[clackers]

Ha ha Are you being serious or, as I think, pulling my leg? (I'd say taking the pi** but we're on a public forum lol)

Not joking. If you take a second shot, same setup, but upside down, and the opposite side is soft, that's a giveaway. Probably decentred.

You need to rule out that one of the right angles in your testing wasn't quite perpendicular.

[Digitalis]

one lens - a zoom - showed that it is very slightly softer on the right side than the left (at the outer 10% of the image);

Classic case of de-centering, this isn't uncommon, zoom lenses are optically complex the more elements in the lens design the greater the chance is for something to go pear shaped.

another - also a zoom - showed that, wide open at the longer end, the slight halo around white lettering against a black background is always slightly up and to the right side of that lettering rather than a consistent "glow" all around. This is looking at 1:1, even 2:1, and even then, the issues were far from bad - but they *were* noticeable.

Sometimes when people see a problem their anxieties tend to focus, and magnify a problem - rather than the real consequences of it, which can be mild to non-existent. The solution to this issue is simple - stop the lens down. Which is what you should be doing in any case, as a vast majority of lenses perform best at 2 stops down from their widest apertures.

If you want to make a living from architecture, product, or astrophotography for example I imagine even slight optical problems will be important.

^ this. I own a specialized optics testing bench which is used to project various laser etched test patterns through a lens, each pattern is designed to show up any optical flaws. Using this bench removes the camera from the list of variables that affect image quality. Every lens I have ever bought has been mounted on it and scrutinized*. I test new and old lenses for colleagues because many of them simply don't have the time or equipment for the kind of precise optical testing I can do. I have saved myself and my colleagues both time and money from substandard copies of lenses - and from agonizing over it. If I don't the lens a thumbs up, it is time to get another copy.

*which isn't easy as the bench, which is made by Zeiss is actually designed to test medium format, 4X5 and 8X10 lenses. I have to use modified lens boards to mount 35mm lenses on it. Lenses with electromagnetic diaphragms and by-wire focusing systems in Nikon F/Canon EF mounts have a battery powered mounting board that cost nearly a thousand dollars that allows for focusing and aperture control.

[stevebrot]

and then, there's my testing inaccuracies.... sigh

...gotta make sure the sensor is parallel to your target before assuming one-sidedness is due to a decentered element.

Back to your original question...
I do a thorough inspection (visual and mechanical) of all lenses (both new and used) on receipt
Included in the visual inspection is a thorough look-see of the inside using a magnifying loupe, a darkened room and a small flashlight shining at oblique angle through the other end. The aim is to detect obvious defect or corruption of the glass. Yes, I once found a cracked element in a brand new lens from a major maker.

• Included in the mechanical inspection is evaluation of all controls, aperture operation, and focus operation (manual and auto)
• Any rattles or looseness?
• I then do a little shooting of ordinary stuff to put the lens through its paces. If there are serious optical flaws or stuff like poor infinity calibration, it will generally show. (Yes, Samyang, we are talking about you...)
• If things look disappointing in real world shooting, I will then check against a test target under controlled conditions. If you want to screen specifically for a decentered element use a Siemen's star or LensRentals handy substitute:
  
  LensRentals.com - Testing for a Decentered Lens: an Old Technique Gets a Makeover

Have I ever sent a lens back? You betcha! The one with the cracked element went in on warranty and used stuff...dozens, mostly for fungus and one for what appeared to be a missing or backward internal element.


Steve

[stevebrot]

I have saved myself and my colleagues both time and money from substandard copies of lenses - and from agonizing over it. If I don't the lens a thumbs up, it is time to get another copy.

It just occurred to me that what you describe might be a decent business model.


Steve

[ScooterMaxi Jim]

You know what, "it probably is decentered" is a fact with every single lens ever produced that isn't designed for military or outer-space use. As such, I have purchased 387 lenses over the past seven years because upon testing I have found every one of them to be slightly decentered when blown up to 500x. This is really disturbing. It has prevented me from going on a real photo shoot for the past seven years. Besides that, I'm frustrated that there are no good lenses out there.

Actually, seriously, you need to watch closely when you're looking at UWA zooms because they are very difficult to get right even for very normal viewing and printing purposes (some of us use mundane real-world standards). One particular brand that specializes in extreme WA AF zooms such as 8-16 for crop and 12-24 for FF - but shall otherwise remain unnamed - is especially challenged in getting these optics centered properly (but, hey, they put in nearly silent motors that actual stand up to rigorous use). So, yes, you need to check things out, especially if you have an extreme lens. Otherwise, pixel peeping is simply an exercise in frustration - hence the vulgar but appropriately assigned term for it relating to, umm, down under.

[Digitalis]

It just occurred to me that what you describe might be a decent business model.

It's actually one of the perks of being on my good side, that and being able to hire some of the lenses I own with the guarantee they will meet expectations.

One particular brand that specializes in extreme WA AF zooms such as 8-16 for crop and 12-24 for FF - but shall otherwise remain unnamed - is especially challenged in getting these optics centered

There is only one manufacturer who makes those lenses = Sigma. I went through about 5~6 copies of the sigma 8-16mm until I was sent a copy from sigma japan which was tested by sigma techs, and myself, and is perfectly centered. Though I have to admit that was one of the hardest lenses I have ever tested - 8mm isn't a common focal length, I had to alter the optical projection apparatus on my testing bench to completely fill the fov with the test pattern on that lens.

[stevebrot]

I went through about 5~6 copies of the sigma 8-16mm until I was sent a copy from sigma japan which was tested by sigma techs, and myself, and is perfectly centered.

Thanks for doing your part to keep them honest. With any luck they plugged what they learned into their CQI process.


Steve

[Digitalis]

Thanks for doing your part to keep them honest.

Sigma Japan was being honest, it was the Australian distributor that was being an intransigent nuisance. So I went over their heads and contacted Sigma Japan about the issue personally, their response was prompt and resolved the situation. I had similar issues with the Australian distributor with the 18-35mm f/1.8 ART, but I resolved that situation myself by having the lens rebuilt and optically re-aligned by a specialist.

[pathdoc]

So did the rebuilding and realignment cure the Dreaded Art Focus Issue?

Digitalis, if I ever get back to Adelaide, God knows when, I seriously need to come look you up and see how you do what you do.

ScooterMaxi Jim, I think you might have OCD...

As for me, if I can't see any problems with the images on my full-size Toshiba laptop I'll give a lens the thumbs up, knowing I will never make a physical print bigger than that. Once, I had to resort to my professionally calibrated screens at work to sort out the differences between a pair of lenses, but that was an exception.

[januko]

Quality control for the last 6 or so years has been bad from all manufacturers.

Very frustrating because we the buyers are paying good money (read: huge and hard earned money) and yet we are cheated by them.

Makes me think they are always in a hurry to make a buck!

Not just cameras. Cars. Gadgets. You name it!

[Digitalis]

So did the rebuilding and realignment cure the Dreaded Art Focus Issue?

No. My copy of the AF is still only 78% accurate. The re-alignment fixed the lateral rotation of the rear cell. The focus plane was once again flat, rather than running at an angle to the sensor.

[FantasticMrFox]

I do two quick tests - I use a chart to test for focus accuracy and fine tune it in camera if necessary and I quickly check for decentering by focusing at infinity and taking images of an object far away (like a church tower), placing it in different corners of the image. I don't see why I would want to waste and sharpness a lens has by not correcting for inaccurate focus, and I want to be able to tell whether it's decentered (as that seems to be a growing problem).

I only do more thorough testing (taking pictures of test charts at various focal lengths and apertures, shooting into the light to provoke flare etc.) if I want to review a lens for my blog.

[Nicolas06]

To be honest, I expect that the manufacturer has a responsibility to implement respectable quality control. That being said, I am quite appalled at the numerous reports of quality issues with camera body and lens equipment in this modern (digital) age!! My past experience with extensive camera equipment purchases were from 1984 through 1988 involving Olympus/Zuiko 35mm, Hasselblad/Zeiss MF, Sinar/Schneider/Copal LF film systems and you just didn't have any manufacturing and assembly quality problems like today. A faulty aligned lens was unheard of to me back then.

Back in time, almost nobody looked at pictures at 100% crop equivalent magnification. Looking at 100% crop is like looking at a 60" print from near distance looking for issues.

If you used only 24x36 and common print sizes you would never see the issue. And with MF being much more common , most soft shoots issues were not the lens fault.

We can be nostalgic of the past photographically wise, but today you get much more for far less and while it doesn't mean much for the artistic side, technically we get much better pictures theses days that what we had to pay back in time.

[pathdoc]

Back in time, almost nobody looked at pictures at 100% crop equivalent magnification. Looking at 100% crop is like looking at a 60" print from near distance looking for issues.

If you used only 24x36 and common print sizes you would never see the issue. And with MF being much more common , most soft shoots issues were not the lens fault.

We can be nostalgic of the past photographically wise, but today you get much more for far less and while it doesn't mean much for the artistic side, technically we get much better pictures theses days that what we had to pay back in time.

This post needs to be a locked sticky at the top of the Lenses section.

[Sandy Hancock]

Digitalis, if I ever get back to Adelaide, God knows when, I seriously need to come look you up and see how you do what you do.

Make sure you look me up too....