Nothing in focus - Canon 6D or Canon 24-70 f4L - which is to blame?

I recently bought a 6D and a 24-70 f4L both used. 6d has 7k actuations and the lens is 4 months old and hardly used (looks new).

I'm trying to find out if it's the lens or the body to blame as I can return the lens in the next few days but don't want to send back a good lens if the body is faulty!

I've taken about 140 test shots with different combinations of micro adjustments on the body to compensate for possible front or back focus.

I used a printed focus chart with the camera on a tripod and various outdoor shots which were handheld, fast shutter speeds and mostly F8
Nothing has worked.

I also tried my 70-200 F4L which was better but still not as sharp as on my 40D. I haven't micro adjusted this lens yet.
Any ideas as to which is more likely to be the culprit would be appreciated.


See 2 shots, one close one distant.

1 - 1/500sec, f8, iso 400, 70mm

https://photos.google.​com …OOiNWe15qAnaqgk​33uFHxYxX3


2 - 1/1000sec, f8, iso 100, 70mm

https://photos.google.​com …C__xvkzTIa5x5QR​trZnyaXvSe

photos aren't showing up. maybe you can try uploading them elsewhere?

how does the 24-70mm f/4L hold up when you focus manually?

also, do you have any other lens that works fine? like a 50/1.8 or something?

steve18 wrote in post #19044985
I've taken about 140 test shots with different combinations of micro adjustments on the body to compensate for possible front or back focus.

That's not the way to do it.
Take a photo. Hopefully somewhere in the photo is sharp. If not, that's a problem.
Assuming somewhere in the photo is sharp, is it behind or in front of the point you focused on? Once you determine that, you can use micro focus adjustment. (if results are consistent. If some are front focused, some are back focused, no point using MFA)

To just test the lens, focus using live view. It is rare for live view focus not to be accurate.

Finally, if you are viewing results at 1:1 (or actual size), 6D results will not look as sharp as 40D.
This is normal as the 6D is 20MP vs 40D 10MP. Viewing actual size means you are magnifying the 6D image much more than the 40D image.

Step 1, before you do anything else. Set up the camera and lens on a tri-pod with good light and a good target. Using live view, zoom in 10x to your focus point and manually focus. This will establish your baseline, the best the camera and lens are capable of, something to compare to. If the image is still out of focus and not you your satisfaction, you have bigger issues than a front or back focus problem and I would start looking at the lens and camera themselves. A manually focused image using live view zoomed in 10x will be the best you can get.

If that image is good and meets your expectations, which I think it should as the 6D and 24-70 combo is a really nice combination, then you start with other tests, like using the live view auto focus and make sure that works. If so, then move on to the phase detect focusing system.

gjl711 wrote in post #19045010
Step 1, before you do anything else.

Whenever you buy a used body, I’d like to offer that step 1 should be setting the camera to factory defaults and clearing all custom functions. Then retest.

Maybe this.
Is Focus removed from shutter. Press the shutter half way. If a Back Button Auto Focus user owned the camera, they may have not set it back to Shutter.
My cameras have been set to Back Button AF for ten years now. When some else tries to use the camera, I have to show them how to focus.

John from PA wrote in post #19045012
Whenever you buy a used body, I’d like to offer that step 1 should be setting the camera to factory defaults and clearing all custom functions. Then retest.

Good point, I just assume that anytime you get a used camera, you do this, sort of step zero. I was just addressing that when you start testing for focus, before anything else, make sure that the equipment is capable in the first place an no better way to do than than to do it manually removing all automated functions. But yes, make sure everything reset back to factory defaults for sure.

deanedward - Manual focus better but still not sharp. Only other lens I have that fits the 6D is the 70-200 F4L which is better than the 24-70 but still not spot on. I'll try to upload pics somewhere else, don't know why they're not working.

Choderboy - I did 3 tests on a tripod with a focus chart, slightly different angles to accentuate the blur. AF, no IS, F4.
Results were:
1. 24mm +5, 50mm ok, 70mm -1
2. 24mm ok, 50mm -1, 70mm -1
3. 24mm -1, 50mm -1, 70mm -1

Took photos with all the above settings and every shot OOF.
Haven't tried live view, will do that next. Coming from crop to full frame I have a lot to learn!

gjl711 - Will try live view as you suggested.

John from PA - I didn't do this but did go through every setting before taking shots.

Nick5 - It's on back button focus as it was on my 40D for many years, AF-ON, single shot.

Thanks for the replies, appreciate it.


1 - 1/500sec, f8, iso 400, 70mm

Just curious, does the lens have a filter on it? Usually filters affect long lenses more then short, but that picture looks all around blurry. Nothing looks sharp.

You can also use the yardstick test to diagnose what the problem is after resetting all camera functions:

https://www.picturecor​rect.com …brate-your-cameras-focus/

Have you put the camera in Live View mode, then used the on-sensor focus to achieve best focus as seen on the LCD, and assessed that photo on a PC? It SHOULD give you perfect results even when microfocus is AFU (all fouled up). Assuming it does, then you can monkey with MFA on the camera knowing that it is not the lens!

steve18 wrote in post #19045071
... Manual focus better but still not sharp. ... 70-200 F4L which is better than the 24-70 but still not spot on

This statement really concerns me. When using live view for focusing, it's the very same sensor for focus and capture. If the camera is stable (tripod) and you're focusing accurately in live view, the images should be spot on.

Regarding AF, I have a 6D and the AF system is what I like the least on that camera (compared to my 5DIII). The (only!) 11 AF points and only the center point being cross-type, is the main crippling factor on that camera (IMO). But, when using the center point, I've always had great results with the 6.

Still, I can't let go of that bad feeling about the manual focus test. The picture above with the brick wall and (lots!) of foliage makes me wonder if that was manually focused or using AF. With all of the surrounding branches and many of them closer than the wall, did the camera simply mis-focus. If it was manually focused, then what were you focusing on? You've got that large red box, but it encompasses a lot at different distances. The second picture might be the concrete patch below, but it's not the same picture magnified (different EXIF data).

According to DXOMark, the sharpest settings for the camera/lens are with the lens set at 24mm and f/4. How about trying that and a simpler subject? That lens doesn't rate the highest, but still, the image that you posted looks way, way worse.

The other question I'm wondering about is what kind of post-processing has been applied? Are we looking at a JPG out-of-camera or did you shoot in CR2 and then how did you convert it for posting here?

steve18 wrote in post #19045071
Coming from crop to full frame I have a lot to learn!

Framing differences aside, and even with different MP densities, you should expect reasonably sharp images out of either sensor size.

1. If there was a problem with the reflex mirror being out of alignment, or there is a problem with the focus screen not resting properly in its holder,
   in either case the user would not be able to see a sharply focused image in the viewfinder ever.
   
2. Since the supplemental focus mirror is afixed to the main reflex mirror, its displacement would cause the wrong assessment of focus by the AF sensor in the base of the camera, too. And since neither the eye or the AF sensor is seeing things properly because of a main reflex mirror out of position, all taken images would be out of focus.
   
3. Live View focus bypasses focus error caused by the reflex mirror or by the focus screen, and should be sharp and in focus in spite of either being out of whack.
   

gjl711 - No filter on any shots.


dangermoney - Next thing to do is to reset everything and if it doesn't work it's all in the bin!

Wilt - Just did 2 shots with LV, manual focus and auto, see below.
Image looks sharp in the viewfinder so I guess it's not a focus screen problem.


dasmith232 - LV pics are better than non-LV auto but the 24-70 F4L is much lower quality than the 70-200 F4L (I will try and upload 2 more pics in my next reply as POTN only allows 2 pics per post).

The 2 pics you've seen are from different places. The house in the trees I focussed on the chimney pot, AF. I took 3 or 4 shots and refocused each one.

No post on any shots except auto contrast, all highest quality jpg.

Appreciate the DXOMark info, wouldn't have bought it if I'd seen that!


AF Liveview / tripod

Forgot to say all pics are 100% crops.

1st pic 70-200 F4L
2nd pic 24-70 F4L

Both autofocus.