Breitling65 wrote:
IanCale wrote:
28-135 and a 50mm, both with the same results nothing sharp
Come on Canon Wake the F' up,



I wouldn't do AI Servo tests on plastic-fantastic first place


why not I have sold stock images through getty with these lenses, in servo, if anybody knows how picky getty is you'll understand why I want my images razor sharp at 100%,

IanCale wrote:
With any product, no matter how good, there will always be a few defective units out there. Nothing to fear...just exchange it for a different one.,

should not be that way, canon is really messing up, I just read in the past 5 mins online, about 8 posts with problems, from , errors', to batteries frying the camera itself, other focus issues, not good as now more are hitting the shelves, and since the major issue with the 1D3, you would think canon would make sure that these things were friggin perfect.


Oh brother!!! cars have issues also why not Canon? let's count all the perfect 7D first instead of listening to a few. Also I saw that post with the battery not the camera's fault, not Canon's fault they do not make the batteries

Please don't shoot the messengers. There apparently are either bad copies out there, possibly more then normal, or there is an inherent problem especially with servo modes, low light and busy backgrounds. When my 40D, 5DmkII and especially mkIIn can shoot more sharper images in servo mode then the 7D then something is wrong.
I am close to contacting Canon and sending mine in for repair.

In good light, single shot or servo modes, my camera is great.

I wonder how many of these peopel have no idea how to set up the camera? My 7D is excellent

Photon wrote:
Sorry yours is bad. Here's someone who seems to have a 7D with excellent AF performance:
http://www.fredmiranda.com/forum/topic/821100


most of them look way oversharpened, look at the artifacts in the blue skies . halos etc..

"Please don't shoot the messengers."

Please don't be ignorant messengers. Until you try to get your camera fixed, don't try to blame your problems on some kind of systemic, product-wide conspiracy. There are thousands of 7Ds in the field, and hundreds of people posting shots every day, the vast majority of them work perfectly fine.

Your anecdotes are not evidence. Some kind of "OMG Canon you're screwing us again!" outrage is just plain ignorant.

tanglefoot47 wrote:
I wonder how many of these peopel have no idea how to set up the camera? My 7D is excellent

so lets have it then, enlighten us, oh mighty one

OK, Ian sent me some picts that illustrate what he is talking about. The red squares indicate where he focussed and the white squares indicate where he thinks optimal focus is. I will post them below with EXIF data and then will add my comments in another post. I did reduce the files in size a bit so that I could use image upload.

My first take is that the files aren't that bad at all. I do agree that the first one could certainly be sharper, but 1/60 for with a 85mm lens on a crop camera is a pretty margin shutter speed and could be causing the problem. In fact, if it were me I would want faster shutter speeds on all the shots--but I do not have the steadiest hands. Also with these narrow apertures the DOF is going to be quite large, so I think finding optimal focus is going to be difficult. IMO, the second and third shots both have the eyes in focus, which is what I would care about. I don't think the camera is performing nearly as bad as Ian suggests--at least that is my take from these samples.

thanks steve, these are 100% crops, coverted from raw, saved as jpgs,and are the much better ones of last bunch I took,

OK, on the second image, the eye is in focus. The third, the eyes are just outside of the focus plane. I think that in those two images, the AF sensor (which is larger than the indicator in the viewfinder shows) appears to have found the contrast boundary of the eyes.

I think that in the first image, the AF sensor again found the contrasty boundary between the shirt and the neck rather than the low-contrast boundary between the teeth and the mouth.

One suggestion - try activating the "high precison" spot AF points and see if the AF doesn't pinpoint the region/area on which you want more readily. While the "normal" mode is best for tracking and general photography, if you're attempting precise focus plane alignment, you will probably do better with the spot AF turned on.

EDIT - actually, when I look a bit more, I see a possible backfocusing trend. Have you tried to microadjust AF on the lens(es) in question?

"Let not shooting the messenger", as I recalled many people already went through with the 1D-MkIII's AF problem soon after Canon released it, and as IanCale mentioned that he "sold stock images through getty", then I wouldn't question his ability with a camera - or in fact, any camera. The problem IanCale just raised only confirm my fear on Canon's QA aspect. Let wait and see how many will report their problem with 7D before cocking the guns.

Sam

Lots of possible explanations for these problems. But, I would say that Ian's comments on micro adjustment seem to indicate that he is either using a faulty process or his camera is faulty. Here is a solid process:

http://www.precisioncaps.com/training_af_micro_adjust.html

I would revisit that first. If you can not get solid lens calibrations following this procedure, your camera needs calibration.

Finally, this comment is not directed at Ian, but just my personal thoughts on the 7D AF. I doubt if 1 out of 10 owners will ever understand what it is doing and how to use it correctly. Everyone has wanted a professional sophisticated AF system ... well the flip side of that is going to be learning how to use it.

Is there a way you can tell us with some certainty what the distance to subject was? I'm wonder if, for the last shot, you were right at MFD or a tad too close? Maybe you said earlier and I missed it, but what lens was used for these?

I also agree with Steve that the first shot, at that aperture and focal distance, should have enough DOF to keep the whole face in focus. But I guess we'd need to know distance to subject first.

And the second one, not a lot of contrast on his nose could have contributed to AF missing?

Can you post some of the "worse" ones as well? That might make it easier to analyze.

Could be any/all those things above, or, maybe you have camera issues, too.

the second and third shots both have the eyes in focus, looked at at 100% to me the eyes are not in focus , I edited everything at 100% and my stock stuff I edit at 200% or better, the images must be flawless, and tac sharp without sharpening, which there is spots that are tac sharp in all 3 of these images, just not at the focus point

When you say focus point ... were you using spot AF? If so, the smaller AF points need something to lock onto and the nose is probably not enough.

There's probably no point in looking closer than 100%, unless you're talking about cloning small imperfections, in which case zoom 500% if you want to. But for overall viewing of an image, closer than 100% can be misleading. Your second pic shows conflicting white boxes, they are not on the same plane, I think that photo might be a bad example of your complaint. You're going to be able to fix this with the micro-adjust feature, don't freak out just yet, the camera is probably fine.

EDIT - actually, when I look a bit more, I see a possible backfocusing trend. Have you tried to microadjust AF on the lens(es) in question?

yes, I went one click at at time from +20 to -20 and it basically just made things worst with both lenses.

Steve Spencer wrote:
My first take is that the files aren't that bad at all. I do agree that the first one could certainly be sharper, but 1/60 for with a 85mm lens on a crop camera is a pretty margin shutter speed and could be causing the problem. In fact, if it were me I would want faster shutter speeds on all the shots--but I do not have the steadiest hands. Also with these narrow apertures the DOF is going to be quite large, so I think finding optimal focus is going to be difficult. IMO, the second and third shots both have the eyes in focus, which is what I would care about. I don't think the camera is performing nearly as bad as Ian suggests--at least that is my take from these samples.

yeah those shutter speeds are WAY too slow for such a high density sensor!

and #2 everything seems backfocused so microfocus might help

IanCale wrote:
tanglefoot47 wrote:
I wonder how many of these peopel have no idea how to set up the camera? My 7D is excellent

so lets have it then, enlighten us, oh mighty one

Still learning the camera but so far it has been better than any camera I have had in the past Oh little one


When I bought my 50D I got the 28-135 in a kit and that lens sucked on the 50D maybe you should try a better lens?