Acutance Comparison of
Slide Film vs. Neg Film vs. Digital

I have rewritten much of this comparison to incorporate new information I have uncovered regarding the acutance of digital capture. My former take on digital was based on frames with in-camera sharpening turned on. Because in-camera sharpening is one-size-fits-all, it often leaves incipient haloes lurking in areas of fine detail.

The importance of acutance

Cameras are put to nearly as many different uses as there are different photographers. For many of those uses sharpness, or crispness of detail, is not important. Indeed, for portraiture, too much sharpness is detrimental. Sharpness is especially important for extreme enlargements of complex subject matter, such as the 16x24 prints of woodland scenes I love to make.

The word "sharpness", however, is ambiguous in photography. When I say an image is sharp do I mean that it is precisely focused or do I mean that very fine detail is discernible? The technical term acutance refers to the crispness of edges that largely forms our impression of the degree of focus of an image. The technical term resolution refers to the fineness of detail that can be seen.

In digital post processing we routinely use an Unsharp Mask filter (USM) to increase the apparent "sharpness" of an image. USM sacrifices resolution to add acutance. Unless we are creating images for the purpose of technical documentation, such as in forensic evidence or in surveillance, the perceived acutance of an image is at least as important as its actual detail resolution. If we are doing art, then in most cases perceived acutance is far more important than actual detail resolution. (Note: I am not saying that art demands high acutance, only that the acutance – whether high or low – is usually more important than the actual resolution.)

For the past ten years, as digital image capture has graduated from sub-megapixel image sizes to multi-megapixel image sizes, the overriding concern with this medium has been whether it has sufficient resolution to replace film for a given photographic task. For close-up portraiture and photojournalism, the required resolution has long been achieved. For group portraiture and much of landscape photography, the bar for resolution and – even more – for acutance is set much higher; although how high that bar gets raised varies greatly with how large a maximum print size is required.

Acutance compared

The following three images are small 100% crops from frames taken with three different media:

35mm Provia, the finest-grained 100 ISO slide film
35mm Reala, the finest-grained 100 ISO colour negative film
6 megapixel digital via the Nikon D70, at its lowest ISO: 200.

Provia is the film of choice for most professional photographers who still use film:

Fig 1. Provia (unretouched crop from a 4000 PPI scan)

Reala was originally developed by a Fuji research team to provide the fineness of detail required for achieving acceptable prints from the smaller APS film format:

Fig 2. Reala crop (unretouched crop from a 4000 PPI scan)

The Nikon D70 is widely renowned as having excellent resolution for a 6 megapixel dSLR:

Fig 3. D70 crop (very slightly sharpened crop from a 2000x3000 pixel frame taken with no in-camera sharpening*)

Each of these crops has been saved in Photoshop as Quality 11 JPEGs. These images are typical examples from my own collection. They do not represent carefully controlled test cases, but are typical of what I see over and over again in my own work. If you want a more carefully controlled test, I urge you to do it yourself. My purpose here is to motivate you to make that effort. ;)

Please feel free to save each of the above crops to your hard drive then attempt to sharpen them with USM or any other filter. .

In the film scans the fineness of detail (resolution) is not dramatically different; however, the acutance is significantly better in the Reala frame. In the Reala frame we can visually separate the green leaves and grass blades; their edges are reasonably distinct. In the Provia frame acutance is so low one would swear it was out of focus. Yet both frames were carefully focused and taken on a tripod with the same lens (35-70/2.8 zoom), at approximately the same focal length, at approximately the same distance from the subject, and with a large DOF aperture. Clearly, the Provia is going to require much more aggressive USM to achieve anything like an acceptable level of acutance than the Reala will require. And that aggressive USM is going to destroy a greater proportion of the Provia’s vaunted resolution.

The D70 crop differs from the film crops in that the D70 frame is only 2000x3000 pixels, while the film frames, scanned at 4000 PPI, are closer to 4000x6000 pixels. The D70 frame will require twice the amount of enlargement compared to the film scans to produce a given print size. However, its acutance is superior even to the Reala crop, which means that it can be given clean edges with very moderate amounts of USM compared to film.

Dynamic range

We can also see that dramatically greater dynamic range of the colour neg film over the slide film in these two crops. Both were shot on sunny mornings. Even though the Reala frame was exposed around 10 AM (with the sun higher in the sky) and the Provia frame was exposed around 7 AM, the fence post in the Provia frame is overexposed in the sunlit portion and underexposed in the shadow portion.

Good 12-bit digital, such as in the case of the D70, has at least another two full stops of dynamic range than slide film. The latitude is just as tight in the highlights, but with careful metering digital can handle nearly all existing light situations.

Grain and noise

Another urban myth is that slide film has inherently less grain/noise than neg film.

Fig 4. Provia sky vs. Reala sky vs. D70 sky

In my experience scanner noise is more directly proportionate to "grain" (dye cloud) size than to slide vs. neg. In general, the lower the ISO, the smaller the dye clouds and therefore the less the scanner noise. So far as I can see, the only reason the sky noise of Provia might be considered marginally less apparent than that of Reala, is because the acutance is less - the Reala noise is a tad sharper.

...But of course such distinctions are moot in the face of the absolute purity of the low ISO digital's complete lack of grain.

Conclusions

Slide film is strongly favoured by most pro and serious amateur photographers. Being able to see the un-reversed colours of the scene on the film itself is a critical advantage in high-volume production environments. Colour slide film also has a reputation of greatly exceeding colour neg film both in resolution and acutance that originates from the 1950s and ‘60s when Kodachrome 25 was king. We can see that the 100 ISO descendants of that superb film do not retain those kingly virtues. Instead, digital has come from behind to overtake both in acutance and grainlessness - and is fast catching up in resolution.

Note: for an excellent numeric preçis of film characteristics see Bill Tuthill’s Film Characteristics Table.

- - - - -