 |
| M.W. Dunn filling station and convenience store (click to enlarge) |
The Law of Variation from the American Society of Quality website:
According to the law of variation as defined in the Statistical Quality Control Handbook:
“Everything varies.” In other words, no two things are exactly alike.
“Groups of things from a constant system of causes tend to be predictable.”
We can’t predict the behavior or characteristics of any one thing. Predictions only become possible for groups of things where patterns can be observed.What does this mean to you and your photo gear? It means nothing is perfect. Your gear isn't perfect and no two products (lenses, cameras, etc.) are identical. Everything varies, even if a tiny bit. The absolute best and most precise manufacturing techniques will produce minute variations in the final product. Each manufacturer knows this and sets maximum allowable tolerances for variation in their products. If a product is within these set tolerances, it goes out for sale. If, during manufacture, products regularly exceed the set tolerances, production stops and the manufacturer institutes problem solving to ascertain why the excessive variation is occurring. They then identify the problem to bring the products back within tolerances and resume manufacture.
For example, a manufacturer may set a variation scale for acceptability of -5 to +5 (for illustration here only) and the desired finished product is supposed to measure 0 on the scale. A product that measures +4 and a product that measures -3 both still are within manufacturing tolerances and go out for sale. A product that measures -6 fails. The ideal is for all completed products to measure 0. That rarely happens.
This product variation occurs with our cameras and lens as well. Flange distances from lens mount to sensor surface are set to very low tolerances, but still can vary minutely. Lens focus performance is set to minute tolerances as well. Lens element placement can vary. Sensor placement within a camera body is highly important but can vary by a couple of thousandths of a millimeter.
For example, if lenses focus within a set tolerance, they go out as a product that meets manufacturing standards. But few focus on the absolute perfect point. One camera with IBIS may give you more image stabilization than another copy of that exact mode. One lens may give you sharper and more consistent images than another of that exact lens. One tripod ball head may hold a long, heavy lens just a bit better than another of the same model. Variances occur and gear performs slightly differently.
But what happens, say, you buy a camera that is at a +5 on the tolerance scale and a lens that is also at a +5 on the tolerance scale? The result is a camera and lens that don’t focus together properly. Together they are out of spec in the + direction. However, if you have a camera at +5 and a lens at -5, you lucked out as the variations will largely cancel each other out and all is well with this particular camera body and lens combination. This is one of the reasons why Nikon and Canon (there may be others) that allow you to adjust your autofocusing point for each of your lenses. They know that there are variances (as well as issues with phase detection AF) and give the owner a methodology to counteract those variations to achieve sharp focus.
I’m greatly simplifying things for illustration and understanding’s sake but you get my point. Even minute tolerances can make a difference. And they do.
When Olympus put their m4/3 lenses on sale in November or December of last year, I purchased a 17mm f/1.8 prime. I currently carry an Olympus 25mm f/1.8 prime in my little Olympus kit (which also contains the wonderful 12-100mm f/4 PRO lens), but I wanted the option of a lens with a wider field of view. In this case, the equivalent of a 34mm full frame lens versus a 50mm full frame lens.
I received the lens from one of the large internet vendors and proceeded to thoroughly test it. To my dismay, the lens, in my opinion was not very sharp at all. I'm sure it was within Olympus' tolerances but it wasn't at all satisfactory to me. I know they do better. In fact, even the focus seemed somewhat inconsistent. Olympus has excellent quality control, so I chalked this up to product variation. The lens just did not work well with my particular camera body. I was disappointed.
Within a couple of days I returned the lens and asked to exchange it for another copy. Last week I received the new copy and, as I normally do, I thoroughly tested it. Wow! This copy exceeds my expectations! I smiled. From f2 to f/8 the lens is very sharp in the center, edges and corners. No smearing of detail in the corners at any aperture. At f/11 I can see just the start of a tiny bit diffraction softening as I do with my other m4/3 lenses. But you really have to look hard at 100% to find it. At f/16 I clearly saw the effects of diffraction, again, as expected. To me, I would not hesitate to use this particular copy of the 17mm f/1.8 lens from wide open to f/11.
The lens seemed so good, I tested it against my Olympus 12-100mm f/4 PRO lens and, with the exception of the extreme edges at f/4, I don't see any differences in performance. At f/4, again, I really have to look hard at 100% to see a very slight difference at the extreme edges. The zoom was just barely noticeably better. As I said, this particular copy of this lens has exceeded my expectations.
This is a good example of product variation. Two lenses, identically manufactured and identical quality control but an easily seen difference in image quality. Product variation is an excellent reason to always test your gear when you first receive it so you know what you have and how well it will serve you.
Thanks for looking. Enjoy!
Dennis A. Mook
All content on this blog is © 2013-2018 Dennis A. Mook. All Rights Reserved. Feel free to point to this blog from your website with full attribution. Permission may be granted for commercial use. Please contact Mr. Mook to discuss permission to reproduce the blog posts and/or images.
1. Oh dang. I gotta get back in the books about autofocus adjusting.
ReplyDelete2. Nice post! This one takes me back 25+ years to the glory days of TQM and statistical quality control!
I had written a paragraph but deleted it about how I studied W. Edwards Deming TQM and statistical process control in graduate school, and how, after WWII, the American manufacturing industry rejected him, his ideas and SPC as they saw no reason why they would want to change their wasteful manufacturing ways as America had no real competition. Deming went to Japan where they embraced it and it dramatically improved their quality and efficiency to quickly pass America in manufacturing. Quality Circles and later, Six Sigma were outgrowths of Deming’s early work.
DeleteGosh, I guess I just did write the paragraph!
Dennis, you - or your lenses - have me slightly confused. So you have that wonderful 12-100/f4 zoom. Now, you get a 17/f1.8 with a focal length within the range of your zoom. Kind of following one of the lens-selection strategies Mike over at TOP advocates. You come to the conclusion ".. The zoom was just barely noticeably better." (at f/4). Wow, shouldn't a prime, well above (in numbers) its maximum aperture, be better than a zoom at its max aperture? Well, physics/optics would suggest that. Of course, the prime is smaller and lighter, but being rather wide angle you won't gain a lot of depth of field control either by having f/1.8 instead of f/4. So what was your objective for getting the 17/f1.8?
ReplyDeleteWolf, thank you for your comment and question.
DeleteI adopted a two lens strategy long before my friend Mike Johnston over at The Online Photographer recently wrote about it. However, my thinking is a bit different from his. I use a high quality wide to telephoto zoom as an all around (walk around) lens, in this case the 12-100mm f/4, that will serve the majority of my photographic purposes. In the past, I have utilized a Nikon 18-200mm lens for the same purposes. The moderately wide angle relatively fast prime primarily is for those low light instances I may encounter. For example, I may be out walking around at night or indoors in a church or other relatively dark structure where I would want to photograph. The fast prime gives me a couple more usable apertures so I can keep my ISO at a reasonable level. Everything is a trade-off.
This copy of the 17mm f/1.8 lens is much, much better than the two previous copies I have had. The first one I bought a couple of years ago was pretty good but not great. I sold it as I thought I could find a better lens. The one I received in December of last year, which I sent back produced poor image quality in my opinion. It also didn't focus correctly consistently. This copy is the best of the three, hence the post on sample variation. This copy is very sharp from center to edge, but at about three times the price, the 12-100mm f/4 PRO lens is, overall, even a bit better. Slower maximum aperture but a little better at comparable apertures at the edges. I am happy with both.
I hope this helps explain my rationale.