Lens QA: How and how much is really done?

Started 4 months ago | Questions thread
kenw Veteran Member • Posts: 5,342
Re: Lens QA: How and how much is really done?

RCicala wrote:

WhiteBeard wrote:

RCicala wrote:

WhiteBeard wrote:

I have been reading for years on "bad copies" of otherwise fine lenses; some folks even trying 3 or more copies of a lens - especially zooms - before finding a "satisfactory" copy or just giving up. I also have a nagging feeling that a fair proportion of folks satisfied with their first copy haven't really tested it thoroughly enough; not pixel peeping mind you, just being methodical about possible de-centering.

Which - finally - brings me to my point: How is lens QA done - or not - to let pass apparently so many "bad copies"? I would complicate things further by asking how much better the QA on so-called "Pro" lenses (Oly Pro line or Pana-Leicas, for example) is compared to the basic variety, given that the number of adjustment screws often rises along with the price.

Inquiring minds want to know...

I won't try to write (or rewrite) 6 blog posts as a response, but I'll try to summarize what I know.

1) Only Sigma, Leica, and manual focus Zeiss lenses are all tested optically after assembly. Optically.

2) All AF lenses are tested with a thick bar chart on a computer jig after assembly to set AF firmware, but this is NOT a real optical test (although some manufacturers claim it is).

3) During assembly of the lens, subassemblies are tested to some degree optically. Usually this is only done on axis (straight through the center). The most shocking thing I've ever discovered is there are big, huge, well known manufacturers with Ph. D. lens designers who truly believe if the center is OK, then the edges are OK. This is totally false except in one place: on the computer design programs. This has a lot to do with why you see people complain about one side is off, corners are muddy, but not as often that the center is soft.

4) Paper test charts, often ink-jet printed, are still the norm in many service centers and some factories. In other words, your tests at home are as accurate, and probably more carefully done, as their tests at the service center.

5) Electronics are always tested because that's easy and automated.

The biggest thing is people think of QA as testing the lens at the end of assembly. Most optical QA has to take place during the design of the lens (especially the mechanics) and the design of the assembly process.

A couple of points that those of you who can step away from 'what I expect' and think put on the 'what would I do if I was them' hat, here's some food for thought.

1) Only one company has an optical designer CEO: Sigma.

2) Two companies have optical designers in very high executive positions: Leica and Zeiss.

3) Two more companies treat optical designers as revered specialists, not in the main decision making process always, but consulted, opinions considered, and their names freely given: Canon and Nikon. Other companies usually have (some surprisingly don't have) optical designers on their staff, but mostly outsource their designs.

When the CEO is an MBA type, the CEO thinks about the bottom line. He doesn't want an awful design, it won't sell. He doesn't want awful quality, it hurts their reputation. He mostly doesn't want to spend money he doesn't absolutely have to. So he makes a business decision about 'what % of lenses have to be returned to make it worth upgrading QA'. Especially in today's market, where many of the manufacturers aren't certain they'll be here in two years, putting millions into QA isn't the highest priority.

I appreciate your input and reply, but as an engineer myself, I am very knowledgeable about the fact sh*t can happen with any new product / service. Our job is to minimize this but it is by no means easy or cheap. My problem lies in the fact that supposedly high line products have such shabby QA that a simple test by a casual user can show the lens is if not worthless, at least sub-par. Maybe I'm just wishing the engineers take over the world...

Oh, and by the way, what does the "Leica" in Pana-Leica stand for if not for a certain degree of quality?

The latter part I can partly answer: it stands for at least design consultation and a fee (or perhaps trade in kind in the form of Leica banded Panasonic cameras) paid to Leica for use of the name. But whether it stands for anything other than that only Leica and Panasonic know.

Years ago Panasonic commented on this in an interview.  It came down solely to the design - Leica had to "approve" the design by some means.  There was zero mention of quality control at all.

In one sense the Panasonic Leica zoom lenses do harken well to their Leica heritage.  Leica only specifies performance at three specific focal lengths for their "zooms" (aka Tri-Elmar).  For the true Leica they tell you what those focal lengths are.  For a Panasonic Leica zoom the "good" focal lengths are different for every lens manufactured.

(and yes Roger, I know all manufacturers' zooms have significant performance variations- but the three PL8-18 I've tested are surprisingly whacky)

-- hide signature --

Ken W
See profile for equipment list

 kenw's gear list:kenw's gear list
Sony RX1 Panasonic Lumix DMC-G1 Panasonic Lumix DMC-GM1 Olympus E-M5 II Panasonic Lumix G Vario 14-45mm F3.5-5.6 ASPH OIS +34 more
Post (hide subjects) Posted by
(unknown member)
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark MMy threads
Color scheme? Blue / Yellow