How do you check a lens to see if it is a good copy?

dubcat

dubcat

Messages
1,310
Name
Amir
Edit My Images
Yes
One thing I have never done when buying my previous lenses (10-22 and 17-55 2.8 IS are the current batch) is to check whether I have a good copy. Is this something I should do?

I have just bought a 1.4x TC and a 70-200 f/4L IS - if I want to check that they are good copies how would I go about doing that?

Thanks in advance,
Dub
 
99.9% (a figure I just made up!) of "bad copy" threads and post are down to user error. probably a fraction are due to mis-handling during carriage.

Take pictures. Enjoy. Don't worry about it. Don't look for problems that are not there and don't take a picture of an angled test chart as this will find non-existent "problems".
 
thumbs up to all replies in this thread. Any lens will show up problems if you push it.
 
Agree with some of the posts above, however, if you want to be sure the lens is fine shoot a brick wall that fills the frame fron a reasonable distance, with a deep DOF, and check that the sides are equally sharp.

E.g. Is a crop of one third of the leftmost image the same as then right hand side?

Also worth setting the lens on a tripod and autofocusing on a few things, then switching to Live View and checking by zooming in that you can't tweak the focus significantly by hand.

If it's equally sharp all over and focuses fine, then what more coild you want?!
 
Don't pixel peep, it makes most lenses look bad. Your never going to print a picture at 100%, so why check one at that size and start worrying about non existing problems.

Just use it and enjoy it.

John:)
 
if you want to test your copy for sharpness, which you obviously do ,then you need to rule out camera shake and DOF differences, so set up on a tripod, stop down to something like f8 (although 5.6 is where most lens' sweet spots are) and shoot head on to a level surface, like a brick wall as suggested above. keep the iso on 100 to rule out noise also
 
if your talking about used lenses, then so long as you ask all the relevant questions you'll be fine. If you on about new lenses, they're mostly made by machine on big production lines, and QC and QA will be controlled very well, especially with the big brands, canon, nikon, sigme, tamron etc, so no need to worry at all, if by the slightest chance theres something worong that will make a meaningful difference to your photos, theres more chance the manufacturer will issue a recall before you even notice it!
 
I hear a lot of people going on about sigma's and bad copies...personally I've never had any problems, is it really a case of mountains out of molehills?
 
silverstar189 said:
I hear a lot of people going on about sigma's and bad copies...personally I've never had any problems, is it really a case of mountains out of molehills?
Click to expand...

I think its a combination of things. The people who you hear from on the net are usually the ones with problems while 99% of sigma owners are out enjoying their lenses. Its this combined with the paranoia it generates which leads to people testing their lenses incorrectly. You just have to search sigma 10-20 threads on flickr groups to find a load of people worried about their lenses. Threads like "do i have a bad copy??" and "help i think my lens is faulty!!" crop up all the time but most of the time it's people testing their lenses on landscapes at f/4, or shooting pictures of their TV remote control hand-held indoors. I mean no wonder the edges are soft :bang:

The fact of the matter is that only a handful of people have any actual problems. I myself ummmed and aaahed over which ultrawide to get because of the so-called "soft copy" issue with sigma, but having tried both the canon and the sigma i'm glad i went with the siggy because a) its actually sharper and b) its better built and its MUCH cheaper. OK, so i bought off a forum member and saw the pics first but i suspect all this bad rep is hocus-pocus and the chances of getting a duffer are actually much lower than reading internet forums would lead us to believe...
 
I'll second all that. I had worries about my 40D and my 24 - 105.

I took the bull by the horns and did a check for back focussing. As far as I could tell there was absolutely nothing amiss. Worrying about whether your lens or body are
not up to scratch is the photogrphy equivalent of hypochondria and should be treated in the same way. 99.9 % of the time there's nothing wrong.

The other 0.1% kills you.

ROFL
 
Well I've bought and sold far too many lenses for my own good and so far I've found no definite issues with any of the Canon ones. I had concerns with 40D & 50D bodies I owned. The 40D was definitely screwed and even 2 attempts at recalibration failed to fix the focus, the 50D was minutely out but within spec. Shows the potential for user error though, as I had suspected it was slightly out in the opposite direct to the actual situation.

I agree that many perceived errors are user problems, but here's a few I'm sure aren't:

I've tested 19 (yes 19!) Sigma 30mm 1.4 lenses and 18 of them front focussed (tested on three different bodies so confident it is not a body issue). The 19th back-focussed. Focus was fine at close distances, over about 8 feet errors crept in.

My Sigma 50mm 1.4 lens worked great on my 5D - absolutely perfect. Occassionally it mis-focussed on my 50D.

I had a Sigma 50mm 2.8 Macro that focussed poorly if used as a standard lens, but was fine for Macro work. Recalibration didn't help it.

I've not got it in for Sigma though - two 150mm Macros have performed just great and I had an old 70-300 APO which was very good as well.

I've experienced focus errors with the 1D MKIII - in less than brilliant light it can lose accurate focus when taking static objects in servo mode. It also can be unpredictable with objects moving directly towards the camera, but tweaks to the many settings help to improve this.

Most errors are user generated though!

Phil
 
Those are all premium grade lenses you've got. Very unlikely to be anything wrong.

If you want to know if you've got a good copy or not, it's too late! The only way to tell that is to test two or three samples side by side before you buy. I have done that very many times, mainly because of internet derived paranoia, and including several Sigmas, and they've all been fine.

If you want to check if your lens is faulty, the majority of image quality issues are down to poor centering during assembly, or if the lens takes a bad bump. Very easy to check in two minutes if you have a decent LCD.

Distant subject to minimise potential focus errors, like a road sign or car number plate. Good light. Focus carefully. Lock it. Lowest f/number and an absolutely shake-free shutter speed. Crank the ISO to get it. IS on.

Shoot four pictures with the target in each corner of the frame, ensuring the target is exactly the same distance from the frame edges each time. Check the images at full magnification on the LCD. Do it at min, max and middle focal lengths.

What you are looking for is an equal level of sharpness in all four corners. The absolute level of sharpnes is irrelevant, which is just as well as it will probably look pretty poor - this is a tough test. You are only looking for equality and if one or more of the corners is significantly out, the centering is out. If there's an error, repeat the test. If you can't replicate the problem, it's user error ;)

If there is a problem, it will be obvious. If you have to look twice and look carefully and then look again, there is not a significant variance.

Another good way to check general image quality is to test the lens against another lens of known good standard. It won't tell you if the lens is a good copy or not, but it will give you an idea of relative performance.

In all these things you need a direct comparison to draw any meaningful conclusions. Same camera, same exposure, same subject, same light, same processing - the only variant being the lens. Testing a lens in isolation and then looking at the results and wondering if they are any good or not, is just impossible outside a strictly controlled lab. Meaningless, TBH.
 
You must log in or register to reply here.
Back
Top