I recently purchased the Pentax 645D camera, as well as a new Pentax SMCP-FA 645N 80-160mm f/4.5 lens (ordered it from Japan). I also have a Canon and a Sony system, and have learned through experience to test any lenses that I buy with Imatest, as about 30-40% of the lens copies that I get are not sharp, are decentered,…
I thought I would share some quantitative Pentax 645D sharpness testing results. You can download a spreadsheet that shows the Imatest results, by viewfinder position, for the SMCP-FA 645N 80-160mm f/4.5 lens at 80mm, 120mm, and 150mm, and by f stop. The spreadsheet is at
http://cid-67997732cc8711ff.skydrive.live.com/redir.aspx?page=play&resid=67997732CC8711FF!227&type=5&authkey=NeXLMH7Adm4%24&Bsrc=EMSHOO&Bpub=SN.Notifications. I also tried to attach it to this post.
I did not test all the lenses at all f-stops, so if you see “NA” in the results, simply choose a different f-stop in the drop down menus. In the spreadsheet, blue signifies a drop-down menu.
Imatest is very sensitive to a number of factors, including lighting, distance to target, RAW conversion, post-processing,… To normalize, I try to keep all of these items the same from lens test to lens test. For my tests, I use Lightroom for raw conversion, Auto-Tone, Auto-White, no sharpening or noise reduction, and a Linear Curve with Black set to zero. I also run through a series of focuses and report the highest LW/PH. I’ll see some decentering this way, but can miss some as well (this is a relatively complex topic). Fundamentally, this approach tells me what a lens
can do, and it is the same as what it
will do the vast majority of the time.
Key findings:
• I got a pretty good copy of the 80-160mm lens. I run Imatest to determine whether to keep the lens or return it. This one is a keeper, but it’s not quite as good as I hoped it would be
• The best f-stop at all focal lengths, in terms of sharpness, is 9.5, followed by f11, then f8, and then f6.7. I don’t regard the sharpness at f5.6 as usable for my landscape work. The effects of diffraction are pretty large by f16. I did not measure f13.
• Sharpness at 80mm is better than at 120mm, which is better than 150mm. Racking out the lens all the way to 160mm resulted in significant softness, so I think the usable range for my copy is 80mm-150mm, and it is very good from 80mm-120mm.
• One measure of relative sharpness across camera systems is the average LW/PH for the lens-camera system divided by the number of pixels in the picture height. I compared the 80-160mm zoom to my Canon 50mm f1.4 (probably my sharpest lens, and I’ve got quite a few sharp lenses) and to my Canon 100mm f2.8 (non-IS lens). I chose these lenses as they fall within the vicinity of the 80-160mm lens, which is approximately the equivalent of a 65-130mm lens for 35mm systems. The relative sharpness of the Pentax at 80mm and f8 to the Canon 50mm lens at f8 is 91% - the Canon 50mm lens is ~10% sharper. You might say comparing a zoom to a fixed focal length lens is unfair. True. The comparable figure for my Sony Zeiss 24-70mm zoom lens at 50mm and f8, the best zoom that I’ve measured, is 108%. A number > 1 means that the Pentax 645D+80n-160mm lens system at 80mm is relatively sharper than the Sony Zeiss zoom at 50mm + a900! Given this, I believe the Pentax 80-160mm zoom performs quite well at the lower end of its range. This metric is shown in my spreadsheet comparator. The Pentax does relatively better (the ratio increases) at smaller f-stops. The anti-aliasing filter (or lack of it) will also be a factor here.
• The average increase in LW/PH for the Pentax 80-160mm lens vs the Canon lenses was about 130% (it ranged from 110% to 145%). From just the number of pixels, if the lenses were of comparable sharpness, we would expect a 145% increase – maybe more given that the Pentax does not have an anti-aliasing filter and the Canon does. This result makes the Pentax zoom seem even weaker relative to the Canon 50mm lens. However, a 30% increase for (approximately) the same field of view is a 30% increase, and I’ll take it.
• I did not see the edge softness to quite the same degree on the Pentax 80-160mm at 150mm as Lloyd Chambers saw, but I saw some softness that is reflected in the Imatest numbers. I believe the difference is simply copy-to-copy variation.
• The Pentax files sharpen up very well; perhaps better than the Canon files
• For non-Imatest target shots (real world pictures), I was very pleased with my initial results
I hope that you find this interesting and/or useful. If you do your own Imatest testing on Pentax 645 lenses, please contribute and add to this post.
