Testing the Coolscan’s Optical Resolution With a USAF Target
A Nikon Coolscan can produce remarkably detailed scans from 35mm negatives and slides, but its advertised dots-per-inch figure does not tell the whole story. The useful question is how much fine detail the scanner can separate with acceptable contrast. A USAF 1951 resolution target provides a practical way to examine that performance without relying solely on manufacturer specifications or a visually sharp-looking scan.
The test is especially useful when comparing older models such as the Coolscan III, LS-40, LS-4000, LS-50 and LS-5000. It can reveal whether a scanner is resolving close to its nominal optical resolution, whether focus is accurate, and whether film alignment, dust, vibration or software sharpening is affecting the result.
For Australian hobbyists, this kind of measurement can also make second-hand buying less of a gamble. A scanner found through Gumtree, eBay Australia or a camera shop in Melbourne may have a good transport mechanism but tired LEDs, a misaligned film holder or a lens that no longer focuses properly. A controlled target scan gives you evidence before committing time and money.
What The USAF Target Actually Measures
The USAF 1951 target contains groups of black and white bars arranged at progressively finer sizes. Each group has six elements, and each element contains three lines in one direction and three in the opposite direction. As the lines become narrower, they eventually merge into grey, lose contrast or disappear.
The smallest clearly separated pattern provides an estimate of limiting resolution, usually expressed in line pairs per millimetre. One line pair consists of one dark line and one light line. This is different from pixels per inch, because optical resolution depends on contrast, lens quality, focus and the scanner’s sampling system.
A target scan is therefore best treated as a comparative measurement rather than an absolute laboratory result. A Coolscan may technically sample at 4000 dpi while delivering useful detail below that level. The scan can still be excellent, but the finest recorded pixels may represent blur, grain or sharpening artefacts rather than additional image information.
The target must be positioned in the same film plane as an ordinary negative or mounted slide. A printed target placed directly on the scanner glass tests a different optical path and says little about the film holder, autofocus system or film-channel alignment. For the most meaningful result, use a calibrated transparency or a high-quality USAF target designed for photographic resolution testing.
Preparing A Repeatable Coolscan Test
Clean the scanner’s film path, holder and target before beginning. Dust is particularly troublesome on dense black bars because a small white particle can make a line appear broken. Australia’s dry inland air can encourage static and airborne dust, while coastal areas such as Sydney, Brisbane and Perth can add humidity and fine residue. Handle the target by its edges and use a blower rather than touching the emulsion or printed surface.
Disable settings that can disguise the scanner’s true resolving power. Turn off Digital ICE, ROC, GEM, grain reduction and automatic sharpening for the baseline scan. Nikon Scan’s colour management and exposure controls can remain active, but record every setting. If VueScan is used instead, save the input profile, bit depth, infrared options and sharpening state.
Useful checks before the first scan include:
- Confirming the correct film holder and film orientation
- Cleaning the target and scanner aperture with suitable tools
- Allowing the scanner to warm up consistently
- Selecting 100 percent, 200 percent or 400 percent viewing without resampling
Scan the target at the Coolscan’s nominal optical settings, such as 2700 dpi for an LS-30 or 4000 dpi for an LS-4000, LS-5000 or LS-9000. Also make a lower-resolution scan for comparison. Save a 16-bit TIFF where practical, because JPEG compression can create false edges and obscure the point at which the bars disappear.
Run at least three scans without moving the target. Then repeat the test after turning the target around or placing it in another holder position. If the result changes significantly, the issue may be holder flatness, film curvature, autofocus behaviour or uneven alignment rather than the scanner’s basic resolving ability.
Comparing Nominal And Visible Resolution
Read the USAF target at native pixel size rather than fitting the entire image to a browser window. Find the last element where all six bars remain distinguishable and retain reasonable separation between dark and light areas. A line that can only be guessed from a faint tonal change should not be counted as resolved.
The approximate spatial frequency of a USAF 1951 target element can be calculated from its group and element numbers. For the standard target, the frequency in line pairs per millimetre is commonly represented as 2 raised to the power of the group number plus the element number minus seven. The exact chart format should be checked, because some targets use different labelling conventions.
The following comparison gives a practical interpretation for common Coolscan settings. These are working ranges rather than guaranteed specifications; a sharp target, clean optics and accurate focus may produce better results, while an aged scanner or soft target may produce less.
| Coolscan Setting | Approximate Sampling Pitch | Useful Target Result | Practical Interpretation |
|---|---|---|---|
| 2700 dpi | About 9.4 micrometres per sample | Around 45–55 lp/mm | Typical of a well-focused LS-30 or LS-2000 |
| 4000 dpi | About 6.35 micrometres per sample | Around 55–70 lp/mm | Strong result from a healthy LS-4000 or LS-5000 |
| 4000 dpi with soft focus | About 6.35 micrometres per sample | Around 35–50 lp/mm | Check film height, autofocus and holder alignment |
| 4000 dpi with heavy sharpening | About 6.35 micrometres per sample | Apparent detail may exceed true detail | Confirm with an unsharpened scan |
| Reduced output resolution | Larger effective sample size | Lower visible frequency | Useful for workflow comparisons, not optical diagnosis |
A 35mm frame scanned at 4000 dpi is roughly 5500 pixels across, although the exact dimensions vary with cropping and the scanner model. That file size does not mean the scanner resolves every pixel as separate photographic detail. The USAF target helps distinguish genuine line separation from interpolation and edge enhancement.
It is worth recording both the finest visible element and the contrast at that element. Two scanners may resolve the same pattern, yet one may show strong black-and-white separation while the other produces a weak, mushy result. The first scanner will generally preserve more usable texture in foliage, hair and fine architectural detail.
Controlling Focus, Film And Software
Focus is often the largest variable in a Coolscan resolution test. Nikon scanners focus through the film base, and the sharpest plane can vary with film thickness, curl and mount height. A mounted slide may sit at a different height from a strip of unmounted negative. If the scanner supports autofocus, compare autofocus with a manually selected focus point when the software permits it.
Film flatness matters as well. A curled negative can place one side of the USAF target outside the depth of focus, producing an uneven result across the frame. Holders that are warped, cracked or missing their anti-Newton-ring glass may cause inconsistent sharpness. This is one reason a single central crop should not be treated as the whole story.
Common causes of misleading results include:
- Digital ICE or grain reduction smoothing the narrow bars
- Excessive unsharp mask creating bright halos around lines
- A target copied from a low-resolution file
- Film grain being mistaken for fine resolution
- Cropping or scaling software resampling the scan
Film stock changes the test too. A slow, fine-grained slide film can show a cleaner target than a fast colour negative, even when both are scanned perfectly. Dense transparency film may also challenge the scanner’s dynamic range, while an underexposed negative can lose bar contrast through the film itself.
For a fair comparison, use the same target, holder, orientation, scan mode and file format. Compare scans in unedited form first, then apply identical sharpening to separate optical performance from output processing. If you are testing a scanner bought in Adelaide or shipped across Australia, repeat the test after transport and allow the mechanism to settle at room temperature.
Turning The Result Into Better Film Scans
A resolution target is most valuable when it explains real photographs. After measuring the USAF pattern, scan a frame containing fine grass, distant branches, brickwork, fabric or lettering. These subjects contain irregular detail that behaves differently from perfectly spaced test bars. A scanner that performs well on both the target and real images is more likely to produce convincing fine texture.
Compare the edge character of the scan. Genuine optical detail usually appears with natural transitions and stable texture. Sharpened or digitally enhanced detail often has bright outlines, dark fringes or a brittle appearance. Nikon Scan’s unsharp mask can make a modestly resolved image look impressive at first glance, but it cannot restore detail that the lens and sensor failed to capture.
The result can also guide the choice of scanning resolution. If a 4000 dpi scan resolves only modestly more target detail than a 2700 dpi scan, the larger file may still help with cropping, dust removal and output flexibility, but it will not necessarily contain proportionally more texture. A good 2700 dpi scan from a healthy Coolscan can be preferable to a poorly focused 4000 dpi scan.
Keep a simple test record with the scanner model, serial number, software version, target type, scan resolution, focus mode and observed line-pair result. This is especially helpful when maintaining an older unit, since LED output, transport alignment and film-holder condition may change over time. It also creates a useful reference when comparing a Nikon scanner with a dedicated modern model or a local lab scan.
For everyday archiving, the practical target is consistent, natural detail rather than the highest possible number. Clean the film, keep the holder flat, disable enhancement during the baseline test, and judge the finest USAF bars alongside a real negative. That combination will show whether the Coolscan is genuinely resolving film detail or simply producing a larger, sharper-looking file.