Tips for knowing a new lens isn't a lemon?

Discussion in 'Native Lenses' started by Mike_D, Jul 15, 2017.

  1. Mike_D

    Mike_D Mu-43 Regular

    Jul 2, 2017
    I read many posts about variations between copies of certain lenses, or a recent post where a reviewer found a prestigious lens less than desirable perhaps due to q/c issues with the lens given to the reviewer.

    With that in mind, any advice for the simplest and most straightforward ways to test a new lens to ensure it doesn't have QC problems (i.e. is a "lemon")?
    • Like Like x 1
  2. wjiang

    wjiang Mu-43 Hall of Famer Subscribing Member

    The good old brick wall test?
    • Like Like x 1
  3. eteless

    eteless Mu-43 All-Pro

    Jun 20, 2014
    I generally take a picture of my dogs, if I'm happy with the picture I acknowledge that even a defective lens is more capable than I am.

    As long as it functions as expected (autofocus, iris, etc), I don't really care about most of the metrics used in reviews...

    Edit: As to why I do this: Every lens you have ever bought is a lemon - any test to the contrary is just not precise enough to detect it. All consumer products are built to a price, no matter what the lens is if you look too closely you're going to be disappointed, it's better to just enjoy using them.
    Last edited: Jul 15, 2017
    • Agree Agree x 6
    • Wow Wow x 1
  4. gnarlydog australia

    gnarlydog australia Mu-43 Hall of Famer Subscribing Member

    Feb 23, 2015
    Brisbane, Australia
    Damiano Visocnik
    make that wall at a decent distance because decentring will be more evident there
    • Useful Useful x 2
    • Informative Informative x 1
  5. maflynn

    maflynn Mu-43 Regular

    May 7, 2012
    I just use it, and if there's significant softness, or CA, then I figure its not a good one. I noticed once, that my Nikon Lens that I spent a lot of money was performing worse then the kit lens. took some test shot of roses, and cropped them to 100%. Did the same thing with the kit lens and brought the lens, and pictures into my local camera store.
  6. Dziwaczka

    Dziwaczka Mu-43 Regular

    Dec 20, 2016
    • Useful Useful x 2
  7. pdk42

    pdk42 One of the "Eh?" team

    Jan 11, 2013
    Leamington Spa, UK
    Since I do a lot of landscapes, even sharpness at infinity is important to me. So, I usually take a shot of a scene with distant trees, stopped down 1-2 stops from wide open. I then turn the camera upside down and repeat. If there's any decentring or such like i'll quickly see it by comparing the opposite sides of the image in both orientations.
    • Useful Useful x 2
    • Like Like x 1
    • Agree Agree x 1
    • Informative Informative x 1
  8. retiredfromlife

    retiredfromlife Mu-43 Hall of Famer

    May 15, 2016
    Sydney, Australia
    I may give this chart a try, any tips on how to use one like;
    - How big do you print one of these charts for testing A4 or A3
    - Is there a certain distance the chart should be from the camera, for different focal distances or
    Since the chart is not the same aspect as a Mu-43 sensor do you fit the chart to the sensor width and that determines the distance ?
    - Have to find some way to keep it dead flat as well
    - or any other tips as I have never tried to test a lens before.
  9. Dziwaczka

    Dziwaczka Mu-43 Regular

    Dec 20, 2016
  10. cptobvious

    cptobvious Mu-43 Veteran

    Jan 8, 2013
    My routine when buying a lens:
    1. Check outer physical condition for any damage
    2. Use a flashlight to check for debris inside the lens (rare but it happens; dust is ok)
    3. Test all the switches and rings on the lens
    4. Point the camera at a blank wall and cycle the AF a few times, listen for any odd noises from the AF mechanism.
    5. Optical test:
    a. Take the camera out on the balcony (I live on the third floor). Rather than a brick wall test I take a picture of the ground with the camera as parallel to the ground as possible, using in-camera levels. At the height I am at, there is enough DOF to cover any minor tilt, even when the lens is shot wide open (especially for MFT where the DOF is deeper). I will take test shots wide open and each stop down to f/5.6. For a zoom lens I'll repeat that for the wide and tele end and the middle focal length.
    b. I'll test centering at infinity by focusing on something at a distance (e.g. a row of trees)
    c. I'll do a field test on the ground by setting up a tripod and taking shots at various distances with the aperture stopped down. I'm mainly checking the foreground to see if it is evenly in focus.

    I'll open the files in Lightroom (all RAW) and look at the files at 100%. A really bad lens (rare) is blurry across the frame at all apertures, but more commonly a decentered lens will be blurry on one side but sharp and in focus on the other. If I think a lens is decentered, I will repeat some tests but with the camera upside-down, to confirm that the blurry side is reversed.

    Even though this sounds like a lot of work, it probably only takes me 20-30 minutes now. Sometimes a lens will have very minor decentering on the 'flat field' test but it won't really show up in the field test shooting, so I'll keep it. However, a badly decentered lens will show it in every shot.
    • Useful Useful x 2
  11. alex g

    alex g Mu-43 All-Pro Subscribing Member

    Mar 30, 2016
    New York / Bath
    I think the distance at which you test the lens can often be very significant. Most designs have an optimum focal distance at which they perform best, but that information is rarely made available to the consumer. This is a separate issue from sample variation — the latter is a manufacturing issue, whereas the former is intrinsic to the actual design.

    It makes a lot of sense, therefore, to test a new lens at the sort of distances you typically shoot at — if you mainly shoot portraits, test it at typical portrait distances; if you mainly shoot wildlife, test it at wildlife distances etc. Naturally, the choice of test subject/methodology will depend largely on your chosen distance. A standard test chart is perfect when the distance is small, less so when the distance is large.

    Most reviewers these days test lenses at short distances, and consequently many designs are optimised to perform well under those conditions, which is great for portrait and garden bird photographers, but less beneficial for landscape shooters or for more distant kinds of wildlife. If you regularly shoot at 10m +, it's definitely worth testing a new lens at your favourite distances, even if you have no reason to think you have a bad copy — it might just not be a good design for what you need it for.
    • Like Like x 1
    • Agree Agree x 1
    • Informative Informative x 1
    • Useful Useful x 1
  12. Jonathan F/2

    Jonathan F/2 Mu-43 Hall of Famer

    Aug 10, 2011
    Los Angeles, USA
    To check focus accuracy shooting a subject in both AF and MF is important. I've had a lens that was not focusing accurately even with CDAF.
    • Agree Agree x 1
    • Informative Informative x 1
  13. Turbofrog

    Turbofrog Mu-43 Hall of Famer

    Mar 21, 2014
    @pdk42@pdk42 's test looks like a very good one. For a number of reasons.

    Not only does it test performance that will be important to the way that he actually uses the lens, but the methodology is self-calibrating. He doesn't need to make sure that his planarity to the surface is absolutely 100% perfect, since he's flipping over the camera. If his focal plane was non-parallel, producing softness, it would appear the same way in both images when mirrored vertically.

    That planarity problem is a big difficulty when people try to use test charts. Not only is the depth of field often a limiting factor when shooting a test chart from short distances, but most lenses (especially fast ones) are not even necessarily corrected for flat-field performance.

    The main thing is to be conscious of what you actually use your lens for. And to realize optical issues like field curvature are not necessarily the same as soft corners. They manifest differently, and have a different impact on real world use. Being mindful of cross-frame performance, stopped down, near infinity with a wide-angle lens like the PL8-18mm is reasonable. Obsessing over cross-frame sharpness when wide open at 2m with a lens like a 45mm/f1.8, or even a 25mm/f1.8, is generally pretty silly, since it's hard to find real world subjects that will even fit with such a slim focal plane...
  14. HRC2016

    HRC2016 Mu-43 Regular

    Dec 6, 2016
    Does BH sell brick walls?
  15. Impulse

    Impulse Mu-43 Regular

    Apr 18, 2014
    Puerto Rico
    Agreed, that kinda test is a lot easier to perform and less susceptible to tiny methodical inconsistencies than brick wall and test chart setups... I'd still perform it wide open tho just because any flaws are usually more evident that way, if you see something bothersome or questionable you can then stop down and figure out whether it'll be an issue in actual usage.

    The other thing you wanna do is use manual focus and keep it set exactly the same between shots, avoiding AF minimizes any other testing variables. Another variation of said test is the four corner technique... A riff on the same concept but it works better with certain focal lengths or scenes.

    Just put something like a far away building (or buildings) in the center of the frame, then either shift or rotate the camera such that the targets end up on all four corners. The center is a control (if it looks worse than a corner then there's something seriously wrong!), you'll mostly wanna compare the corners against each other.

    There's another interesting test suggested by ProfHank which involves using OOF PSD, it might actually be the simplest to setup by far but interpreting the results seems a little more ambiguous to me... Maybe I just haven't used it enough or run across the sample of a really faulty lens. Some examples of the '4 corner' & OOF PSD;

    12-40: sharp enough? decentered?: Micro Four Thirds Talk Forum: Digital Photography Review

    Simple decentering test

    Or, you know, you can always go out and shoot a lot with it and use common sense when reviewing the images critically... I'm too impatient/neurotic/geeky for that. :p