home > photography > lens tests

Lens Test Using Test Charts -- Take 2

Purpose

Compare the 18-105mm power-zoom lens against other lenses, using a Nex-6 (and hopefully improving over previous tests).  Testing on this page was done on a tripod with OSS left on and using an infrared remote.  Results have to be considered somewhat approximate, as they are affected by various factors that are prone to errors.

ISO 400 was used for most of the shots, as light was a bit low, and raw were used directly by mtf_mapper (which should not apply sharpening); I don't believe the ISO would largely affect the resolution, as NR was turned to low, but in-camera sharpening (even with the setting at "0") might.

Testing

For this round of testing, I used MTF Mapper.  It automatically calculates MTF 50 values when using a certain kind of test chart.  The precision of the results can be affected by the quality of the chart print-out and even in-camera sharpening.

In the below numbers, I'm listing the maximum result as well as the top 5% from the test area (not the whole frame).  It's approximate, because for the wider angles, I'm only sampling the center of the frame, due to the size of the chart needed to cover the entire frame.  The "max" number is approximate (could be off by as much as 2 lp/mm) due to lack of precision in the test chart results; the 5% number may be more accurate for judging peak performance, but I don't know.   At a future time, I'd like to test edge performance as well.

To convert lp/mm to lp/ph, multiply by 15.60192.  [3264/(1000/4.78)]   (For other cameras, you'll need to find out the pixel size and/or the sensor size.)

Center

Sigma 30, 30mm, F2.8: 54.4 lp/mm max., 51.03 top 5% of center test area
Sigma 30, 30mm, F4.0: 56.5 lp/mm max., 56.91 top 5% of center test area
Sigma 30, 30mm, F5.6: 60.7 lp/mm max., 57.24
Sigma 30, 30mm, F8.0: 54.4 lp/mm max., 52.28

18-105, 18mm, F4.0:  48.1 lp/mm max., 50.30 top 5% of center
18-105, 18mm, F5.6:  52.3 lp/mm max., 57.55 top 5% of center
18-105, 18mm, F8.0:  48.1 lp/mm max., 54.68 top 5% of center

18-105, 30mm, F4.0: 46.0 lp/mm max., 47.25
18-105, 30mm, F5.6: 54.4 lp/mm max., 54.06

18-105, 50mm, F4.0: 43.9 lp/mm max., 43.84
18-105, 50mm, F5.6: 58.6 lp/mm max., 57.82
18-105, 50mm, F8.0: 52.3 lp/mm max., 52.88

18-105, 105mm, F5.6: 41.8 lp/mm max., 37.40

E20, 20mm, F2.8: 54.4 lp/mm max., 55.86
E20, 20mm, F4.0: 58.6 lp/mm, 59.88

E20 w/UWA, 15mm, F4.0: 46.0 lp/mm, 45.33  (Confirmed with 3 tests)
E20 w/UWA, 15mm, F4.0: 35.6 lp/mm, 39.54  -- Edge
E20 w/UWA, 15mm, F5.6: 48.1 lp/mm, 49.64

E16, 16mm, F2.8: 35.6 lp/mm max., 37.23
E16, 16mm, F4.0: 48.1 lp/mm, 49.54
E16, 16mm, F5.6: -- lp/mm max., --
E16, 16mm, F8.0: 56.4 lp/mm max., 57.90

16-50, 20mm, F4.0: 43.9 lp/mm max., 43.90
16-50, 20mm, F5.6: 56.5 lp/mm max., 55.10

16-50, 30mm, F5.6: 46.0 lp/mm max., 41.50  (Only one reading of 46 lp/mm found, so I'd consider the peak to be 43.9.)
16-50, 30mm, F8.0: 46.0 lp/mm max., 46.04

First session with 18-55:

18-55, 21mm, F4.0:  52.3 lp/mm max.,  50.01 top 5% of center
18-55, 21mm, F5.6:  56.5 lp/mm max.,  59.40 top 5% of center

18-55, 30mm, F4.0:  43.9 lp/mm max., 41.66
18-55, 30mm, F5.6:  54.4 lp/mm max., 48.86
18-55, 30mm, F8.0:  52.3 lp/mm max., 50.55

Second session with 18-55 (with modified lighting and position):

18-55, 18mm, F4.0:  56.5 lp/mm max.,  59.05 top 5% of center
18-55, 18mm, F5.6:  58.6 lp/mm max.,  55.58 top 5% of center
18-55, 18mm, F8.0:  58.6 lp/mm max.,  57.58 top 5% of center

18-55, 20mm, F4.0: 52.3  lp/mm max.,  49.99
18-55, 20mm, F5.6: 62.76 lp/mm max.,  62.89
18-55, 20mm, F8.0: 58.6 lp/mm max.,  N/A

18-55, 30mm, F4.0:  41.8 lp/mm max., 40.97
18-55, 30mm, F5.6:  52.3 lp/mm max., 50.03
18-55, 30mm, F8.0:  52.3 lp/mm max., 52.83

18-55, 55mm, F5.6:  35.6 lp/mm max., 32.70
18-55, 55mm, F8.0: 41.8  lp/mm max., 41.55 (one entry found  of 43.9)

Edge:

18-55, 18mm, F4.0:  56.5 lp/mm max., N/A
18-55, 18mm, F5.6:  48.1 lp/mm max.,  46.42 top 5% of edge
18-55, 18mm, F8.0:  39.7 lp/mm max.,  43.68 top 5% of edge

 

 

Conclusion

Between using the new methodology and no sharpening, the numbers are much lower than the previous test page, but probably are closer to reality (which may be helpful for comparison with other tests).  Numbers would likely be higher on a higher-resolution sensor.  So far, results are not unexpected -- in earlier tests, it seemed pretty clear that the 18-105 benefited greatly from stopping-down to f5.6. While the 18-105 has trouble competing with the primes, it's not bad, and it is correct to consider it to be a kit-lens-upgrade.

While many people have stated that the UWA adapter does not degrade the image quality, results here differ from that.  However, even with those poor numeric results, it's hard to tell with everyday photos.  The often-derided 16mm prime is pretty sharp in the center when stopped down.  (Not shown here is poor corner performance, which may or may not matter depending upon the subject.)

Originally created: July 18, 2015.
Last updated: July 25, 2015.