I give up with the focusing thing!

Discussion in 'Nikon DX DSLR' started by Czechman01, Jul 30, 2005.

  1. Although I wasn't noticing any OOF shots with my D2x I was seduced by all the focus hype on the Internet so I decided to test mine at all 11 focus points.

    I did the Tim Jackson focus test and was careful to do the test exactly as described, and I repeated the results with no less than 3 Nikkor lenses. Sure enough, several focus sensors appeared to be off, some by quite a bit. Why hadn't I noticed this in my work? Hmmmm....

    Two weeks ago I sent my D2x to Melville to have the focus corrected.

    It came back yesterday and today I re-tested all 11 sensors and guess what? Some are still off and some were off by quite a bit. Man, was I steamed! I spent $85 to send this thing to NY and they returned it still screwed up? @#$(&%!!!

    Then I recalled that there was some discussion about the Jackson test and how having the camera at an angle wasn't quite correct for some reason. It's been suggested that the target should be at 45 degrees and the camera level instead of the other way around.

    So, why not? I fastened the Test to a piece of lumber so I could set it up on it's edge and I used a machinist's square to set it at a 45 degree angle on the table. I then positioned the camera so it was level and aimed straight at the target and re-ran the test.

    Guess what? A remarkable improvement! It's not "perfect" in the far right and left sensors but now I have 7 out of 9 that are spot on instead of 2 running the test the other way.

    So what does this prove? For me it proves that if the thing works under real-life working conditions, but fails a "test" then the test is screwed up!

    I'm not saying that there aren't SOME cameras with legitimate focusing problems, just as there are some cameras with any other problem but it's VERY possible that this entire focus issue has been blown completely out of proportion by this stupid Jackson test.

    The hell with all this focus stuff. If it hadn't been for the hysteria on the Internet over this "issue" I'd have never wasted $85 to have my camera "repaired" when there was nothing "wrong" with it in the first place.

    I love this camera and I'm letting everyone know right now that if they come up with something else to complain about, I'm looking in their camera bag! $10 says they have Canon in there.

    Woody
     
  2. How real world shots come out is all that really matters anyway.

    I never would have done the tests if I hadn't seen problems there. And the images I sent to Nikon were ordinary in-use photos, not tests.
     
  3. Beezie,

    I know, I saw the pictures you posted. You have a legit complaint and I sincerely hope you get what you paid for. Your situation had nothing to do with my paranoia.

    My point is that the Jackson test is no true test of the camera's focusing ability and sooooo many people are doing it under all sorts of conditions and assuming they have focusing problems and then going on these discussion groups and trashing the camera.

    One thing I failed to mention in my original post was that after I shot the Jackson test I shot the eye of a mannequin (I live alone) at all 11 focus sensors and all but the extreme right and left sensors were perfect.

    So, I guess if anyone is shooting thick black lines on white paper they may have a problem.

    Woody
     
  4. I need to get me a mannequin. I can't get my three year old to sit still unless he is asleep. ;)

    PS, you should have seen the adventure mine went through last week.

    I sent it to the Torrance address the DAY before they moved to El Segundo. And don't you know, the delivery folks didn't forward it until it bounced twice.

    I was worried it would sit in a post office (I used overnight Express Mail) for a week then be returned, but then they suddenly figured it out and forwarded it to El Segundo.

    THEN (argh) it bounced there and Nikon had to go to the post office to pick it up.

    I think I will drive it down myself next time. ;)
     
  5. Woody,

    Thanks so much for your post and test!

    Paul
     
  6. Steve S

    Steve S

    Feb 1, 2005
    SE Florida
    After my recent photo shoot from yesterday, I'm worried (pic

    nearly all my 200 shots were taken in portrait mode, using the "right center" focusing reticle, and a high percentage of them show signs of missed focus point, with either slight FF or BF. I used the "focus then recompose" method. My 1st body, a very early example, was send back into Nikon and has been performing much better since it's servicing, but it looks like this one's going in too.
    This shot was taken with my 28-70 wide open @f2.8, and reduced to 25%, and sharpened for web. Note how her left hand area is in perfect focus, but not her face. What are your thoughts?
    original.
    Date/Time 31-Jul-2005 07:56:00
    Make Nikon
    Model NIKON D2X
    Flash Used Yes
    Focal Length 40 mm
    Exposure Time 1/320 sec
    Aperture f/2.8
    ISO Equivalent 200
    Exposure Bias +1/3
    White Balance (-1)
    Metering Mode multi spot (3)
    JPEG Quality (6)
    Exposure Program aperture priority
     
  7. Re: After my recent photo shoot from yesterday, I'm worried

    Hi Steve,

    If I understand you, by "focus and recompose" you mean that you first focused on her eyes in portrait mode (with the right center reticle), and then moved the camera to ensure the composition you wanted??

    If that is so, then you very well may have moved the camera enough to throw the eyes out of perfect focus, since dof at f/2.8 is so narrow.

    Make sense??

    Or perhaps I've misunderstood what you did.

    Best wishes,

    David
     
  8. Steve S

    Steve S

    Feb 1, 2005
    SE Florida
    Yes, that's exactly what I did, but,

    How could recomposing the shot have "moved the camera" enough to throw the eyes out of focus? Is the DOF at 40mm focal length and f2.8 that razor thin? You do see where it's missed it's mark here, right?
     
  9. Re: Yes, that's exactly what I did, but,

    Yes, of course I see what you are referring to in the shot.

    As far as my point, it's a question of geometry (trigonometry?). Somewhere there's a great diagram that illustrates the point. Here's a VERY rough attempt:


    1. [camera lens] ---------> Subject

    2. [camera lens] ----------/

    See what I mean? (despite the poor diagram). Distance from plane of lens to plane of subject is the same, but distance from plane of lens to POSITION of subject is altered once you move the camera body. This alters the point of focus.

    Perhaps someone else can explain (and illustrate) this better.

    Best wishes,

    David

    P.S. Very sexy model, by the way.
     
  10. Ruffles

    Ruffles

    482
    May 5, 2005
    Rexburg, ID
    I go back and forth on this. Every time I convince myself that nothings wrong and that I've taken enough "controlled enviorment" test shots (stuffed animals etc., no paper test targets) to prove that I can get good focus with each sensor, I go and take a real world shot and it ends up OOF. Here's an example from this weekend...

    [​IMG]

    I guess I need MORE testing to figure out how to get this thing to work :(.
     
  11. Steve S

    Steve S

    Feb 1, 2005
    SE Florida
    I get it, but I don't get it David

    I get what you're saying, and if the DOF was razor thin, (which it isn't @44mm & f2.8), such repositioning could've made a small difference in the focus distance. I hope others will weigh in on this subject. All I did was pivot my camera the slighest bit to get her out of the center of the frame. Maybe, just maybe if I was shooting my 50mm @f1.4 I could agree with you, but not at my 28-70 f2.8.
     
  12. Steve S

    Steve S

    Feb 1, 2005
    SE Florida
    That shot's pretty clearly back focusing, Steve

    Sort'a hard to tell with the water in the bg, but since I don't see Anything in focus in front or at the point of your focus. I'd send that one back.
     
  13. Thanks, Paul.

    That is a good link, and is exactly the issue I was suggesting to Steve.

    Steve, I don't know, of course, whether this explains the results you are seeing.....but it's worth thinking about. (Note the point in the link Paul provided that the focus/recompose error increases if you are shooting at closer distances to the subject. So, if you were shooting around 44mm, I assume you were within 15 feet or so of the model??)

    Best wishes,

    David
     
  14. gvk

    gvk

    388
    Jun 17, 2005
    Mystic, CT
    To add a bit of quantitative analysis ...

    Assuming that the young lady is around 5 1/2 feet tall, and accounting for her 3 inch heels and about 6 inches from the top of her head to the edge of the frame, I deduce that the frame long dimension at the subject plane is likely a bit over 6 feet. The angle of view in the long dimension for an APS sensor with a 40 mm focal length lens is around 32.8 degrees, resulting in a camera to subject distance of about 3.25 m, or 10 2/3 feet. Assuming a COC of 0.0167 mm, the depth of field at f/2.8 is 0.9 feet in front and 1.1 feet behind the focused plane.

    Analyzing the geometry of the described focus techinque, i.e. recomposing after focusing on the face with the "right center" bracket, gives a shift in the plane of focus proportional to the camera to subject distance, and the factor (1 - sin a/tan a) where "a" is the angle that the camera was rotated during recomposition. An upper bound on this angle is half of the narrow dimension angle of view, about 11.1 degrees. This results in a maximum focus error of only about 2.4 inches.

    It is difficult to say for sure, but based on the model's pose, her hand is probably a couple of inches behind her eyes, and thus close to the actual plane of best focus. Clearly this small focus shift is not sufficient to explain the entire perceived focus problem. However, using the corner focus bracket would minimize the shift.

    I don't have any direct personal experience with the 28-70 f/2.8 zoom lens, but all of my zooms are a bit softer at the edge than in the center when used wide open. Could edge softness be a contributing factor to this sharpness problem?

    Gerry
     
  15. 44mm seems kind of thin to me given that scene. It is easily that far from her right hand to her eyes. I could see how turning the camera from centered on her face to where we see it above would land her face at the back edge of that DOF.

    One thing I would try is fire off a shot before you recompose, then recompose and fire again. Then you have something to compare to after.

    A problem I have had now and again is a tendency to want to approach the subject as I focus. And by the time I release the shutter I have moved in a few inches. Argh.

    I keep reading how the pros use the AF On button and turn off shutter button AF. But I never have understood why they do that.

    Just random thoughts. You won't get any doubt from me as to the possibility that the camera needs calibrating. I get a feeling the fourth time is the charm though, and I look forward to getting my camera back THIS WEEK PLEASE. :lol:
     
  16. Steve S

    Steve S

    Feb 1, 2005
    SE Florida
    "...depth of field at f/2.8 is 0.9 feet in front and

    1.1 feet behind the focused plane." Well, thanks for the scientific analysis. That tells me that my ever so minor *lateral* shift from recomposing should not have made any difference. But, as you also pointed out, the corners may be a bit soft at wide open on this lense. I still think it needs to go in for focus calibration of the outboard focus reticles.
     
  17. Steve S

    Steve S

    Feb 1, 2005
    SE Florida
    Interesting link, Paul, *but*,

    in his illustration, he focused on the subject's belly button, then moved the camera upward to her face. I only moved ever so slightly laterally. There's a significant difference there. Plus, according to Gerry's analysis, my DOF is much greater than the small amt of camera movement I could've induced. So, I'm not at all convinced I could've moved out of the focus plane. And yes, David, I was approx 12-15 ft away.
     
  18. Re: Interesting link, Paul, *but*,

    Hi Steve,

    Yes, I'd have to agree, given Gerry's math, and especially if your movement was lateral.

    Edge softness? Gosh, I don't know. I doubt that is what we're seeing here. But maybe.

    It will be a shame if you have to send your camera off for adjustment, but I suppose it may come to that.

    In the meantime, why not do some more portrait testing, and see if your results vary with different reticules.

    Best wishes,

    David
     
Loading...
Similar Threads Forum Date
Question on a4 and a5 setting for D5600, Rangefinder and MF Focus Nikon DX DSLR Aug 18, 2017
D7200 focus problems Nikon DX DSLR Apr 5, 2017
D500 Focusing Setup for Sports Nikon DX DSLR Nov 14, 2016
D200 focusing without me touching a thing ???? Nikon DX DSLR Oct 23, 2010
The focusing thing comes to an end Nikon DX DSLR Aug 11, 2005