Results 1 to 11 of 11

Thread: My LensAlign MkII Results - Opinions?

  1. #1
    BPN Member David Pugsley's Avatar
    Join Date
    Dec 2009
    Location
    Central Florida
    Posts
    255
    Threads
    65
    Thank You Posts

    Default My LensAlign MkII Results - Opinions?

    After spending far more time than I probably needed to I've come to the following results with the LA MkII.

    For my naked 500 I've gone from a previous setting (determined by prior "home-grown" tests) of +10 to +5.

    Here is one of the best images of the series. I found my biggest problem was getting a series of enough consistent results to verify the best MA.



    For the 500 +1.4x I went from a +5 to a +11. Apparently I've been front focusing.

    Again, here is a representative image.



    Note that both images are 100% crops of SOC files.

    Does anybody see anything obvious that I'm not? Good, bad or indifferent? The real test will be in the field tomorrow.

    Cheers and sharper images!
    David

  2. #2
    BPN Viewer
    Join Date
    Sep 2008
    Location
    London/Essex, UK
    Posts
    92
    Thank You Posts

    Default

    Maybe it's my eyes, but....

    The first thing that strikes me is how soft the images look. I'm surprised you can determine anything with much precision. I don't have a 500/4 and probably never will, but for a highly praised prime lens (I assume you're talking about Canon, but maybe that assumption is wrong) I would hope to the images looking sharper than that, somewhere in the frame at least. I'm really struggling to work out exactly where the image is sharpest.

    How solid is your tripod and head? Is the tripod on solid flooring or carpet? Are you using mirror lockup? Are you using a remote release? What are your sharpening settings for the files? Are you shooting raw or JPEG? Which camera is it? Can you get sharper results with Live View and Live AF? Do you have a filter in place? Is it of good quality and clean? The files have no EXIF. What are your exposure settings? What is the ISO? What NR is being applied?

    It is possible that there will be small variations in AF performance from shot to shot. Manually defocus towards infinity between each shot. Take several readings and average the results.

    If you can hook up to a computer for remote shooting you might also look into Arash's technique for focus nudging to determine how much adjustment is required.....

    http://www.stanford.edu/~ahazeghi/Photos/MA-web/

    Again I'm assuming Canon gear (well, the camera this time) for this. If it's something else...... :(
    Last edited by Tim Dodd; 02-18-2011 at 05:03 AM.

  3. #3
    BPN Member David Pugsley's Avatar
    Join Date
    Dec 2009
    Location
    Central Florida
    Posts
    255
    Threads
    65
    Thank You Posts

    Default

    Maybe it's my eyes, but....

    The first thing that strikes me is how soft the images look.

    I agree they're soft. I think I can attribute that to two reasons.
    1. These are, as I said, SOC (straight out of camera) with no sharpening etc. Just RAW files into LR3 and exported as JPGs.
    2. The images are heavy crops to just show the target. I haven't done the math, but with my 10MP MkIII we're only looking at 3-4 MP on target so there's not a whole lot of resolving power. I'm no tech guru so I may be WAY off base here...just my thoughts.

    How solid is your tripod and head? No worries
    Is the tripod on solid flooring or carpet? Outside
    Are you using mirror lockup?No
    Are you using a remote release? No
    What are your sharpening settings for the files? Are you shooting raw or JPEG? See above
    Which camera is it? 1D MkIII
    Can you get sharper results with Live View and Live AF? No live-view AF with MkIII
    Do you have a filter in place? No
    Is it of good quality and clean? NA
    The files have no EXIF. What are your exposure settings? 1/3250 f/5.6 and 1/8000 f/4 What is the ISO? 400
    What NR is being applied? None

    It is possible that there will be small variations in AF performance from shot to shot. Manually defocus towards infinity between each shot. Take several readings and average the results. Did this

    I basically used BAA recommendations for camera settings and target set-up with the exception of tethering as I can't AF with Live View.

  4. #4
    BPN Viewer
    Join Date
    Sep 2008
    Location
    London/Essex, UK
    Posts
    92
    Thank You Posts

    Default

    Lightroom applies sharpening by default to raw files and my experience with my 1D3 tells me that at the pixel level I can get very sharp results from that camera when I do my job right and view in Lightrrom with no further edits. Here's a 100% crop example of a moving subject taken with my 100-400 handheld without IS at 190mm and only 1/500, left at Lightroom defaults.



    The shot could do with an adjustment to levels and for use at 100% it might benefit from a little more sharpening, but to my eyes it looks at least as sharp, if not sharper than your static target shot with a prime lens mounted on a tripod. Something seems not to add up.

    Again, maybe it is my eyes.

  5. #5
    BPN Viewer
    Join Date
    Jan 2008
    Posts
    157
    Threads
    9
    Thank You Posts

    Default

    Quote Originally Posted by Tim Dodd View Post
    The shot could do with an adjustment to levels and for use at 100% it might benefit from a little more sharpening, but to my eyes it looks at least as sharp, if not sharper than your static target shot with a prime lens mounted on a tripod. Something seems not to add up.

    Again, maybe it is my eyes.
    Tim, I don't think you can make any reliable comparison of the relative sharpness of your image to the lensalign test images. They are obviously made from very different distances with very significant differences in depth of field.

  6. #6
    BPN Viewer
    Join Date
    Sep 2008
    Location
    London/Essex, UK
    Posts
    92
    Thank You Posts

    Default

    All very well, but at the plane of focus, no matter how great or little the DOF, the image should be sharp. In AF calibration the DOF is really secondary to identifying where the point of sharpest focus (i.e. focus) has been achieved. Of course a shallow DOF helps it easier to pinpoint where that sharp plane is, whereas a generous DOF makes it trickier.

    But that wasn't really my point. My point was that a file from a 1D3, when processed through Lightroom at default settings should be able to produce sharp images. David is agreeing that the images seem a little soft and is saying that the reason is (a) because the files are SOOC without sharpening; (b) these examples are crops. Well (a) assuming Lightroom is at default settings then the images are sharpened; (b) my example is a 100% crop. In other words I'm not sure David's reasoning stands up.

    Anyway, I can't contribute anything further to the thread so I shall bow out. David, I hope you get things sorted.

  7. #7
    BPN Viewer
    Join Date
    Jan 2008
    Posts
    157
    Threads
    9
    Thank You Posts

    Default

    I do my lens resolution tests with a USAF resolution target. I've found that I often have difficulty determining differences in resolving power of different lenses unless the comparison images are taken at the same degree of magnification. In other words, differences in image magnification can account for the differences in ability to distinguish separation of horizontal and vertical line pairs. It gets even more difficult to make determinations when the comparison images are of totally different subjects in different lighting conditons. There are other factors involved in the apparent sharpness of an image. Personally, based on my experience in controlled test situations, I couldn't make a determination that Tim's image was any sharper than David's test images.
    Additionally, the LensAlign is not a resolution test, but a test to determine whether the camera/lens combination is front- or back- focusing. So to get back to David's question, it appears that his test images do provide the info he needs to make the adjustments. My only concern would be his mention of difficulty in getting repeatable consistent shots. I read that Michael Tapes, the LensAlign developer uses flash to eliminate any camera vibration issues. He places the LensAlign in his garage (protected from wind) with remote flash units. The camera/lens is mounted on a tripod in his driveway.

  8. #8
    BPN Viewer
    Join Date
    Dec 2007
    Location
    Cary, North Carolina
    Posts
    789
    Threads
    64
    Thank You Posts

    Default

    Quote Originally Posted by David Pugsley View Post
    Does anybody see anything obvious that I'm not? Good, bad or indifferent? The real test will be in the field tomorrow.
    David
    Hi, David,
    The second image looks like it's just about perfect--with the depth-of-field balanced on either side of the "0" mark. I would note that you had the target ever-so-slightly out of perfect alignment. The red of the central bull's eye should be centered within the opening, but in yours there's a little more "red" on the left side of the viewport than on the right. That said, you probably were close enough that your micro-adjustment was good. It's just a note for future reference.

    Also, in the image on top, you were not front-focusing, but back-focusing: when the numbers on the top of the ruler are sharper than those on the bottom, you are "back-focusing," and when the numbers on the bottom half of the ruler are sharper than those on the top, you are "front-focusing."

    As to the comments about the "sharpness," I'd like to say that I've actually been using these products for a few months now as a reviewer for Michael Tapes Design and this is actually very representative of how it looks at 100%. Keep in mind that you don't need to process the files--this is one time when you can just look at JPEGs right out of camera. No levels, no curves, no sharpening necessary. The only processing you might do is to bring the photo into Photoshop and use the Emboss filter to help evaluate where, exactly, the DOF begins and ends.

    You can read more about that in my reviews of the LensAlign Pro and the LensAlign MkII. Cheers!
    David
    Last edited by David Kennedy; 02-23-2011 at 02:44 PM.

  9. #9
    Forum Participant
    Join Date
    Jul 2009
    Location
    CA Central Coast
    Posts
    311
    Threads
    25
    Thank You Posts

    Default

    David,

    I struggle also with both the original LensAlign and MkII to get unambiguous results. Not condemning LensAlign, I just think that slight errors are really hard to see. So don't take this reply as great expertise.

    I interpret your first 2 images as showing a very slight amount of front focus. The front 6 and 8 on the right look sharper than the back 6 and 8. They ought to be at least the same, or the back even a tiny bit sharper. The latter if the usual rule that DOF extends further to the back than the front is visible on the ruler.

    Alan

  10. #10
    BPN Viewer
    Join Date
    Dec 2007
    Location
    Cary, North Carolina
    Posts
    789
    Threads
    64
    Thank You Posts

    Default

    Quote Originally Posted by Alan Lillich View Post
    David,

    I struggle also with both the original LensAlign and MkII to get unambiguous results. Not condemning LensAlign, I just think that slight errors are really hard to see. So don't take this reply as great expertise.

    I interpret your first 2 images as showing a very slight amount of front focus. The front 6 and 8 on the right look sharper than the back 6 and 8. They ought to be at least the same, or the back even a tiny bit sharper. The latter if the usual rule that DOF extends further to the back than the front is visible on the ruler.

    Alan

    Hey, Alan,
    I think you're right that sometimes it is hard to figure out what, exactly, is going on. After working with these things for a while I've noticed that it's actually easier to "read" the LensAlign using the smaller columns of numbers on the left. That's why I actually disagree with your assessment: if you look at the left-most column of tiny numbers in the first image, I can read (somewhat) clearly up to "21" on the top half of the ruler, but only down to about "12" on the bottom half. I should note that, yes, I can tell what some of the more "distinct" numbers are lower down on the bottom half, such as "14," but I can't for the life of me read "13," so I don't count it. This tells me that I can read more numbers on top than on the bottom, so it is back-focusing.

    In the second image, now that I'm looking at it a little more closely, I might argue it is still slightly back-focused, as I can read the whole series of numbers up to about "13" or "14," but only to about "11" or "12" again on the bottom half. Sliding over to the second-smallest column of numbers, the top "36" is definitely a tad more legible than its bottom counterpart.

    However, as you observed yourself, if one only evaluates the "large" numbers, these differences are less distinct. I hope this helps give you a new look at how to use LensAlign!

  11. #11
    BPN Member David Pugsley's Avatar
    Join Date
    Dec 2009
    Location
    Central Florida
    Posts
    255
    Threads
    65
    Thank You Posts

    Default

    Thanks for the input folks. FWIW, here are the embossed images.

    Naked 500


    500 +1.4


    Real-world results on it's first test were promising, but I'm always willing to tweak further to get the best possible results.

    Cheers
    David

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  
Web Analytics