DiscussionCopy variation: how many 'bad copies' are actually user error?

Every second lens thread on every forum has someone saying they 'went through three copies' before they got a sharp one. I'm sure decentred lenses exist. I've had one myself, a zoom years ago that was clearly soft on the right side at every focal length, and the replacement was fine. But I've also watched people 'test' a lens by shooting a brick wall handheld at 1/60, at a slight angle, with stabilisation on and the AF point sitting on a gap in the mortar, and then post 300% crops as proof.

If you're going to call a copy bad, the bare minimum is a tripod, a flat target square to the sensor, several AF attempts or manual focus with magnification, the same aperture across the test, and comparing left side against right side, not centre against corner. Corners are always worse. That's not a defect, that's optics.

So, genuine question. Those of you who've returned a lens as a bad copy, how did you test it? And did the replacement actually test differently, or did it just feel better because it was new?

rawtherapist said:

how did you test it?

I test every new lens the same day it arrives. Tripod, a newspaper page taped flat on the wall, camera levelled with the bubble so the sensor is square to it, IBIS off, two second timer. Five AF attempts at each aperture from wide open to f/8, and I compare the left and right edges, not centre against corner. Then the same at a distance, usually rooftops from my balcony, because some problems only show at infinity.

I've done this since 2015, maybe fifteen lenses. I returned exactly one, a wide-angle zoom where left side was clearly worse at all apertures and also at distance. Replacement was fine. So it exists, but it is rare.

What I see more often with friends is front or back focus that is actually the body, or one AF point they always use that is not where they think. People blame the lens because it is the newest thing in the bag.

rawtherapist said:

did it just feel better because it was new?

Honestly yes, at least once. Sent back a lens in 2019 that I was convinced was soft and the 'good' replacement tested exactly the same when I finally did it properly. I'd spent £1,100 on it so my brain needed it to be perfect lol

rawtherapist said:

Corners are always worse.

Half my lenses are fifty years old and have been dropped by at least two previous owners each... if I worried about copy variation I'd never leave the house. Most people's 'bad copy' is shutter speed. Just saying

No, this is not only user error. With wide angle for landscape a decentred lens is very visible, one side of the mountain is sharp and other side is not, at f/8 on tripod, focus on infinity. I agree most complaints on internet are technique, but for landscape people it is real problem and they are not imagining it.

When you shoot panorama of mountains with wide lens, you see it immediately, because the left frame and right frame overlap and one is soft in the overlap and other is sharp. You cannot blame shutter speed on tripod at 1/250 in sun. Also with wide lens on landscape you look at the corners and edges, because there is the rocks and the trees, not only centre with the subject. So landscape people find problems that portrait people never see, and then portrait people say it is user error.

I checked my 14-30 on the first day exactly like this, on a ridge above Zakopane with snow, both sides the same, then I stopped thinking about it. This is the right way I think. Test once properly, then go and shoot.

TomekKRK said:

one side of the mountain is sharp and other side is not, at f/8 on tripod

Agreed, and that's exactly the kind of test that means something. Distant subject, tripod, stopped down, compare the sides. My complaint is the brick-wall-at-1/60 crowd, not you.

Here's my two cents... Out in the desert I once had what I was sure was a soft lens. Every frame mushy past 50 yards in the afternoon. Turned out to be heat shimmer coming up off the ground, and at 600mm you're shooting through a quarter mile of wobbling air. Same lens at 7am, sharp as a tack. I'd already printed the return label, swear to God. Before you blame the glass, check the air.

• Walt

rawtherapist said:

the bare minimum is a tripod, a flat target square to the sensor

From the business side, I'd go further. I buy from the same dealer every time, I test in the first week inside the return window, and if anything's off it goes back, no agonizing, no forum thread. I don't have time to run a lab and I don't want to own a lens I'm suspicious of on a paid job.

But the bigger thing is what 'sharp' actually means to the people paying me. Clients don't care what lens I used. They care that the headshot is sharp on the near eye, the skin looks good, and the files arrive Thursday like I said they would. Out of a few hundred sessions a year, the frames I've thrown away for softness came down to me missing focus at f/1.4 or a subject leaning in, basically every time. Never once a copy problem I could prove.

Most forum copy variation talk is hobbyists overthinking, respectfully. Test once, then go make money with it.

dmitri_k said:

I test every new lens the same day it arrives.

My sigma 18-50 is 'perfect' and I have never once tested it on a tripod lol. Ignorance is bliss

Most people would be happier if they stopped looking.

mfrost said:

Most people would be happier if they stopped looking.

This is true for life in general also.

only time ive ever had a 'bad lens' it was my own fault tbf, the focus limiter switch had got knocked on in the bag and i spent a whole morning at Rutland Water thinking the thing couldnt focus on anything close, bloke in the hide next to me pointed it out in the end, never lived it down

andyp_1974 said:

the focus limiter switch had got knocked on

This is the most honest post in the thread and should be pinned.

Ha, not pinning it, but I did laugh. Been there with the VR switch more than once!