Discussion in 'Canon EOS' started by adam_k_krause, Oct 29, 2004.
Just curious on how I'd go about testing if my camera has front or
back focus problems?
Are you unhappy with the pictures you are getting?
Don't get sucked in, very few people have a real problem and you would
notice if you did.
Just by trying to test for this in a non laboratory setting can convince you that
you have a problem where none really is and that will spoil your enjoyment of
using what is probably a very fine camera.
Be a photographer not an anorack.
Sure, just sit around and wait till your camera is out of warranty before realizing the film plane isn't in line with the plane of focus in the viewfinder. My 10D has a minor problem with it, I've used a friends 10D with a BIG focus problem, and it's totally f---ing annoying for those of us shooting with good glass and anything less than F8.
It's easy to test for. Lay a ruler on a table pointing away from the camera, and manually focus at a particular mark on the ruler and shoot wide open. If your 20D isn't a 'problem child', you'll get that same mark in focus in the capture image. If you have the common focus problem with many 10D/20Ds, you'll get a different point in the ruler in focus than what you saw through the viewfinder.
"Sure, just sit around and wait till your camera is out of warranty before
realizing the film plane isn't in line with the plane of focus in the viewfinder."
And if you haven't noticed for a years worth of images you probably don't have
As far as I understand it (and canon indicate in their latest litterature) the
problem can be camera or lens specific so if you will need to check every lens
you currently own or buy in future (even if the body waranty has expired).
Basically I stand by what I said earlier if you can't se any problem then don't
worry about it.
Test with multiple lenses at multiple zoom positions. Focus depends on BOTH the lens and the body, and the focal length setting of zooms.
Focus problems are not unique to Canon and not unique to any particular body or lens. The more paranoid you are and the closer you look, the more problems you will find. Nothing is perfect.
If you see no problems in actual use of the camera, that's a pretty good test. If you perform accurate "scientific" focus tests and you look hard enough with enough lenses and settings, I'm sure you'll eventually find a problem.
Reminder to all: The ONLY stupid questions are the ones you don't ask...
>>The ONLY stupid questions are the ones you don't ask...<<
will the SUN come up tomorrow? The only stupid answers are the ones you don't give
It's just that the focus "issue" has been beaten to death, buried, excavated, sent to the morgue for further exams, buried again...and again...and again...and again.
It's understandable that people get impatient with those questions, especially since the archives must have 100s of posts with the same question and therefore, the same answers.
A focus test doesn't change with the camera model, make, year of production.
I don't think we should look for problems with each piece of equipment we purchase. Use them and *if* you see a REAL problem during the course of normal use then you will have to identify it and test it before sending it in for repairs.
To each his own, I guess. For me, I prefer using than testing...
HA! The SUN indeed - this is cloudy rainy England - even if it does come up (not a foregone conclusion based on this morning I shall certainly not be there to watch it! Have a good weekend...
Focusing on a ruler or other oblique object is not a good test. In order for this to work the focusing sensor and the focusing mark on the screen must be 100% aligned - if not you are aiming the sensor at the wrong place. How often when shooting in real life to you shoot oblique objects ?
That's why you use MF for this test.
Oh for God's sake - you need to have the ruler at infinity anyway. Otherwise you're into some very complex trig regarding WHERE the focus point is. You paid a fortune for the camera, and a King's ransom for the lens. If it isn't good enough then take it bck to Canon. With the photos of course...
Whatever you do, DO NOT take a full frame image of a uniformly illuminated flat area (like a grey card), then stretch the histogram to full scale to see if the sensor has uniform response / the lens has uniform illumination.
Be warned. Don't do this or you'll never get a good night's sleep again.
Focus accuracy is a very valid issue. Some cameras with some lenses have indeed revealed a problem. It's worth checking. Just don't get too hung up on it if things look good. If they don't look good, then that's a different matter.
Once upon a time, people would use film for their photography, and used to be happy. Then came the digital era, and all of a sudden, images are blown up to 200, 300, and 400% in non-calibrated monitors. Oh, my, my lens is not that sharp afterall, I will never be able to sleep restfully again. It must be defective. Let me go on-line and see if some body else has this problem. Will this affect my 10x15cm prints?
And on, and on, and on...
Adam, just use the damn thing, and if you do see a problem under your typical shooting conditions, send it back for repair.
I'm not sure why so many of you have bothered to post "don't worry be happy" etc. It's a simple question -- how do I test? That deserves a good answer (which was given in the link). Moreover, this test will potentially avoid questions of this sort -- "my images are soft -- do I need a new model?" If one quarter of the people posting such questions did a test instead, and learned more about focusing etc. then we'd be ahead.
It's a simple test, which will cost you a few hours at most. I recommend reading the entire thread previously mentioned, and also using the Canon DPP / Remote Capture in tethered mode for quick high resolution review of images. It's your hobby/craft/etc., go wild, get it over with, learn a few things in the process, and then do what you want.
And if Canon's quality improves with consumers holding them to higher standards when they have verified defects and poor performance, so much the better from my perspective.
I'm not sure why so many of you have bothered to post "don't worry be happy" etc. It's a simple question -- how do I test?
First, Ignore the inbred, time for tea, "Oh my god, my mother was actually scotish" trolls here saying crap like Oh for God's sake - you need to have the ruler at infinity anyway.
Lay a frikken ruler on a table, manually focus on a specific point in the middle of the ruler, and take a picture at a wide open F-stop. Or, use the rim of a glass and shoot the edge of it as another good subject - I don't care. Your captured image should display the exact point you focused on, and many 10/20Ds won't because they have misaligned mirrors/prisms. At 3-feet away my own 10D is a little over an inch off, which may not seem like much to somebody who always shoots at F8 and in AF mode - like Claire and a few others here. Seriously, they need to go to a point -n- shoot forum with other in the equivelant skill category because they obviously lack the abiltiy to take a picture of something in manual focus at a wide aperture.
I can sit at my desk with my 10D, take a picture of my office phone 3feet away at F2.8, and the button on the phone I'm manually focusing on *IS NOT* the one that is in focus in the final image. I've used some 10D's where the problem is *worse*.
I don't think we should look for problems with each piece of equipment we purchase.
I paid $1500 for my 10D, jerk, and I expect it to perform as well as my RB and FE-2 in terms of basic operation - obviously you don't. This includes things like metering and focusing. If you find these concepts 'rocket science' and should be not be tested with a new camera, you are seriously naive.
Oh, my, my lens is not that sharp afterall, I will never be able to sleep restfully again. It must be defective. Let me go on-line and see if some body else has this problem. Will this affect my 10x15cm prints
Another troll that I hope finds a life and goes elsewhere because he doesn't want us to hold the same standards with our digital gear as film. I shoot and sell lots of macro work Mr. Troll, and I prefer it to be as sharp if not sharper than my MF macro work, and focus on what it's in focus in the viewfinder, got it? My FE-2, F3 and RB do not have this problem, and 10 years ago if a certain brand of film camera were to exhibit the same problem, it would be all over the bulletin boards, in photo magazines, and in every camera store in the Westrern Hemisphere that it had a problem and to avoid the brand.Since these is a digital issue though, the hypocrites/trolls are screaming their bloody holes out though (as evident here) since you're not allowed to hold the same standards with digital even though you paid more for the camera.
Canon and Nikon are far more profit concious than 10 years ago, and build these machines in the cheapest factory they can at the lowest labor rates as possible. This is not 1985 when a $600 film SLR was unlikely to have such a basic problem as film/sensor plane misalignment. Just wait till production moves to China and then Cambodia.
If Bob can, I wish there was a way to lock this thread because enough has been said and he knows I'm right.
Scroll down to the bottom of the page, and instead of focusing on an inclined piece of paper, use the domino test on a tri-pol to ensure that your subject /w contrast are parallel with the lens' focal plane. Between shots, manually reset or set your lens focus to macro or infinity. Use consistent lighting, and manual everything except focus. Use a tri-pod. Try to minimize all variables to isolate the camera's autofocus circuit.
Well I just got my 20D last week, took it to school to shoot the Halloween
parade, and most of my pictures were not sharp! The focus accuracy of my
EOS-1v and EOS-3 were deadly sharp, so I decided to do the inclined paper
When I manually focus, I find that I do get a sharp picture--which tells me that
my prism/film plane are in good order. When I autofocus, I find that it depends
on the lens. I tested 15 f/2.8 fisheye, ranges of 17-35 f/2.8, ranges of 17-85
EF-S, 35 f/1.4, 35 f/2, 50 f/1.4, 100 f/2.8 macro. The 100 f/2.8 macro was
dead-on. The 35 f/1.4 and 35 f/2 were very close. The 50 f/1.4 was a bit off,
the zooms were more off, and the fisheye was way off. I used only the center
focus spot, and did the tests at f/2 or wide open (if the lens did not do f/2).
Ten years ago I switched from manual focus to autofocus because the EOS
620 could focus better than my eyes could. My eyes have gotten worse, and
somehow the 20D has gotten much worse.
I did try the domino test, but did not find it easy to set them up in a way that
gave me very testable results.
Should I expect better results from the 20D? Should I expect results as good
as my EOS-1v from a 1D MkII?
Ask Scott - if he's back from his anger management class!
Wow. Although this was a, how do I put this.... emotionaly charged discussion, I decided to go and test my 300D since I manualy focus. Thankfuly, the mirror is dead on.
I must agree with Scott here. If you pay for something, it better be built with damn good tolerances. I don't want my camera to be built with the production tolerances of a Ford. Scott is right, what good is a good lens if it cannot be focused correctly?
If you find that you have good focus with one lens and poor focus with another, it's probably a problem with the lens. Contact Canon and explain the situation. They can (and will) make adjustments to lenses to correct the problem.
I upgraded from the DRebel, which I had for a year, to the 20D. I am not so happy with the results. Looking back over my thousands of photos with the DRebel, I got better results. There is something just not quite right here. I can get better performance using manual focus, something I never needed to do on the DRebel. Doesn't matter if I use the fabulous 50mm/1.8 or a Sigma zoom lens, I still am not getting the focus nailed via the autofocus.
I am going to see if Canon service can help here, it is just getting too frustrating.
I just got through reviewing the 20D for photo.net. Mine nailed focus dead on every time. I was mainly testing with the 500/4.5L, 300/4L and 50/1.8. I was testing on resolution targets using both autofocus and manual focus (with eyepiece magnifier), so I'm pretty certain I'd have detected problems if they were there.
I have been using a 50mm/1.8, 28mm/2.8, and the new Sigma 18-50/2.8 lens.
I'll post some before/after photos when I get the 20D back from Canon service.
The problem was pretty consistent over a large number of photos, most obvious in portraits, where I am accustomed to focusing on the eye, and all the images come out with stuff forward of the eye being in sharp focus (hair, fabric on collar or sleeves) , and the eye being just out of the sharp area. My photos from the 300D were much more consistently in focus in the same situation.
"Another troll that I hope finds a life and goes elsewhere because he doesn't want us to hold the same standards with our digital gear as film. I shoot and sell lots of macro work Mr. Troll, and I prefer it to be as sharp if not sharper than my MF macro work, and focus on what it's in focus in the viewfinder, got it? My FE-2, F3 and RB do not have this problem, and 10 years ago if a certain brand of film camera were to exhibit the same problem, it would be all over the bulletin boards, in photo magazines, and in every camera store in the Westrern Hemisphere that it had a problem and to avoid the brand.Since these is a digital issue though, the hypocrites/trolls are screaming their bloody holes out though (as evident here) since you're not allowed to hold the same standards with digital even though you paid more for the camera."
Scott, YOU should have gone elsewhere a long time ago...
The 20D came back from service and works fine with all Canon lenses, but the Sigma 18-50/2.8 still seemed to front-focus, so I sent it in to Sigma and when it came back, it was acting much better. I still have some problems with it when used in AF and the 420EX is doing it's focus assist, but otherwise it seems to be happy now.
Hmm...seems like there's a new business opportunity for the corner photo store: to hand-test and submit for service all the kit they sell that didn't get done right the first time. Humpf!
I had the same problem with a new 20d kit. 18-55 lens and it would not focus correctly. Any movement at all even at fast aper.'s and it would focus on something close by. Even using the middle a/f selector I was having problems. I sent the kit back for another. If it continues I would recommend the same. Regardless of what the Canon knights of the Templar say, there is a real issue. chris
This is the first time I've tried to find out if there are others having 20D focussing problems. I'm a little surprised at some of the attitudes out here. It seems obvious to me that there is a focussing flaw with the Canon digital cameras and I don't understand why we all just don't band together and demand that Canon address this, fix all our cameras at no cost, or replace them. They are pushing new models out as quickly as possible, and it's obvious that the quality control is poor - if not the engineering design. This is simply unacceptable in digital equipment costing thousands of dollars. I'm not going to explain my particular problem because I'm not in the mood to get yelled at, but I've learned enough from reading this page that I know I am going to address my problem with Canon and demand results. Thank you.
Separate names with a comma.