EOS 5D MkIII Focusing

Discussion in 'Canon EOS' started by Gary Holliday, Oct 28, 2017.

  1. I reported a focusing problem on a previous thread. I've since had the body sent off for a warranty repair to the CMOS:

    "Dismantle to reset body depths, full reprogram and re-calibration to return to standard."

    I then sent the repaired body off to a Canon approved centre to calibrate my 135mm L 2.0 lens to the Mark III.

    I'm not convinced that this camera is pin sharp...what do you think of the three images attached?

    A few shots taken on an overcast day a moment ago.
    400 ISO
    250s - f8
    JPEG.
    No additional sharpening in Photoshop

    View attachment 5S9B0854.JPG

    View attachment 5S9B0860.JPG

    View attachment 5S9B0858.JPG
     
  2. Looks sharp enough for me :)
     
  3. Really? I'm looking at the brick photo; the red coloured bit left of centre but the image looks like a confused mess to me. I really need some opinions as I'll have to return the body this week. This doesn't look pin sharp at all.
     
  4. It looks like a wall full of bricks to me... even at 100%. The red bit doesn't look quite right, but it's a chunk of missing brick, so there's no real telling how textured the surface is (or isn't)... Of more concern might be image 2, especially if the point of focus was the center flower.
     
  5. Look just fine to me. Picture 2 is difficult to tell because we don't know which flower you focused on (do you know?) but it still looks good. I'd just go out and start taking photos.
     
  6. The single centre focusing was selected so you can see which flower by the composition. To me there's no sharpness whatsoever. I think this body is going back.
     
  7. Why you trying focus test with F8? I am pretty sure you can google, how to properly test autofocus. Because on shots you provided , it could be camera focus error, lens focus error or camera operator error, you peak one you like most:)
     
  8. Eh? Nothing wrong with f8.
     
  9. F16 will be sharper:)
     
  10. The pictures (flowers aside) don't look alarmingly bad to me.
    I wouldn't trust my handholding skills at 1/250 with a 135mm.
    I also wouldn't expect AF to visibly miss at f8.
    Mine doesn't work great at f2.8. - I'm currently reading up on AF micro adjustments. If your service center really set them note down the value they picked and do a t least some testing. - Doesn't Magic Lantern perform automated adjustments on your Mk III?
     
  11. The single centre focusing was selected so you can see which flower by the composition

    You might have used focus and recompose. If the central flower was the point of focus, then your AF is off. Did you take more than one: are they all like this? If so, then you need to adjust the AF. If you were not using a tripod then you could be rocking back and forth just enough for it to be off (AI servo deals with that), or the wind could be blowing and moving the flower out of the plane of focus. You need to consider these other possibilities too before you decide the AF is off.
     
  12. Looked at EXIF on the first photo.

    Camera: Canon EOS 5D Mark III
    Lens:
    Canon EF 135mm f/2L
    Exposure:
    Manual exposure, 1/128 sec, f/8, ISO 400

    1/128th second according to the embedded exif data in the photo.

    At only 1/128th sec I wouldn't rule out camera shake.

    Try doing the test again with a tripod.
    The Canon 5D Mark III can do micro adjustments and save those adjustments data for a specific lens.
    Here is a guide on how to perform this test and task and save yourself some money by doing it yourself.
    LINK
     
    Last edited: Nov 1, 2017
  13. Agree, you need 1/250th or up for the 135L, better 1/500. One of the reasons I no longer have a 135L
     
  14. If camera shake were the issue, then ALL the flowers would likely be smeared. You shouldn't have sharp ones behind the plane of focus. That makes zero sense. unless a random breeze puts those particular flowers moving in the exactly the same direction as the camera at that particular moment... possible? yes. but also extremely highly improbable? yes.

    That center point one shot was used means that it is likely there is a fault. Redoing w/ a tripod seems the obvious next step. Also redoing at f2 would be a good idea, since shooting at f8 covers up any faults by a good degree.

    At 135mm I often have no problem (even w/o IS) pulling crisp shots out @ 1/125. However, putting it on a tripod is critical to do an objective measurement of AF accuracy, as well as shooting near and far stationary subjects.
     
  15. I agree with you on the flower shot. My comments were based on the first photo, brick shot. One problem for us to Monday morning quarterback this on the flower shot, where was the original focal point when the camera achieved focus lock. Was it a focus and recompose? We can't be sure.I agree I have achieved good focus at 1/125sec. But I can not be sure of the OP, this is 135mm, how far was he from the subjects, how steady are his hands for the first shot. Did he focus recompose on the flower shot? If he is going to do this right, set up a tripod and do a real test. There are ways to display the focal point to share in a post like this. Gary please follow up with better test shots off a tripod and what you found, was it actual front or back focus issues or just shooting error.
    FocalPointExample.JPG
     
    Last edited: Nov 5, 2017
  16. I probably uploaded the wrong photo out of the test photos. The first shot would have been that shutter speed, but then I moved up a stop to 1/250s.

    All my shots are not sharp, even subjects that don't move. my handheld technique has always been good, hold camera tight to face, don't breathe or move, wait for wind. My critical indoor studio tripod shots are also not 100% sharp.

    The retailer has been good, so will receive a new body later this week. I'll upload some comparison shots.
     
    Last edited: Nov 8, 2017

Share This Page