Sharpness results derived from Websites

Several websites publish images of ISO 12233 test charts that can be entered into Imatest SFR to obtain detailed sharpness measurements. This page contains sharpness results derived from charts downloaded from two sites:

The measurements in the tables below come from two areas of the ISO 12233 image, illustrated by red rectangles in the image below.

  • First row: near the center. The rectangle just above the center of the chart (about 15% of the distance to the corners).
  • Second row: near the corner. The rectangle in the lower right of the chart (about 78% of the distance to the corner). Results include Chromatic aberration, which cannot be measured near the center.

Regions of interest (ROIs) for sharpness comparisons

These measurements are limited to

  • One lens. For DSLRs this is usually one of the sharpest available, such as the 50mm f/1.4 prime (non-zoom).
  • One focal length. Sharpness is a strong function of focal length for zoom lenses. The sites aren’t very consistent in their choice of focal length for zooms in compact digital cameras; long, short or intermediate focal lengths may be chosen.
  • One aperture, typically around the optimum: f/8-f/11 for DSLRs; f/4-f/5.6 for compact digital cameras.
  • One ISO speed, which strongly affects noise and dynamic range, but has little effect on sharpness.
  • One sharpening radius, typically 2, used in standardized sharpening calculations, indicated by (corr.) and red columns.
  • One or two vertical edges (in most cases), resulting in horizontal MTF measurements. Sometimes vertical MTF (from horizontal edges) is different.
  • One RAW converter (usually the JPEG converter built into the camera) with one setting, usually the default. The choce of converter and sharpness setting has a strong effect on sharpness measurements.
  • Gamma assumed to be 0.5, typical for digital cameras. A 10% gamma error results in about a 2.5% MTF50 error. Though gamma error is rarely severe, gamma should be measured with Stepchart or Colorcheck for greatest accuracy.
  • Exposure and lighting are not always optimum. If the image is too dim or bright there may be some clipping on data (where it goes pure black or white), which reduces the accuracy of the sharpness measurement. Lighting may not be as even as it should be.

If you want to compare your own camera or lenses with the published charts, test lenses at various focal lengths and apertures, or observe the effects of ISO speed and signal processing, you’ll need to download Imatest. The trial version allows up to twenty runs.

Explanation of results

 

Megapixels The total megapixel count for the camera.
Picture height pixels The picture height of the camera. Important because sharpness is measured in units of Edge Rises or Line Widths per Picture Height.
Oversharpening % The amount of sharpening relative to standardized sharpening, which results in a modest amount of overshoot, similar to what you might get after manually sharpening the image. Undersharpening (as reported in Imatest SFR) is displayed here as a negative number.
10-90%
rises per PH
(raw)
The number of intervals per Picture Height where the average edge signal (with gamma removed) rises from 10% to 90% of its final value. The more the better, though aliasing can be a problem for extremely high values (significantly larger than the picture height), especially with Bayer sensors. See Understanding sharpness and SFR tour.
10-90%
rises per PH (corr.)
The number of 10% to 90% rise intervals per Picture height after standardized sharpening. This allows cameras with different degrees of sharpening to be compared on a (relatively) fair basis. A sharpening radius of 2 is assumed, unless indicated below. This radius works well for oversharpened compact digitals, but radius = 1, used by Imaging-resource.com, may be more appropriate for find DSLRs.
MTF50
LW/PH
(raw)
The spatial frequency where contrast drops to half its low frequency value. This is the best indicator of sharpness. See Understanding sharpness and SFR tour.
MTF50
LW/PH
(corr.)
MTF50 with standardized sharpening applied (corrected). MTF50 with standardized sharpening is best for comparing cameras with different amounts of sharpening; MTF50 without standardized sharpening (uncorrected) is best for testing and comparing lenses on a single camera model. See note on sharpening radius, two rows above.
Chromatic aberration (CA) A measure of “color fringing” near the edges of the images. Units are pixels. Tends to increase linearly with the distance from the center of the image. Under 0.5 pixels is insignificant. Over 1.5 is severe: quite visible in large prints. See Chromatic Aberration. An improved CA measurement, percentage of distance from the image center, is being implemented.
Shannon
capacity
MB
The information capacity of the image, based on sharpness (with standardized sharpnening) and noise. It correlates well with perceived image quality. But it should be interpreted with caution because it is strongly affected by signal processing, especially noise reduction. Best when used with RAW files. The number below is for a signal with the same contrast (100%) as the target (around 80:1). Shannon capacity is a relatively new measurement. We are still learning how to interpret it.
Comments Contains the source of the image (DPR or IR), and, if available, the focal length, aperture, and ISO speed.

Tables of results

DSLRs: Digital SLRs, which take interchangeable lenses. Used by professionals and serious amateurs. Sensor size is at least 22 mm diagonal. Larger and heavier than the compact digital cameras, below. This table is more complete than the table for compact digitals because it’s where my interest lies (and there are far too many compacts to keep up with). For each camera, the results in the upper and lower rows are for the image center and edge, respectively, as illustrated here.

Reminder: Remember the limitations listed above: one lens, one focal length, one aperture, one ISO speed, one RAW converter & settings (usually) built-in, one sharpening radius (2 unless otherwise indicated), etc., when interpreting these results. Also remember that several important image quality factors such as noise and dynamic range have been omitted.

Camera Mega-
pixels
Over-
sharp-
ening %

Center

10-90%
rises/PH
(raw)

Ctr

10-90%
rises/PH
(corr.)

Ctr

MTF50
LW/PH
(raw)

Ctr

MTF50
LW/PH
(corr.)

Ctr

Chro-
matic
Aber-
ration
Ctr
Shannon
capacity
MB

Ctr

Comments
Picture
height
pixels
Corner Cor Cor Cor Cor Cor Cor
Canon EOS-1D Mk II 8.2 -15 1351 1741 1355 1559 5.56 DPR. 50mm f/9 ISO 100. Very little built-in sharpening. Optimized for speed. Same pixel count as EOS-20D, but larger pixels.
2336 -21 1178 1589 1173 1443 0.40 4.99
Canon EOS-1Ds 11 -17 1509 2045 1521 1860 7.11 DPR. 100mm f/9 ISO 100. 24×36 mm sensor. Little, if any, built-in sharpening.
2704 -17 1485 2016 1519 1852 0.30 7.06
Canon EOS-1Ds Mk II (DPR) 16.7 -16 1919 2520 1922 2240 11.0 DPR. 50mm f/9 ISO 100. 24×36 mm sensor. Little, if any, built-in sharpening. Sharpening radius = 2.
3328
(Image cropped (1/4 area); corners unavailable.)
Canon EOS-1Ds Mk II (IR) Horiz. -39 1908 3310* 1938 3084* —  IR review results are shown. 100mm f/5.6 ISO 100. Horiz. and Vert. refer to the MTF direction, not the edge. *Sharpening radius = 1 improves results. Similar to results from the DPR image with radius = 1.
Vert. -47 1559 3063* 1630 2909*
Canon EOS-10D 6.3 0 1419 1508 1332 1335 4.02 DPR. 50mm f/9 ISO 100. Nearly identical to the EOS-300D. Very good 13×19 inch prints.
2048 -1 1410 1480 1331 1339 0.55 3.95
Canon EOS-20D 8.2 5 2017 1907 1758 1725 6.13 DPR. 50mm f/9 ISO 100. Successor to the EOS-10D. Excellent 13×19 prints: slightly better than the 10D.
2336 3 1807 1686 1603 1576 0.52 4.66
Fuji S2 Pro 6/12* -12 1869 2104 1819 1832 7.19 DPR. 50mm f/9.5 ISO 100. *SuperCCD diagonal pixel array. Uses interpolation to increase pixel count. Unusual ringing
in edge. High noise reduction?
2848 -7 1699 1987 1708 1756 0.73 6.72
Kodak SLR/c 14 7 3571 3457 3184 3177 15.1 DPR. 24×36 mm sensor. No anti-aliasing filter. (MTF > 0.5 at Nyquist.) Exceptional sharpness, but has problems with color Moire, illustrated
here.
Strong noise reduction. Discontinued.
3000
(Image cropped (1/4 area); corners unavailable.)
Nikon D2X 12.4 -5 2082 2382 2074 2098 9.04 DPR. f/8 ISO 100. 23.7 x 15.7 mm sensor.
2828
(Image cropped (1/4 area); corners unavailable.)
Nikon D100 6 -3 1381 1437 1243 1262 3.44 DPR. 50mm f/’9. Competitive with the
Canon EOS-10D.
2000 -4 1300 1373 1186 1227 0.74 3.3
Nikon D70 6 5 1590 1897 1766 1759 4.72 DPR. 50mm f/9. Weaker anti-aliasing filter than the D100. Excellent sharpness for the pixel count. Relatively high MTF at Nyquist (0.4) could indicate Moire problems, but I’ve heard few complaints.
2000 4 1769 1849 1700 1688 0.71 4.61
Olympus E1 5 -8 1174 1351 1142 1215 2.89 DPR. 50mm f/5 ISO 100. (Wider than optimum aperture) 4/3″ format. Curious spikes in noise spectrum at quarter multiples of Nyquist. Otherwise relatively flat noise spectrum indicates little noise reduction.
1920 -10 1121 1325 1105 1201 0.21 2.86
Sigma SD10 3.4 -2 1574 1618 1508 1508 3.53 DPR. 50mm f/9 ISO 100. Foveon sensor: no Bayer array. No anti-aliasing filter. Outstanding sharpness for the number
of pixels (if only it had more!) Very high MTF at Nyquist (0.5), but little apparent problem with Moire. See this Foveon report.
1512 -2 1463 1487 1381 1383 1.0 3.18

Compact digital cameras: These are primarily consumer cameras, though many produce excellent image quality, especially for letter-size or A4 prints. Most have non-interchangeable zoom lenses. Maximum sensor size is 11 mm diagonal.

Camera Mega-
pixels
Over-
sharp-
ening %

Center

10-90%
rises/PH
(raw)

Ctr

10-90%
rises/PH
(corr.)

Ctr

MTF50
LW/PH
(raw)

Ctr

MTF50
LW/PH
(corr.)

Ctr

Chro-
matic
Aber-
ration
Ctr
Shannon
capacity
MB

Ctr

Comments
Picture
height
pixels
Corner Cor Cor Cor Cor Cor Cor
Canon G3 4 18 1636 1360 1265 1157 2.19 DPR. 17.6mm f/4.5. Oversharpened.
1704 11 1346 1186 1134 1052 1.27  
Canon G5 5 20 1931 1565 1458 1346 2.88 DPR. 14.4mm f/4.0. Oversharpened. The chromatic aberration is not quite as bad as indicated because R and G are together; only B, to which the eye is less sensitive, is separate.
1944 12 1774 1544 1399 1327 1.81 2.78
Canon G6 7 12 2024 1780 1707 1612 4.39 DPR. 15.2mm f/5.0.
2304 -1 1738 1726 1599 1606 0.9 4.24
Fuji
F810
12.3* 13 2085 1962 1944 1791 6.02 DPR. 15.1mm f/5.6 ISO 80. *SuperCCD diagonal pixel array. Pixels are doubled in the test image. The lens is much better at the edges than the S7000.
3040 -6 1906 1937 1829 1762 2.19 5.87
Fuji S7000

UL
LR

6/12* 22 2014 1853
1912 1625   5.21 DPR. 13mm f/5.6 ISO 200 (a high ISO speed for a compact; somewhat noisy). Oversharpened. *SuperCCD diagonal pixel array; 6 megapixels orig; 12 used for measurement. Uses interpolation to increase pixel count. UL = upper left; LR = lower right (the usual) ROIs. Softness on the left was noted by dpreview.com. Edge falloff is considerable.
3040 -17
-22
1137
1202
1567
1591
1241
1230
1388
1446
2.67

2.04

4.52

4.62

Fuji S9000 9.1 -3 2233 2534 2233 2242   DPR. “Prosumer” 21 mm f/5.0 ISO 80. Impressive performer.
2616 -13 1554 1979 1669 1814 0.41 7.49
Konica-Minolta
DiMAGE A2
8 8 1829 1696 1611 1537 4.71 DPR. 13.3mm f/5.0 ISO 64.
2448 -3 1500 1580 1403 1432 1.29 4.37
Sony DSC-F828 8 9 1949 1751 1694 1617

4.69 DPR. 22.3mm f/5.0 ISO 64. 8 megapixel. Noisy at ISO speeds above 64. Superb lens.
2448 5 1884 1763 1669 1627 0.11 4.65
Sony
DSC-P150
7 9 1696 1689 1622 1537 4.37 DPR. 15.1mm f/8 ISO 100. Stopped down well beyond optimum aperture.
2304 -8 1331 1504 1340 1427 1.9 3.95

Interpretation of MTF50

What MTF50 do you need?  It depends on print size. If you plan to print gigantic posters (20x30 inches or over), the more the merrier. All the cameras in the above tables can make excellent 8.5x11 inch (letter-size; A4) prints. At that size the best of them wouldn’t look much better than the 4 megapixel Canon G3 (or any other high quality 4 megapixel camera).

With fine lenses, careful technique, and a little extra sharpening, my 6.3 megapixel Canon EOS-10D (corrected MTF50 = 1335 LW/PH) makes very good 12x18 inch prints. Prints are quite sharp from normal viewing distances, but pixels are visible under a magnifier or loupe; the prints are not as sharp as the Epson 2200 printer is capable of producing. Prints made with the EOS-20D (corrected MTF50 = 1581 LW/PH) are have a slight edge: I’d call them excellent.

Softness or pixellation would be visible on 16x24 inch prints made from the EOS-10D. The EOS-20D would perform slightly better, and several of the cameras listed in the table— especially the full frame (24x36mm sensor) DSLRs— perform even better.

The table below is an approximate guide to quality requirements. The equation for the left column is

MTF50(Line Widths / inch on the print) =
MTF50(LW/PH)


Print height in inches

 

MTF50 in
Line Widths/inch
on the print
Quality level— after post-processing, which may include some additional sharpening
150 Excellent— Extremely sharp at any viewing distance. About as sharp as most inkjet printers can print.
110 Very good— Large prints (A3 or 13×19 inch) look excellent, though they won’t look perfect under a magnifier. Small prints still look very good.
80 Good— Large prints look OK when viewed from normal distances, but somewhat soft when examined closely. Small prints look soft adequate, perhaps, for the “average” consumer, but definitely not “crisp.”.

Example of using the table: My Canon EOS-10D has MTF50 = 1335 LW/PH (corrected; with standardized sharpening). When I make a 12.3 inch high print on 13x19 inch paper, MTF50 is 1335/12.3 = 108 LW/in: “very good” quality; fine for a print that size. Prints look excellent at normal viewing distances for a print this size.

This approach is more accurate than tables based on pixel count (PPI) alone. Pixel count is scaled differently; the numbers are around double the MTF50 numbers. The EOS-10D has 2048/12.3 = 167 pixels per inch (PPI) at this magnification. This table should not be taken as gospel: itwas first published in October 2004, andit may be adjusted in the future.

Some observations

There has been a good deal of skepticism about 7+ megapixel compact digital cameras. They live up to expectations for sharpness, but their noise may be objectionable at high ISO speeds. Dpreview.com has found the Sony DSC-828’s noise at ISO 64 to be as high as the Canon EOS-300D (similar to the EOS-10D) at ISO 800.

The 6 megapixel raw/12 megapixel interpolated Fuji S7000, which uses Fuji’s SuperCCD, has sharpness comparable to the Sony F828. It also has some problems with noise. The SuperCCD is based on an interesting concept. Thanks to gravity, most detail in nature is vertical or horizontal, and evolution has made our eyes are more sensitive to those directions. SuperCCD pixels are located on a diagonal grid where the spacing between x and y-pixel spacings is 1/sqrt(2) = 0.707 that of a normal grid the same number of pixels. The diagonal grid is interpolated to a horizontal grid that has twice the pixels, keeping the reduced pixel spacing. This results in an impressive gain in horizontal and vertical resolution, though sharpness is slightly below that of a sensor with the same number of uninterpolated pixels.

I analyzed the sides of Fuji F7000 because dpreview.com noted a problem with softness on the left side. Uncorrected MTF50 is the appropriate metric for comparing portions of a single image. The problem is indeed very evident. It seems to be slightly worse on the left, where MTF50 (uncorrected) is 62% of the center value. The lens alsohas a problem with chromatic aberration. The Sony DSC-F828 does much better: edge MTF is 93% of the center value. It apears to have a finer lens, at least at the focal length and aperture chosen for the test.

The Sigma SD9 has high resolution for its pixel count because it doesn’t have, and probably doesn’t need, an anti-aliasing filter. If only Foveon could break the 3.4 megapixel barrier…