Focusing issues EOS 1Ds III

My trusty EOS 5D II was stolen a few years back and so I recently bought a used 1Ds III - a great camera except the AF seems to be less reliable: I took some shots of our brick wall to demonstrate and, as is typical for this kind of shot, 2 out of 8 are not sharp at 100% magnification (see attached photos). I would say that about 20% of my shots of still subjects are not in focus and I don't recall having this kind of problem with my 5D. I have microadjusted the AF but the problem is inconsistency in both directions so this does not help. Using live view resolves the problem so the problem is definitely with the AF, not the lens. Is this to be expected and does anyone have any idea if this kind of problem is fixable and at what cost? Thanks.

 

 

 

 

 

 

 

 

Can someone please tell me what focusing accuracy I should expect from an EOS 1ds III on still subjects using single focus AF? How many shots should I expect will miss focus. It is an old camera and maybe the AF become less accurate over time.

what settings did you use?

Did you use a tripod? cable release?, 2-sec delay?, mirror lockup?

Are you in earthquake area? (subject movement) have to rule out everything :-)

 

I know my 5d2 is acurate in "one shot" with just central point, and "ai-servo", but not in "ai-focus"

In AI-focus it is sometimes way off.

 

Testing is a profession, too many variables.

J. (used to test software..)

. . . AF seems to be less reliable: I took some shots of our brick wall to demonstrate and, as is typical for this kind of shot, 2 out of 8 are not sharp at 100% magnification (see attached photos). I would say that about 20% of my shots of still subjects are not in focus . . .

 

What AF Mode was selected?

 

If “AF – One Shot” was selected I would expect 100% AF results for shots of a brick wall in hard light.

 

On the other hand, if “AI Servo AF” was selected then I’d expect about 66.6% success, based upon a series of tests conducted by myself and two other members here at PN, several years ago.

 

WW

Thanks! I used AF - One Shot! Handheld, fast shutter, lots of light, standing still. I don't remember having this kind of problems with my 5DII but that was some years back before it got stolen from my car and I never bother testing it - it just worked. Do you have any idea about what it will cost to get fixed or if it is even feasable?

Did you get Auto Focus Confirmation for every shot?

 

What was the distance from wall to camera?

 

What AF point was used?

 

What images are not in sharp focus? (use the numbers please)

 

WW

Edited March 24, 2019 by William Michael

The red autofocus square is always on. I set it this way in the camera. I could try and change it and see if it makes a difference. C FnIII is now set to ON (when focus achieved). Edited March 24, 2019 by johnv

I think it is a good idea to see the AF confirmation.

 

... answers to my other questions please???

 

PS - I might have to get to bed before I get to reply again, it's late here.

I'm using the AF-ON button to focus so I'll try to change that as well so that I focus using the shutter button.

Sorry, yes, I did get AF confirmation for every shot.

That is, it the AF square lights up for a moment and then goes away again. I tried it again coupling the AF to the shutter button, testing with 3 lenses, EF 50 1.8 STM, EF 70-200 2.9 IS II and EF 35 2.0 IS - all wide open. The 50mm had about 20% out of focus while the 35mm and 70-200 had about 10-15%. Anything else i can try to do? Thanks! Edited March 24, 2019 by johnv

Answer to questions:

 

Did you get Auto Focus Confirmation for every shot?

Yes.

 

What was the distance from wall to camera?

About 1 meter. I have tried everything from close-ups to far away. Most shots have been about 1 or 3 feet from wall.

 

What AF point was used?

Center AF only.

 

What images are not in sharp focus? (use the numbers please)

4841 and 4843

I just did another test, this time with my EF 80 1.8 USM and all 20 shots from about 3 feet were in focus and only 1 out of 9 distance shots were not sharp. I really can't explain it. I've spent hours doing these tests and then suddenly a lens is almost perfect.

 

I see some parts of the images 4841 and 4843 are sharp but the center is not. Maybe something is off with this lens or maybe the camera was not completely perpendicular to the wall - I don't know. In real life situations using mostly the 70-200 and the 35mm (the lenses I have used the most) I do have more misfocuses than I would like. Maybe some lens and camera combinations work less well together - this would make sense for my Sigma lens but not the Canon lenses (as mentioned earlier, I've tested all lenses for microadjustments but that has not fixed the problems). Any suggestions about what to try out will be very much appreciated.

Thank for responding to my questions.

 

Obviously I can only respond in someway approximating an analysis, using only the hard data supplied - i.e. Images 4841 and 4843 and the procedures used for that batch of tests:

 

Image 4841 downloaded and viewed at max resolution on my Studio Monitor: scrolling through the horizontal centre of the image from top to bottom, the top of the image appears in acceptable focus and the bottom is not so acceptably in sharp focus, for Image 4843 the opposite is true.

 

At 1000mm Subject Distance, Focal Length Lens 50mm, Aperture F/1.8 using CoC = 0.025mm the Depth of Field is about 34mm: I haven't done the trigonometry, but I think it obvious that you'd only need an hand held camera to be tilted ever so slightly rendering the Plane of Sharp Focus not parallel to the Sensor Plane to notice these effects. Obviously as you mentioned Subject Distances closer than 1000mm, then for these shots the DoF is narrower.

 

Those two images constitute the samples which are unacceptable, but, in all fairness, the testing procedure that was used, is suspect and (although a very small sample group) IMO, 6 out of 8 test shots coming back as 'acceptable' when using a suspect, in field testing procedure, would satisfy me that my gear was working OK.

 

WW

Edited March 24, 2019 by William Michael

I just did another test, this time with my EF 80 1.8 USM and all 20 shots from about 3 feet were in focus and only 1 out of 9 distance shots were not sharp. I really can't explain it. I've spent hours doing these tests and then suddenly a lens is almost perfect.

 

I see some parts of the images 4841 and 4843 are sharp but the center is not. Maybe something is off with this lens or maybe the camera was not completely perpendicular to the wall - I don't know. In real life situations using mostly the 70-200 and the 35mm (the lenses I have used the most) I do have more misfocuses than I would like. Maybe some lens and camera combinations work less well together - this would make sense for my Sigma lens but not the Canon lenses (as mentioned earlier, I've tested all lenses for microadjustments but that has not fixed the problems). Any suggestions about what to try out will be very much appreciated.

 

When we get to "real life situations" there are many more variables.

 

One element which we might deem as 'less of a variable' would be if you were scrutinizing all the shots that you made with you 5DMkII with the same veracity as the shots with the 1DsMkIII and then making A/B comparisons of images made in similar conditions with the same lenses.

 

On the face of what has been presented, it seems more likely than not that your gear’s AF is working within specification.

 

WW

Thanks, William, for your thorough thoughts on the matter. I did see the focal plane issues as you described but the thing is that the centre AF point is not in focus. I guess I would have to use a tripod to make sure I did not move ever so slightly. I haven't compared the numbers of in-focus shots compared to my 5D II - but then again I'm using different lenses, mostly, and I'm out of practice using AF compared to back then so that wouldn't count, really. Mostly, with the 5D II, I used the EF 35L and 85L and their AF accuracy were both excellent and easy to use - easier than for instance the EF 85 1.8 USM in most real life circumstances - that is how I remember it, anyway. I will do an AF focus test later today with my tripod and the EF 35mm 2.0 IS USM and share my findings. Thanks again.

As a general comment: I have done a lot of 'field tests' regarding with Canon EOS Digital gear and many of these tests were 'addressing' AF 'issues'.

 

Importantly many of these tests were done in concert with two other Canon users, both members here, both very experienced photographers. (We also 'field tested' oddment things like how 20D centre AF square is slightly lower than the actual target point and if there is (useful) a constant differential between Metering Modes in typical lighting scenarios).

 

Anyway, my point is about: 'field testing' - I think that it is important to remember that, (because) there are variables in 'real world' or 'field testing', then, if you get a 'reasonable' approxiimately 80% consistency, then that is probably the answer.

 

Additionally, (as a general rule), it is safe to assume, as an initial position that the gear is in specification.

 

Also, I think that the bottom line in the real world is always "Does it affect the Final Product?" - now that was indeed the case when two of us inspected a particular Sigma 28mm Lens: it was indeed Back Focusing on 5D Series Cameras and that was indeed affecting the Final Product because that particular lens was typically used to make the Wedding Cake; Champagne Flutes; Wedding Ring etc., images using Shallow DoF Technique: so, as an alternative to dumping that lens a compensation could be made when using Shallow DoF at a close Subject Distance; on the other hand, the Back Focusing issue didn't matter one hoot when the lens was used for a Group Portrait.

 

I hope that assists.

 

WW

. . . Mostly, with the 5D II, I used the EF 35L and 85L and their AF accuracy were both excellent and easy to use - easier than for instance the EF 85 1.8 USM in most real life circumstances - that is how I remember it, anyway. . .

 

My experiences are:

 

> the 35L has super accurate AF and fast AF with my 5DMkII using centre Point AF (rarely do I use any other AF point)

> the 85L is not so slick as the 35L; but the 85LMkII is much better and is a very nice lens

> the 85/1.8 is not hard to use and represents better value for money (for my uses) and the same, probably quicker AF than the 85L and so I bought the 85/1/.8

> having used the 85LMKII when it was released, (and it is delightful); for me it did not warrant the additional spend as I already owned the 85/1.8.

 

WW

Thanks, William

 

Both my versions of the L lenses were Mark II as I recall. I'll eventually be getting them again.

 

Just did a test with EF 50 1.8 STM and out of 52 images, 7 were not sharp. This time I used a tripod, manually focus the frist two shots and made sure the plane of focus was reasonably perpendicular to the wall.

 

My earlier test of the EF 35 2 IS USM did not go well with about 3 sharp shots out of about 50 photos - also on a tripod, also had checked manually that the camera was perpendicular to the wall. I will retry that one later.

Edited March 26, 2019 by johnv

The EF 50 1.8 STM lens test. First two are manual focus, rest are AF on tripod with about 1 second between shots.

https://photos.app.goo.gl/Pn5o6YSPpSfc7qYv5

 

Unsharp shots are:

5089

5095

5101

5102

5103

5116

5117

I did the same test with the EF 85 1.8 USM and found 1 unsharp photo of 50 using AF.

 

https://photos.app.goo.gl/hzYSt2jKwZKEqQbS6

 

Unsharp shots:

5132

 

I did see some sharpness variation between more shots but nothing too bad.

I did the same test with the EF 100 2.8 IS USM and found 6 unsharp shots out of 51 shots:

 

https://photos.app.goo.gl/fjEdTMDgoADVF4M27

 

Unsharp shots:

5240

5241

5244

5245

5246

5256

 

I did see even more variation between shot sharpness than with the 85mm but I decided only to select the photos that were clearly unsharp.

I did the same test with the 35 2.0 IS USM and found 18 shots out of 56 that were unsharp.

 

https://photos.app.goo.gl/UiPanw8duVPfUBzM8

 

Unsharp shots:

5292

5297

5300

5306

5307

5308

5311

5315

5317

5319

5320

5328

5329

5330

5337

5341

5344

5345

 

Quite a lot more of the shots were not completely sharp at 100% but only reasonably sharp.

BTW, I only used central AF point, Single Shot AF, tripod - so I only checked focus critically at the center point. I think that the 85 1.8 USM works well with this camera, the rest of my tested lenses, not so much, especially not the EF 35 IS USM.

 

I guess I should try the lenses on a different camera body. I do in fact have an old 1D Mark II.

 

I do shallow depth of field portraits from time to time and have noticed some problems with both back and front focusing.

Did the same test above using my EOS 1D Mark II and all photos were sharp with almost zero variation. I realize the camera has a lot fewer megapixels so this is less demanding on critical sharpness. However, many of the unfocused shots on my 1Ds Mark III are out of focus by more than just a little.

 

Zero unsharp photos using the 50mm and 35mm from above :

 

https://photos.app.goo.gl/eUbdbsY9Xv3TsSvj8

 

https://photos.app.goo.gl/ZvJCikcZ8RuMq58o7

 

Is there anything else I can do before taking the camera in for repairs?