STILLWELL SLAYER OF DRAGONS

On his way for a break in Spain, digital expert, Ron Stillwell was anxious to purge a few myths and take a tilt at a few windmills.

Speaking to numerous social photographers over the last year around a third who have bought a digital camera are not happy with the results they are achieving. Some put this down to their lack of digital skills whilst others say 'digital simply isn't as good as using film' and carry on using film, with the occasional excursion to digital for rush jobs or as a supplement to film.

From discussions and articles in the press it appears most are focussed on comparing film and digital to prove which is best. Does it really matter? Market forces will ultimately determine the future, but the move toward the digital highway is relentless. What I believe is far more important is understanding the workings and full potential digital imaging has to offer. To do this properly first requires a separation of myths from facts.

Myth 1:

A 35mm colour neg contains many times more data than an equivalent digital SLR image file.

This is irrelevant! First of all, I regularly produce images (from a Fuji S2 Pro), which are enlarged to 30x20 inches and printed at the lab on photographic paper. They are sharp and 'grain-free' - something I have found impossible to achieve from film at this size, even with 120 film. Secondly, scanning say, a negative at very high resolution and 16-bit is pointless, except for all but mega enlargements, most of the extra data would be thrown away as print output is between 250 - 300 dpi. In the past I have used many film formats and it would take 5x4 film to produce the 'grain-free' tones of a good digital SLR, particularly in skies. With smaller size prints this is less of an issue with film.

So, the important point is not so much capturing an image with a huge colour depth and gamut that cannot be utilised, but fine-tuning an image to excel within the parameters of a given output device.

Myth 2:

Colour profiles - Adobe RGB (1998) gives better colour than sRGB.

In reality, no. First of all, do you know why the former is supposed to be better? When I ask this question the answer is usually, 'someone told me' but with no explanation. I run digital workshops plus a technical helpline via our website, and a large portion of problems revolve around colour profiles. Guess what's the main cause? Answer, using Adobe RGB. When the digital workflow is reworked (using sRGB) the results are better and the problems encountered are resolved. Let me explain why; the Adobe RGB profile was developed for CMYK printing to take advantage of the extended cyan/green printing capabilities of modern 4-colour printing presses. How much of your work is printed in this manner? I'm not talking about inkjet printing, but books and magazine work. I was a publisher of a magazine for over twenty years, ten of which were 'digital' - I have never had any problems using sRGB.The one time I scanned and profiled the images as Adobe RGB I got a call from the printers asking me to come and check the proofs as it appeared the reds in the images were down. The long and short of it was; any colour profiles embedded in images were removed/ignored when ripped at the film separation stage!

As photographers we use the RGB colour space (transmissive) to capture, view (computer monitor) and print (photographic/inkjet) images. Even when printing using inkjets the images are sent to the printer as RGB files in most cases and converted by the printer.

I have been a photographer for more than 35 years and one of my challenges has been to reproduce colour as accurately as possible. Since going totally digital this desire is no less important and it never ceases to amaze me the superb colour range one can achieve with a good digital camera and equally good postproduction technique using Photoshop. As much as anything the output device determines just how good an image will look.

Photographic paper has, and in my view, will always have both the largest and best colour gamut. Although most modern wet-process colour printers have the capabilities to recognise and utilise embedded colour profiles, this option is switched off. Instead labs rely on the printer's built-in calibration system. This is a practical solution and works well, provided the lab checks every few hours for 'drift', which does occur. So, if you use an embedded profile the chances are most high-end printers won't use it!

Modern digital cameras offer both options but the "native" profile is sRGB.This profile most closely resembles the colour characteristics of a CRT (Cathode Ray Tube) monitor. Again, the colour profile of photo paper is closest to sRGB. It is ironic that within the digital workflow, the item we rely on most to evaluate and adjust colour is the analogue CRT monitor.

Taken on Fuji S2

Myth 3:

A digital camera cannot capture as wide a tonal range as film.

Not so. A good digital camera with a low noise level is capable of capturing a greater tonal range than any film! Again, the final result is limited by, in this instance, the tonal gamut of the output paper. For example, the tonal range of images seen on these pages is less than those printed on photographic paper, or even an inkjet printer. Nevertheless it helps to start with an image that can be tonally manipulated to suit and take advantage of the output paper characteristics.

Perhaps the most extreme tonal condition commonly encountered is wedding photography. As a pro photographer I have shot over 1000 weddings, all taken on film. In recent years I have gone on weddings, shooting mainly candids using various digital cameras. Overall, I have found the Fuji S2 comes out tops because of its CCD chip, plus excellent colour and noise reduction tables. Look at the picture of the wedding guests - shot in to the sun and yet all the full tonal range is captured. Taken on film the sky would have ended up burnt out.

I have to say that the way to achieve a wide tonal range with a digital camera does require some basic disciplines; one, a CCD chip gives better results than a CMOS (higher signal, better signal-to-noise ratio). Two, the capture tonal density must be 'compressed' so the brightest highlights are around 210 not 255.The i Fuji on S2 mage is then adjusted in Photoshop using Adjustment Layers (very important) and the tonal range can be 'expanded' to suit the gamut of the output printer. This may sound complicated but it isn't - I have produced a workshop manual showing how the most contrasty situations can be 'tamed' with a good digital camera and results, undreamt of, can be obtained.

This contrast-reduction technique can be applied to any subject, particularly when fine detail needs to be seen in light areas of an image. The whole postproduction process can be automated in Photoshop, adjusting 100 images in a few minutes.

Myth 4:

There is little or no difference between sharpening an image in the camera or after capture in Photoshop.

Wrong! Sharpening performed in the camera or applied to the RGB channels after can alter the colour of literally millions of the pixels. The example seen here compares the eye in a portrait and shows the difference between RGB/Lab sharpening methods. Like colour and to Eye picture ne gamuts, sharpness is dependant on the setting and resolving power of the output device. In addition sharpness varies with image size - one sharpness setting may not suit all output sizes and resolution. For speed, where optimum image quality is not essential, and output prints are less than 10x8 inches, the sharpness could be applied in the camera.

This 'quick & dirty' approach works - most of the time. However, if the image has to be lightened then there is a good chance of increasing the amount of noise, particularly in the blue channel. If the image is not sharpened until Photoshop it is possible to obtain cleaner (and greater) sharpness without noise, by converting the image to Lab mode and using the Lightness channel.

Myth 5:

Digital cameras don't have the same exposure latitude as colour neg film.

They do, but in a different range of the tonal spectrum. Instead of being, say, plus/minus one stop it's all in the minus range. Contrary to what many may believe, more than one stop overexposure of a colour neg will lose highlight detail or, if printed to show highlight detail, midtone and shadow detail will be lost. The only way to get the full tonal range from such a neg would be skilful hand printing - and you know how much that costs, if you can find a lab that does it nowadays! Yes, digital capture is intolerant of overexposure, but it is far, far more forgiving of underexposure. Most of the problems I come across regarding exposure are because photographers expect to pick up a digital camera and use it as they would a (neg) film camera. Those used to working with trannie film have had less problems as they already have the disciplines necessary. Having said this, digital capture allows far greater image adjustment than trannie film, with the added advantage the exposure can be checked immediately after capture. Colour film has a fixed gamma whereas with a digital camera it can be altered. This means (roughly) that the tonal capture range can be varied to suit the subject range. The ability to do this with a digital camera is not well known by photographers, and I think it is one of its most useful benefits. Just how far the gamma range can be extended depends upon the points mentioned earlier. Through my workshops I get to see and test every pro digital SLR there is. The sequence of image tests I run push the cameras to their limits and reveal the best and worst. Price has no bearing on performance, in fact, some of the most expensive cameras produce the worst results in all but ideal shooting conditions!

Myth 6:

Using fill-flash with digital cameras is more likely to give overexposed images.

This can be completely eliminated if a number of simple rules are followed. First, choose a camera which is known to have an accurate flash sensor system. Of all the digital cameras I have seen and used at our workshop shootouts, Nikon and Fuji cameras are the best for consistent accuracy under all shooting conditions. By far the worst is Canon, with its ETTL system - it has frustrated Canon users for years. Several dedicated Canon users have come up with their own workarounds but none are straightforward. Nikon, on the other hand, have a superb TTL flash metering system, which is also incorporated in the Fuji S2.They (Nikon) have even improved on this with D-TTL, which is in their latest cameras, and the new Fuji S3.

However, no matter what camera is used, I have found, from many examples I have seen, that the main problem with overexposed flash pictures, is not caused by the flash! Look at the two pictures of the bridesmaid shown here; the one on the left appears to be over flashed, but in reality it is too much ambient light, caused by the camera metering being fooled by the darker-than-average background. Camera metering systems read most 'foliage' backgrounds as dark and overcompensates. The solution is simply to dial in minus one or more stops and the balance is perfect. Much of the criticism of the performance or capabilities of digital imaging has come from those who have dabbled or apply 'analogue thinking' and find the results - compared to film - less than what they expected. The transition to digital has not always been smooth, particularly when buying cameras, as many have relied on the 'badge name' as a guide to performance. Others have stuck to the same name because they have an expensive array of lenses, and changing to anotherbrand of camera would mean starting again.

Let's just stop for a moment and consider these facts: I still have a 30 year-old Nikon FM2, with some of the best lenses available at the time. Although somewhat battered, it takes better pictures now than when I bought it. Why? Because of the advances in film technology over the years. The same would be true of any top-notch film camera. But what about digital cameras? The opposite is true; from the moment is comes out of the box its performance remains the same (barring optics), apart from some minor firmware upgrades that may be available.

If ever there was a time to 'road test' before buying, it's when choosing a digital camera. The vital elements that create a good image can't be seen or perceived from brochures etc. Not all pro digital cameras are the same - some perform well in the studio but on location, under variable lighting conditions, are simply hopeless. For now, and the foreseeable future, digital SLR bodies should be seen as 'disposable cameras' to be replaced as advances in technology provide practical (not superficial) improvements in image quality. This means, from a business point of view, digital cameras should be treated as a consumable rather than an accounting asset. Considering the falling price of each new digital body, the cost, if amortised over 18 months would be less than £90 per month to keep up with each new major advance in digital camera technology.

Success in photography requires a mix of creative and technical skills, and one should never dominate the other. Digital imaging does require mastering and applying certain disciplines. Learn them, apply them and they will allow your creative side to reach new heights.

By the way, I have just field-tested the new Fuji S3 camera - if you want to see the results, including some 30-inch prints, come along to the Fuji stand at Focus on Imaging at the end of February - hope to see you there.

Fuji pic

Photo Quote: You've got to push yourself harder. You've got to start looking for pictures nobody else could take. You've got to take the tools you have and probe deeper. - William Albert Allard