Popular Photography "Most accurate lens testing"

Discussion in 'Digital SLR' started by RichA, Oct 9, 2006.

  1. RichA

    RichA Guest

    Pg. 48 of the current issue. Basically, they say their lens testing
    equipment provides a far better measure of lens performance than
    testing the lenses on a specific DSLR. They say that many lens tests
    done "out there" can be near worthless.
    RichA, Oct 9, 2006
    1. Advertisements

  2. RichA

    frederick Guest


    "The actual magnifying power of the lens tested greater than 1:1, coming
    in at an impressive 1:0.7"
    Heh - aps-c sensor camera with a test target designed for 35mm...
    Not just once do they make that mistake - and they haven't yet tweaked
    as to how this miracle occurs. Even when everything else from the mfr
    specs and focus distance at 1:1 and f/l says otherwise...

    Methinks elsewhere is better for objective lens tests.
    frederick, Oct 9, 2006
    1. Advertisements

  3. RichA

    Father Kodak Guest

    Which article? or, is there a URL?

    Father Kodak
    Father Kodak, Oct 9, 2006
  4. RichA

    RichA Guest

    I couldn't find it on their website, but I scanned it and posted it. I
    don't think they'll mind
    since it's something of an "ad" for the magazine's tests.

    RichA, Oct 9, 2006
  5. Keep reading Popular Photography. It's dumbed down to about the level
    of something you might understand.
    Randall Ainsworth, Oct 9, 2006
  6. Sure, the same class of mag as the one that a few years ago decided in a
    shootoff that the Tamron 28-200 (second generation) was better on all fronts
    compared to the combo 28-70 f/2.8 AF-S Nikkor and 70-200 f/2.8 AF-S VR
    Of course their "all fronts" didn't include image quality, only weight,
    price, size, and the question if you have to switch lenses once in a while.
    Since the Tamron is smaller, lighter, cheaper, and one lens versus two it
    was clearly far superior.
    Jeroen Wenting, Oct 9, 2006
  7. RichA

    Matti Vuori Guest

    And nobody needs to go and buy the issue... If don't think that they'll
    mind, why didn't you ask them?
    Matti Vuori, Oct 9, 2006
  8. RichA

    Father Kodak Guest

    Thanks. I'm glad I didn't spend too much time looking on the website.
    That article could be summarized as follows:

    1. Use of an actual camera body to test lenses reduces the accuracy of
    the measurements, because the camera body introduces errors of its
    own. We at Pop Photo use a lens bench to eliminate those issues.

    2. Other people haven't made the same investments nor do they
    understand the issues. That is why our tests are the best around.

    3. We at Pop Photo have invested more than anyone else in hardware
    and software tools based on a lens bench.

    Well my reactions are:

    0. What a letdown! I so was disappointed at this "article" that I
    rushed out to cancel my subscription, and then I remembered that I
    stopped subscribing at least 20 years ago. Obviously no reason to
    re-up with them.

    1. Duuh. Duuuuuh again. Big, big duuuuh. Of course. That's true
    for any lab test. Science Lab 101, people. (This guy would be a whiz
    at writing a padded term paper.)

    2. So freaking' what. If I want to take a picture, I actually have to
    use a camera body (digital or film). So for my purposes, the
    effective lens resolution has to consider the camera body.

    3. I don't believe that they are really better until they document
    their claim of best-in-class. What are their hardware and software
    components? what is their procedure (workflow)?


    4. Why not publish some comparisons between Pop Photos' tests and
    Magazines A and B , and Web Sites C and D to prove their point.
    Otherwise it is just an ad for the magazine, as another poster said.

    To repeat my O. point. What a letdown!!!

    Father Kodak
    Father Kodak, Oct 12, 2006
  9. RichA

    RichA Guest

    I think the validity of the lens testing being done on a machine is
    that it gives you absolute results, independent of the camera body. In
    the real world, that may not matter, but it could matter if current
    sensors are capable of showing the shortcomings of mediocre lenses and
    I think most sensors in DSLRs will allow you to see if you have a good
    or a bad lens.
    Testing on a camera body could cause....problems, as two different
    lenses on two different cameras could make the lesser camera seem to
    produce better results.
    However, they said their tests with camera and lens are done using high
    quality primes as opposed to questionable cheap zooms that most cameras
    get kitted with.
    RichA, Oct 12, 2006
  10. RichA

    Father Kodak Guest

    Sure. But that doesn't match up too well with the real world, in
    which you actually need a camera body to hold a sensor/film in order
    to record an aimage.
    How is that possible? Image resolution is roughly

    1/f (image resolution) = 1/f (lens) + 1/f (sensor/film).

    Decrease sensor/film resolution and you decrease image resolution. How
    could it be otherwise? Otherwise a Lens Baby would give you higher
    quality images than the best Nikon and Canon lenses. Which I
    seriously doubt.
    OK, but again in the real world, you can only describe lens
    performance with a camera body. There should be a 'reference' camera
    body for each major product line, and all tests can be keyed off that

    Of course, every few years they might need to upgrade that body. In
    theory, they should then repeat all their tests, but somehow I don't
    think that will happen.

    Father Kodak
    Father Kodak, Oct 21, 2006
  11. RichA

    RichA Guest

    Be hard to do with Canon. Use a cropped sensor model, the people using
    FF would say the test is invalid and really it would be. Use a full
    frame, and you might end up with poor edge definition that might not be
    as bad in a cropped camera.
    RichA, Oct 23, 2006
    1. Advertisements

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.