will further compress the tonal range. Reflective paper can’t possibly hold on to all the detail we see on the screen. It’s very likely that many of our digital images can be significantly improved in image-editing software. This means an image that was exposed with the highlights in mind would probably benefit from shadow detail expansion before it’s printed.

Every now and then I still come across the term “blocked-up shadows” to describe the effects of digital underexposure. You can bet your bottom dollar that the author used to shoot slide film! But happily, digital is different, so maybe the terminology isn’t appropriate.

Because it’s digital we’re dealing with rather than film, we can get a surprising amount of detail out of the shadows. If some of my images, like the more extreme example below, were slides viewed on a light bench, I’d be up the creek without a paddle! I’d say, “Oh no! Key shadows are blocked-up. There’s nothing to be done.” And I’d probably toss them in the bin. However, with digital there’s remarkable scope for improvement.

Recently I was reading through a new book by a talented and knowledgeable photographer. He’s even been on TV! He offered the following advice: “Exposure meters work best measuring from mid-tones…” He went on to say, “Whether an image is properly exposed or not depends on the type of image you want to create. You will have a key tone… that must look right, so expose for that. The rest of the image can fall where it will.”

But part of that creative process should include image manipulation, a point acknowledged by this photographer on the previous page. Not wishing to labour the point, but digital is different! With digital it’s a good idea to capture as many tones as possible. Later, at the computer, you can spend some time looking it over on-screen. Then using the tools available you can decide to sacrifice or enhance tones until you get “the type of image you want to create”.

Here’s what we should remember:

1) Sensors don’t handle highlights very well.
2) DSLRs in particular are very good at holding shadow detail at lower ISO settings.
3) Try to expose so that you favour the lighter tones (a slight reduction in exposure).
4) If necessary mid- and darker tones can be opened up very successfully in software.

An image that lacks contrast is usually less appealing, so we normally won’t want too much shadow detail throughout the frame. But it doesn’t make sense to reduce our options for creativity by making poor exposure choices that damage highlight detail.

The image below was captured by a 5D with a 17-40mm lens at ISO 200 and edited in Curves adjustment layers in Corel PHOTO-PAINT. The top half of a lifted curve was anchored in place so that the shadows could be adjusted on their own.

Enlarged detail shows (below) how effective the Curves Tool has been. Localised histograms show how the tones have been enhanced.

Please don’t let that slightly ragged histogram on the right bother you (main image). We must bear in mind this is a more extreme example specifically chosen to illustrate the power available in software to open up the shadows, so what we’re seeing here is really very good indeed. Very!

A less than perfect histogram is no indication that the image itself will visibly suffer. In this case it’s confined to the darker tones I’ve specifically targeted by the Curves Tool. Generally, if your histogram flattens down after really heavy editing you’ll find there’s been a significant loss of highlight and shadow detail. A badly combed histogram (lots of missing bits) will probably show banding where tones gradually blend.

This DSLR JPEG original was underexposed to the extent that shadow detail has appeared to clip – it’s actually dropped off the histogram altogether. But, there was enough data there all the time to usefully enhance the shot. A second Curves layer was used to add artificial tone to the blown highlights top-right – the extreme top-right of the curve was dropped vertically very slightly. The histogram for the finished image shows that the blown highlights appear to have been held, but there’s actually no more visible information in this area than at the start.

This simple adjustment would certainly help the look of the print, not that I’ll be making one of this! Besides, it still looks a little too dull overall. But the point is it shows how flexible digital images can be in the shadows.

Another layer was added to minimise noise in the bottom fifth of the image were underexposure was too severe. ISO 200 is largely to blame for this. The erasure tool mopped up what wasn’t needed. The paint brush tool applied on the mask adjustment layer would have done the same job more accurately in a more exacting context .

Total time spent: 20 minutes.

So it’s very likely that what appears to be “blocked-up shadows” in your DSLR images can be subtly opened up to give a more balanced result. With this in mind you can usually expose for those significant lighter tones in your images.

If all of this seems too advanced I’d suggest it’s actually not that difficult. In my opinion, the second half of digital imaging is image-manipulation, so we would all benefit from software that allows us to fundamentally manipulate our images.

Practise working with a layers file. In the representation above two layers have been created for manipulation (called "objects" in my software package). They lie over the main background image. The first is a reduced exposure of the coastal scene and will be adjusted to accurately simulate a graduated neutral density filter. The second layer on the top is a precise selection made from the main background image. It can be edited later to enhance detail, colour and tone on the cliff. The layer currently being edited (the active layer) is always outlined in red. Any layer can be hidden by clicking on its eye. If necessary many selections can be made with Mask Tools. After manipulation the layers can be combined into one image and saved under a new name. The above graphic contained 13 stacked layers or objects before being combined and saved as a JPEG for this web page.

Try this: open your image and create a Curves adjustment layer. It will sit exactly over the original.

It’s important to realise at the start that targeted editing using mask tools can be more useful than global adjustments because one way or another you will probably need to exclude areas of the image that won't benefit from manipulation. It's possible to apply software manipulation to the entire image that turns out to be helpful in some areas but damaging to others. This is one reason why I’m not keen on trying to find editing solutions using RAW-editing software only.

Play with the curve’s shape to see what it does. (Out of interest, try Red, Green and Blue curves separately.) Afterwards, if necessary, you can erase what you don’t need in each layer, or try feathering, or adjust its transparency, or refer to the program’s Help text to see how to use the Paint Tool to remove or bring back parts of the image you’re editing.

As well as applying curves adjustments you can select areas of an image and copy them to other layers to correct localised colour balance problems, or reduce noise, etc. Remember though that using layers increases file size substantially so you'll need plenty of computer memory.

All things considered then, it’s clear that digital exposure usually needs intelligent input before we press the shutter button and then afterwards in image-editing software.

home | Personal Pages