Body and Lens focus problems -how common?

[jeremy_tressler]

Hello,

 

I have just started working with digital photography this summer.

 

Discovered that my camera and all but one lens has focus problems. All of my equipment is new, not grey-marlet stuff purchased within

the last four months. Do many people end up sending their equipment back to the manufacturer for calibration of focus issues?

 

1) Performed the excellent focus test at www.focustestchart.com and discovered that my camera (40D) front focuses a little to a lot

consistently on all lenses and is worse when using the outer focus points. From what I've read, this means (in part) that the focus point

targets in the viewfinder aren't properly aligned with the actual focus point sensors. Apparently Cannon will fix this problem if I send it in.

I have already made a call and they didn't seem surprised.

 

How common is this problem and does Cannon usually do a good job with this calibration? I'm under warranty.

 

2) Next I wanted to isolate potential lens issues so I used the LCD at 10x magnification to focus manually on a bundle of horse-hair

using the same magnification through all three lenses (so horse-hair appears the same size in all lenses). Here are the results from

worst to best:

 

In other words, the best focus achievable by manually focusing using LCD at 10X:

 

Sigma 24-70 f2.8EX -EXTREMELY HORRIBLY OUT OF FOCUS

Canon 50mm 1.4USM -HORRIBLY OUT OF FOCUS

Canon 24mm 1.4L -OUT OF FOCUS

Tokina 100mm macro f2.8 atx pro -Perfect or nearly perfect

 

I don't get it, this is all professional level equipment. Is there no such thing as quality control? I have read somewhere that if a lens

doesn't focus properly it is because the lens itself is out of alignment and that can be adjusted and even calibrated to a particular

camera body. Again, I will end up returning all of this to the manufacturer for 'repair' and I sure hope they don't do a crap job and make

me return again further wasting my time..

 

How common is my experience? Any thoughts? I am extremely dismayed..

 

Thank you, jeremy

[j_smith6]

Not properly calibrated body/lens happens. I don't know how common it is but I never had a single problem with any

lens. All are focusing 100% correctly with digital and film cameras.

 

I think your lenses are fine - it's hard to tell without sample pictures but it's normal that zoom lenses are "soft"

especially wide open and both 50/1.4 with 24/1.4 are certainly going to be soft wide open - it's just f/1.4. Do you ever

use f/1.4 in real life photography? If pictures look fine after the body is calibrated you have nothing to worry about.

Macro lenses usually are sharper and that's why you got better results. I think you exaggerate quite a lot, but again -

you may be right, but it's impossible to tell without sample pictures. Just don't expect perfect sharpness when

viewing your images at 100%, especially without sharpening.

 

Did you use tripod, mirror lock up and remote switch?

[sattler123]

The real problem is that with digital photography, people tend to over analyze and pixel peep their pictures. I have had no issues with any of my camera bodies or lenses and their focusing accuracy.

 

Did you ever perform these "focus tests" on a film camera body? And if you did, did you blow up the neg at 100% magnification? I am sure 99.9% of people did NOT. So why are we doing it with our digital bodies? If there is indeed an obvious focus shift, then by all means, send it in and have it repaired. It just seems that many folks do more focus tests than take real photographs because they are so pre-occupied wih having a faulty camera/lens combo.

[henry_s3]

I had continual problems with my 20D and Sigma 18-200 f3.5-6.3 lens. I sent in the lens for it to be recalibrated. It took them forever and a day and then they replied that the lens was fine (which it wasn't - nearly all my prints were telling me so!) and didn't need recalibrating. Two local photo dealers examined the lens on the body and also clearly stated that the lens was not in order. Sigma was not interested. I was later given the tip by the dealers that one should send in both the lens and the body so that the lens is set up for the body and they were surprised Sigma never told me this. I bought this lens because I had been very pleased with the Sigma 28-200 analog lens. However, I got so fed up with fuzzy pictures on the 20D and Sigma's lack of interest, that I bought myself a Canon lens. Now no problems! O.K. it's considerably heavier but I don't need to go to the gym any more!

[ronald_moravec1]

You will have much less problems with manufactures lenses. How do you think they achieve cost savings to sell cheaper than name brand. You always run into the issue the lens guy blames the camera guy and the reverse. Who do you think looses?

 

Don`t check the 24 mm or 50mm 1.4 at a target two feet away. Check it where you use it. It is not made for close up work. Secondly there will be a focus shift as you stop down specially at close distance. A lens can`t be calibrated for different distances. If you make it right for 1.4 at two feet, it will be wrong at smaller stops and/or 10 feet.

 

All close work should be manual focus.

 

I would also look into if the pro level bodies offer berrer AF. The Nikons seem to me to be that way.

 

In short, you get what you pay for and do not attempt to make a piece of equipment work perfectly outside its design perameters.

[sarah_fox]

Misc comments:

 

1. There's a lot of mystery and handwaving that goes with focusing issues. The lens is often blamed, but I can't personally see how a lens could "misfocus," beyond errors that might be attributed to its design compromises (e.g. spherical aberration -- see below). It would seem far more likely that the photographer or the camera body misfocuses it (depending on whether M or AF is used). This could be due to misalignment of the AF points and/or focusing screen (again, AF or M). I can see how a very loose focusing mechanism in the lens could result in "overshoot," whereby the inertia of the mechanism takes the focus a few steps further than the camera is trying to achieve, but this doesn't seem likely with a USM/HSM mechanism (which is a bit like a piezo stepper motor). The bottom line is that the camera body focuses the lens until it is satisfied with the focus. The lens has no idea what it is "looking" at, so all it can do is to "follow instructions." It is the camera body that "sees" the image, makes focusing decisions, and issues those commands. Furthermore, the body has no idea how far away an object is, so it doesn't tell the lens to focus to 5.637 m, for instance. Rather, it tells the lens to focus closer, closer, closer, nope a bit farther, yeah, right there. If there's any distance information passed along anywhere, it's the lens responding, "OK, according to my encoder, that would be about 5.637 m." With all this in mind, it's the camera body that needs adjusting -- to align the light path distance to the AF sensors with the light path distance to the imaging sensor.

 

2. Only some of your equipment is "professional" equipment. The 40D is prosumer. The Sigma EX and Tokina ATX Pro lenses, while in those manufacturers' "professional" lineups, are rarely up to Canon L standards. I'm not criticizing this equipment. It's very good stuff. However, when you're communicating with the manufacturers about it (particularly Canon), you need to know how to approach them. If you were talking about a 1Ds Mark III, THEN you could do a bit more fist pounding and speak with more righteous indignation.

 

3. People on PN will pound you over the head about the evils of pixel peeping. To me, it's no different from looking at a negative with a loupe. Some people want to examine fine details of focus, grain, or whatever. Perhaps the anti-pixel-peeping zealots think pixel peepers expect infinite resolution, just like in Hollywood movies? Perhaps they think pixel peepers have an unrealistic expectation that there should be no little squares (pixels)? I honestly don't know what their problem is, but if they think critical focus can ALWAYS be nailed by AF, by MF with focus confirm, or even with a split image finder, and especially if they think the DoF preview can give them any reasonable notion of critical focus throughout a range of focus, then they are as foolish as *they* think *I* am. So here's the deal: It's OK to peep. It's even good. What you are looking for is critical focus, and for most upper-end lenses, that's achieved when lines and edges are blurred over about a pixel width -- hard to describe/quantify, but you'll see what it looks like. This varies by lens and body, and it varies by center/edge position in the frame. However, you'll quickly learn what your lenses can and cannot do, and with that in mind, peeping is an excellent way to confirm whether you nailed the focus. Your question, when peeping, should always be, "How does the sharpness of this image compare to the best I know my lens can produce at this aperture and focal length?" Furthermore, "Did I successfully achieve adequate focus throughout my entire object field?"

 

5. It seems to me that pronounced spherical aberration in a lens (which is a design issue) could produce focusing errors with phase-detection AF (such as in the 40D). This would happen because the light path convergence through the outer margins of the rear element of the lens would be different from the convergence of light passing through most other areas of the rear element (since it is the information from the outer margins that is used for focusing). This would most likely be a problem with a lens that is "soft" when wide open. So although faster lenses are AF-friendly by providing more light to focus, they are potentially also AF-hostile because of systematic focusing errors introduced by spherical aberration, at least in faster lenses of lesser quality. But because AF is somewhat hit-and-miss anyway, even with the best of lenses and camera bodies, pixel peeping is the best method of confirming that critical focus was achieved.

 

6. Much of this is speculation on my part. I've not found any cogent discussion of this topic anywhere.

[jeremy_tressler]

Hi All,

 

Thanks for your replies thus far. Last night I posted a second comment that seems to be missing.. that is, for my

second test using the LCD at 10x magnification to manually focus through each of my four lenses (comparing sharpness

of focus), I compared all lenses at f8.0 and I focused no closer than three feet away. I believe this should be a pretty

sweet spot for all lenses, correct me if I'm wrong, like I said, I'm a newbie and have no significant ego anyway.

 

I will post images later today, however 10x magnification in a 2-3" LCD can hardly be called pixel peeping, particularly if I

want my 10mp image to be looking sharp at 8 x10 or 13 x 19 print and I don't see how it could not be a good idea to

check ones equipment in this way to see whether the stuff works properly. What do we do, always take the authorities

word for it like with politicians? I got burned with faulty equipment in the audio industry so often I never take anyones

word for the stuff always working to spec. I regret not doing this sooner as I had already done many photo shoots of

dance photography and only started probing further after noticing that my images seemed to be very often soft and not

always just when the lens was wide open.

 

BTW, some of you respond in such a way that it seems like you don't read everything from my original post. For

example, two of my lenses are 'original manufacturer' (Canon), and one of them displays horribly out-of-focus, so no, this

does not make me feel that always buying the same brand lens as body is necessarily the way to go, in fact, the only

lens that displays perfectly in focus (and no I'm not exaggerating as you'll see when I post images in a couple of hours is

the third party lens, Tokina, of course maybe this is only because its a macro and the others not? Still I don't see how

the lenses could look so different at f8.0 from no closer than 3' away, all with the same object in focus to the same size

in each lens. Posting pics soon.

 

Also, I'd like to point out that I spoke with the technical support at Canon and they said that focus problems could be

from body, lens, or both and the calibration can be done to each and all in order to achieve desired results so that tells

me that yes, a lens CAN be not focusing to spec. Certainly if you go to photozone.de and read the excellent lens

reviews, you'll see often a mention about 'bad copies' of lenses etc.. Yes, good photos is about being freely creative and

not hung-up on ones gear, however, to me, the best way to be free of gear is to know that ones equipment functions

properly then one can forget about it and take the pictures. Just like driving, you want to know that your tires are inflated

properly, oil in engine etc, then you can forget about the car and drive knowing it will reliably do its job while you do your

job. Machines are dumb and so I think its imperative that we check up on our equipment once in a while to know that it

works as spec. It won't say when its sick.. I suspect there are a lot more quality control issues with mis-calibrated

equipment than people suspect..

 

Cheers All, jeremy

 

 

Cheers, jeremy

[tomwatt]

I want to comment on Sarah's comment about pixel peeping, but it does relate to the thread as a whole: pixel peeping for me is the same as pulling out the grain magnifier and working focus issues on the enlarger. I can honestly say I have better results with the digital and autofocus than I ever got with film and manual (I have some vision problems already, and age is amplifying it - I can no longer even use a grain magnifier). The whole process of image-making is always a compromise between equipment and ideals. Doing it well requires a nice balance.

[phule]

[[if I want my 10mp image to be looking sharp at 8 x10 or 13 x 19 print and I don't see how it could not be a good idea to check ones equipment in this way to see whether the stuff works properly.]]

 

Then you should test by making 8x10 or 13x19 prints.

[seismiccwave]

You have a few variables to isolate and iron out. From what I have read I noticed a lot more complaint about focusing

issues with the Canon camera than other brands. From my personal experience I do have focusing problem that

prompted me to send in my 5D for adjustment. The service was extremely fast and the work very well done. Apparently

there are a lot of tolerance in the manufacturing process.

 

I don't expect a 40D to have the same focus tolerance as the 1DsMk3. Simply because there are a lot more 40D

manufactured.

 

So to make you feel better, send in the 40D and see what happens. You are only out of some postage.

 

Now let's get to some focusing issue. When you compare auto focus and manual focus you are not comparing the lens

focusing ability. You are comparing the focus distance between the lens flange to the sensor and the lens flange to the

focus screen on your DSLR and/or to the live view sensor. From my understanding the image takes two paths to get

there. The image goes from the lens directly to the sensor but reflect upward via the prism to the live view sensor. I

could be barking up the tree and got this totally mix up here. If I do my apologies and simply ignore me. However I just

want to point out that you need to go back to basics and see what may be the reason you are having the difficult in

focusing.

 

You didn't post any of your test charts and detailed your testing procedure. It will be very helpful if you do.

 

You are not asking for a one answer question so let's start a dialogue.

[john_alexander]

I have had a range of Canon D-SLR's, starting with the D30, on through the D60, 10D, 20D and currently the 30D. I never felt like I

had any focusing issues with the D30, D60 and 10D, but the 20D and 30D both have had issues. There is no doubt in my mind

that it's the camera BODY at fault. The 20D focused TOTALLY differently from the 30D. I have ONLY Canon "L" lenses, and have

found that with some, the 20D either front focused or back focused, and with the 30D body, the lenses focus very differently.

 

This is not a "subjective" idea of "focus", but a REAL issue which is 100% noticeable in even a 4x6 print!

 

As I said earlier, I had no issues with the D30, D60 and 10D, so I'm thinking that the Canon manufacturing tolerances have fallen

off over the years.

 

Rather than send the 30D body and all my lenses in for service, which would mean the corrections would only apply to THAT 30D

body/lens combo, I've decided to get the 50D, which has the ability to "fine tune" any focusing issues and store the tweaks for up

to 20 individual lenses.

[bill_tuthill]

Jeremy, where did you post samples?

I want to see the EXIF before commenting.