fredmiranda.com

Moderated by: Fred Miranda

New fredmiranda.com Mobile Site

New Feature: SMS Notification alert

New Feature: Buy & Sell Watchlist

FM Forums | Canon Forum | Join Upload & Sell

3 end

Archive 2003 · Best 10D Bad AF Example Ever

DavidP
Offline
• • • • •
Upload & Sell: Off

p.2 #1 · Best 10D Bad AF Example Ever

fred_jb wrote:
A lens system focuses in one plane, and one plane only (ignoring distortion). We just need some way of defining how close AF gets to the intended plane. I'm not sure that relating focus accuracy to so-called depth of field as currently defined is a very satisfactory method - but that is a different discussion.

Yes, only one plane of focus.

However, expressing accuracy as a function of wide-open DOF is the ONLY thing that makes sense.

If you use a lens with a max aperture of f/1.4 vs. a max aperture of f/5.6, there's no way the AF can be expected to be more accurate in absolute terms with the f/5.6 lens. And that's whether you AF or MF.

Your ability (and the ability of the AF system) to discern where the actual plane of focus is depends ENTIRELY on being able to "see" where things are in focus and where they're out of focus. And THAT depends on wide-open DOF.

Jul 15, 2003 at 06:49 AM

Canonizer
Offline
• •
Upload & Sell: Off

p.2 #2 · Best 10D Bad AF Example Ever

christo� wrote:
If it makes you feel any better, you're not the only one getting QC defects, that's for sure. I'm lucky with the 10D, but the little battery charger that came with it went POOF in 2 months flat

Ironically, I had initially looked at the little battery charger and thought, man, that thing looks cheap. Wonder how long that will last? Did you also read in another post that I bought a new 1.4x TC II, and it arrived sounding like a baby rattler? This and the bum 10D within 2 months of Canon acquisitions. I'm concerned -- I don't have time for this hit-and-miss QC approach with Canon stuff. I don't suppose many of us do.

Anybody telling you it's all in your mind is crazy, but you have to admit that a lot of people post one or a few shots with dubious subjects for AF test, and for some reason, people seem to use long lenses with large aperture for the testing, and don't give much background info. People are posting with a lot of different experience levels, and clearly many of them upgraded to L class large aperture lenses for the first time along with the 10D. I don't care what AF system you have, when you're shooting with a 1 inch DOF, getting...Show more →

Some truth to this. Still, all my other Canon bodies do a better AF job. BTW, my EOS 3 was $995

Outside my having a bum 10D, I have been shooting the 3 almost since it came out. Canon really broke ground with this camera -- 45 AF sensors! All sorts of new circuitry, eye controlled focus that even works with my bifocals. A piece of art to hold and behold. And the thing worked beautifully from the first roll I put through it. It's never been in the shop. It is one of those cameras that, from the first shoot, you fall in love with. There is something, I think, to your comments about moving "down" from that to a 10D. I get a kick out of people commenting on the great 10D build, for example. Better than a Rebel, no doubt, but the 10D does not have a great build. It's good that there's no body flex, but compared to an EOS 3 ... well there is no comparing.

The 10D is a camera, OTOH, that from the first shoot you think "I paid $1500 for this?" I really struggle with the soft images. Yes, I know all about sharpening, but even that doesn't always get it to where I am accustomed in the film world.

I do hope you have some luck with returning your 10D. Some people haven't been too happy with the results -- hopefully the technicians have had time to learn more about how to re-calibrate the things right. Just don't expect it to come back working like the AF on an EOS 3! It would drive me nuts, but the full time MF ring on the L lenses, the fast AF lock / re-lock cycle time, and the low price on muffed shots all do a lot to make up for the lesser AF system to me. I also really like the...Show more →
Thanks, I'll need the luck. I have definitely learned to lower my expectations for the 10D. The issue is that I do not want to lower my standards for photographic gear. I'm afraid that is what it will come to for the 10D. I go a little crazy inside justifying that price tag when I can beat the 10D's performance with a Rebel. I don't think I can move past that, and I expect that once returned, my 10D ends up on eBay while I can still get something for it.

One suprise about the EOS 3 is that it does not have a built-in diopter. Of course they can by purchased. I've been thinking about buying the eyepiece extender. I'll add that to my list. Thankfully, this stuff can also be used on my EOS 3, which I know I'll have in the future.

Then there's the way that 1.6x factor butchered my lens arsenal in terms of focal length coverage ... but I knew that before I got into this whole 10D mess. Yet here I was looking forward to the boost I would get with my long glass, only to find that's where the 10D's AF is at its worst, for my type of shooting.

Well, enough. I'll work at getting on with life.

Jul 15, 2003 at 07:15 AM

christo�
Offline
• • •
Upload & Sell: Off

p.2 #3 · Best 10D Bad AF Example Ever

Bad news on the eyepiece stuff -- the 10D parts don't fit the EOS 3. Niether diopters nor extender. The EOS 3 viewfinder is physically larger and they just don't fit. I was kind of bummed to have to buy more eyepiece stuff myself, but just another reason why an EOS 3 user should be making a zillion dollars shooting photos and think nothing of plunking down for a 1Ds, I guess.

Jul 15, 2003 at 07:25 AM

rlcramer
Offline
• • •
Upload & Sell: On

p.2 #4 · Best 10D Bad AF Example Ever

Canonizer - my 10D is in Jamesburg now. I woke up in a cold sweat one night and had them send my lenses back to me before they touched them and calibrated them for a bad machine, so at least they are safe. So right now, just the Camera is in their hands. I did calm down a bit and stopped short of selling all my canon gear. My wife convinced me to give it one more go around. I'm going to wait until the 10D comes back, and if it's not fixed I'm going to buy a 10D or maybe a 1D from a reputable dealer who will accept a return if the thing is broken. After that, I'm done with Canon. Who knows though, maybe my 10D will actually be fixed. I spoke to someone over at Canon yesterday and they told me that there was a 15 day glut of 10D's in for repair, but they are shooting to get that number down to a 5 day turnaround by next week. I asked if they got an approved fix from Canon or something, and she pleaded the 5th. Maybe I'm inferring too much, but from the way she was talking she made it seem like I shouldn't have to worry. As soon as I get the camera back, I will do a complete test and everyone here will know the outcome shortly thereafter.

Jul 15, 2003 at 07:55 AM

fred_jb
Offline
•
Upload & Sell: Off

p.2 #5 · Best 10D Bad AF Example Ever

DavidP wrote:
Yes, only one plane of focus.

However, expressing accuracy as a function of wide-open DOF is the ONLY thing that makes sense.

If you use a lens with a max aperture of f/1.4 vs. a max aperture of f/5.6, there's no way the AF can be expected to be more accurate in absolute terms with the f/5.6 lens. And that's whether you AF or MF.

Your ability (and the ability of the AF system) to discern where the actual plane of focus is depends ENTIRELY on being able to "see" where things are in focus and where they're out of focus. And THAT...Show more →

David - I agree with your sentiment, and the point about DOF wide open because of course it is under those conditions that the AF system operates - only stopping down if required to take the shot.

Howevert, the point I'm really making is that DOF itself is a fairly nebulous artificial construct, not an actual optical property - i.e. it depends on how you define circle of confusion, what print size you are going to agree to assess results at, etc, etc. In particular, a convention of a set print size, rather than a set magnification of the image capture area, be it film or a sensor, leads to all sorts of issues of comparability - having to use different CoCs which no one has agreed on, etc.

I think this does not sit well with trying to give a scientific definition of AF performance, expecially as with AF systems you are not directly measuring focus - instead you are measuring contrast and inferring degree of focus - not necessarily the same thing in many circumstances. I think to some extent this vagueness suits the camera manufacturers, but is not so good for consumers of their products! I think Canon and others exploit this.

Fred

Jul 15, 2003 at 09:04 AM

DavidP
Offline
• • • • •
Upload & Sell: Off

p.2 #6 · Best 10D Bad AF Example Ever

I don't see how you can express it any other way. Or at least a way that's not essentially equivalent.

The focus accuracy (in absolute distance) clearly depends on the wide-open aperture of the particular lens. It also clearly depends on how close you are to the subject and the focal length.

You simply can't say something like "AF accuracy will be to within 2.5 mm". It has to be expressed as a function of the DOF.

Exactly which CoC, final image size, viewing distance, etc (what goes into DOF) one uses is immaterial, as long as it's been specified.

Jul 15, 2003 at 09:24 AM

fred_jb
Offline
•
Upload & Sell: Off

p.2 #7 · Best 10D Bad AF Example Ever

Yes I agree focus accuracy has to be expressed as a tolerance with respect to some acceptable range and not in absolute distance. I guess I'm just not comfortable with the current definition of DOF. It is not well understood, open to creative interpretation by manufacturers, and does not take into account (at least as far as I know) different sensor sizes by having a set of industry standard CoC s for different sensor sizes.

Any expression of focus accuracy - i.e 10D supposedly within 1 x DOF is only worth the paper it is written for comparability purposes if all the components that go into the DOF definition used are agreed industry-wide. Is this the case? Do we even know if Canon use the same parameters for all their cameras film and digital?

Maybe it would actually be much better to agree one industry standard CoC and a set magnification factor from sensor to the output print used to assess DOF. Not sure what the other ramifications of this might be though, but the fewer things needed to be agreed on by all the manufacturers the better

Fred

Jul 15, 2003 at 09:40 AM

DavidP
Offline
• • • • •
Upload & Sell: Off

p.2 #8 · Best 10D Bad AF Example Ever

I'm not sure what Canon uses for its DOF definition.

I believe I read somewhere that they use a CoC consistent with a 35mm full-frame sensor (and even that is subject to a range of "acceptable values").

I think that most calculations I've seen for DOF (where final image size, etc. wasn't specified) implicitly assume 8x12 as the final output sized, viewed at a distance of 14.4", sqrt( 8^2 + 12^2 ). This seems to be the norm.

I agree it would be beneficial to explicitly state how the DOF is determined. If Canon truly has an acceptable focus accuracy of 1 DOF (that sounds way too loose a standard for my tastes) on the 10D, then surely they must internally know what they mean by DOF.

Jul 15, 2003 at 09:53 AM

fred_jb
Offline
•
Upload & Sell: Off

p.2 #9 · Best 10D Bad AF Example Ever

I guess that's the nub of the problem - we are all bandying around terms like DOF, which are crucial to any discussion of AF perfomance, without really knowing if we are all (including manufacturers) talking the same language.

Apart from the question of how you define DOF in the first place, I don't know whether the 1 x DOF I've seen quoted for the 10D means :

a) may focus anything up to an amount equal to the DOF in front or behind the intended focus point - this would be bad as anything more than +/- 0.5 x DOF would mean some shots are not sufficiently in focus for the plane of focus to be within the DOF (however that is defined).

or

b) the camera will focus somewhere within the depth of field as calculated for the parameters of that shot - still not very good IMO.

In either case obviously the bigger the aperture the more demanding that target is.

Jul 15, 2003 at 10:20 AM

DavidP
Offline
• • • • •
Upload & Sell: Off

p.2 #10 · Best 10D Bad AF Example Ever

I've always assumed that the "1 DOF accuracy" means b). Option a) is simply too comical to even consider being a possibility.

I've seen MANY cases where the plane of intended focus appears to be at the trailing or leading edge of the DOF. That's another reason I think b) is the actual answer.

This was purely a visual estimate of DOF (by looking at the grass, for example), done at whatever size was posted (sometimes small, sometimes large).

This brings up an interesting item of note: a reason why fast primes are better than slower zooms.

Say you want to shoot at f/2.8

By using a prime with an aperture of f/2 or f/1.4, then stopping down to f/2.8, you'll get more accurate focusing (in absolute terms) because the focusing is done at maximum aperture.

And it doesn't matter if it's AF or MF, either. Either way, the absolute error in focusing is diminished by using the fastest lens possible. Then if you can stop down some, focusing errors can be effectively hidden.

I think this is, in large part, why people say that a lens is better stopped down by about a stop. It's not just that the MTF improves, but that the inherent focusing errors (whether by AF or MF) are hidden by stopping down.

Jul 15, 2003 at 10:58 AM

californiajay
Offline
• •
Upload & Sell: Off

p.2 #11 · Best 10D Bad AF Example Ever

Even I was a bit skeptical at first but after an experience with the 70-200 F4 this past weekend, I am totally convinced that there is indeed an equipment problem.

I've owned the camera for about a month. Had both the 16-35 and 24-85 with no focus problems other than one which I mentioned in another thread (seems the camera focuses erroneously when there is some reflections in the focus area).

Last week I got a 70-200 F4 and put it to the test in Yosemite National Park this weekend.

Result? The lens just does not focus accurately at 70mm. It back focuses, period. It's ok from 100mm on up, but when I go below 100mm, it gets progressively worse towards 70mm. In order to assure sharp images at that focal length, I have to manually adjust the focus collar about 1/4 of an inch closer. It is consistant! The shots I found this out with were infinity focus shots (shots of El Capitan, Half Dome, whatever from across the valley or from another mountain ridge). After making this discovery, I tested with closer subject matter and found the same results. The lens would alway focus behind the subject at 70mm and I would have to manually adjust the focus ring about 1/4 inch or so each and every time. It's not even a matter of the accuracy being within the depth-of-field. It is off by the same amount of manual adjustment of the focus ring (again, about 1/4 inch).

Strangely it's only with this lens (of mine anyway), and only at 70-100mm. Makes me wonder what other lenses I may have a problem with and at what focal lengths.

The lens goes back tomorrow.

There IS a focus problem with the 10D guys. Maybe it's a camera problem, maybe it's related to certain lenses. I don't know, but enough of this "maybe you are not being careful with your focusing" and other excuse stuff. I've been a photographer far too long not to know what I am doing with focus.

Jul 15, 2003 at 08:12 PM

Joe Heiliger
Offline
•
[X]

p.2 #12 · Best 10D Bad AF Example Ever

Wow!! - glad I didn't just go out and grab a 10D now...

When I got my D60 I was aware that the autofocus would be slow.... Grandma was slow but hell she was old...

When the 10D was announced, I thought about selling my D60 rig and, with a little extra cash, getting a 10D, specifically because of the focusing speed. But, boy did you guys make me rethink that...

It just seems to me that Canon should do more testing before they release products, and then have a method of production that is consistent. I always thought is was funny that you never see any "snaffoos" with their "pro" products like the 1D/1Ds. Guess they figure if you are going to pay more than $5k (like $1500 isn't enough...) that you should get a quality product.

It is frustrating when you come over from film, for the equipment to seem so inferior in the digital world. I shot forever with A2's and EOS3 bodies which, in my opinion, were unbelievable machines that performed without flaw. I never had to worry about defects and inconsistencies when replacing a body because I knew what I would get... The same o'le reliability.

I will probably never leave Canon because I believe they are on the edge of R&D and will always be pushing the envelope. But, man they make it hard on people...

I'm just glad they don't design/produce aircraft or weapons... where the inconsistancies would really hurt.

Long live the Great White "L"ephant,

Jul 16, 2003 at 08:53 AM

Canonizer
Offline
• •
Upload & Sell: Off

p.2 #13 · Best 10D Bad AF Example Ever

Hey Bobby,

I joined the club and shipped off my POS 10D yesterday. I can't even imagine what will be returned to me. Instead of front focus -- back focus ... or even manual focus won't work .... or both? I'm scared enough just sending them the camera

I'm wondering how many 10D owners around the globe have been talking to their wives about these 10Ds?

Your scenario sounds just like mine. Hey wife, are these pictures (taken with the EOS 3 and lowly Rebels) in focus? Response: Of course. Hey wife, are these pictures (taken with 10D) out of focus? Response: Of course. Whew, I am sane. I mean, it's not like you have to look close or anything. My wife too suggested giving Canon at least one chance. I refuse to go to a 1D or anything even more expensive. I will not reward Canon for this.

Now that I think of it, perhaps the 10D is a Nikon plant?

So do the Canon techs call you when they receive your camera? Or do you have to call them every day? Or ... ?

rlcramer wrote:
Canonizer - my 10D is in Jamesburg now. I woke up in a cold sweat one night and had them send my lenses back to me before they touched them and calibrated them for a bad machine, so at least they are safe. So right now, just the Camera is in their hands. I did calm down a bit and stopped short of selling all my canon gear. My wife convinced me to give it one more go around. I'm going to wait until the 10D comes back, and if it's not fixed I'm going to buy a 10D or...Show more →

Jul 16, 2003 at 02:37 PM

rlcramer
Offline
• • •
Upload & Sell: On

p.2 #14 · Best 10D Bad AF Example Ever

Unfortunately I call them EVERY day, just to be annoying. They should email you when the camera is received, but they will never call you - even when they ship the unit back to you. I'm sure calling every day doesn't help, but you never know. My wife is SOOOO sick of hearing me whine about this camera that she is about to ebay everything for me!! I'm trying to tone it down for her sake, until I get the camera back. But this was a big investment for me, and I've missed a lot of picture taking opportunities due to this downtime.

Jul 16, 2003 at 02:42 PM

Canonizer
Offline
• •
Upload & Sell: Off

p.2 #15 · Best 10D Bad AF Example Ever

Joe Heiliger wrote:
Wow!! - glad I didn't just go out and grab a 10D now...

When I got my D60 I was aware that the autofocus would be slow.... Grandma was slow but hell she was old...

Don't get me wrong, I'm with you, but I gotta say, I can't even stand to hold the D60. It feels so cheap (to me) it's nasty. The body flexes when I squeeze it!? Then look thru the viewfinder and to see only 3 AF sensors, sheez we're going back a lot of years ... I wish I hadn't gone out and bought my 10D, but I would never have forked out the dough for the D60. Please know I don't mean that in an offensive way.

When the 10D was announced, I thought about selling my D60 rig and, with a little extra cash, getting a 10D, specifically because of the focusing speed. But, boy did you guys make me rethink that...

No, no, no. My 10D focuses on the wrong spot very quickly.

Seriously, it's not slow to focus. The AF speed has not disappointed me.

It just seems to me that Canon should do more testing before they release products, and then have a method of production that is consistent. I always thought is was funny that you never see any "snaffoos" with their "pro" products like the 1D/1Ds. Guess they figure if you are going to pay more than $5k (like $1500 isn't enough...) that you should get a quality product.

I never had a single problem with my Rebels G or 2000 either.

It is frustrating when you come over from film, for the equipment to seem so inferior in the digital world.

Yes, and for you to spend so much more for it.

I will probably never leave Canon because I believe they are on the edge of R&D and will always be pushing the envelope. But, man they make it hard on people...

I probably won't either, but I no longer have confidence in Canon products, and I believe they have turned a corner toward the worse in terms of QC. I bought two bum Canon products within a month. I will only buy from a dealer who I know will give me a full refund if the thing is bum. And there are such dealers.

Jul 16, 2003 at 02:55 PM

rlcramer
Offline
• • •
Upload & Sell: On

p.2 #16 · Best 10D Bad AF Example Ever

Canonizer - FYI. I just got off the phone with Canon and my repair is done. Camera is shipping back to me tomorrow, for delivery on Friday. Fingers crossed....

Bobby

Jul 16, 2003 at 03:06 PM

Canonizer
Offline
• •
Upload & Sell: Off

p.2 #17 · Best 10D Bad AF Example Ever

Great, Bobby! That was a lot faster than 15 days, wasn't it?

But did they call you, or did you call them, one they received your camera? Should I be expecting a call from them?

Jul 16, 2003 at 03:23 PM

rlcramer
Offline
• • •
Upload & Sell: On

p.2 #18 · Best 10D Bad AF Example Ever

I called them. They will never call you, no matter what.

Jul 16, 2003 at 05:36 PM

paganini
Offline
• •
Upload & Sell: Off

p.2 #19 · Best 10D Bad AF Example Ever

Same problem with me, but only at close focusing distances where the DOF is extremely shallow (and only at 70 mm). But at longer distances this is fine for me. i think your lens is probably off.

californiajay wrote:
Even I was a
Result? The lens just does not focus accurately at 70mm. It back focuses, period. It's ok from 100mm on up, but when I go below 100mm, it gets progressively worse towards 70mm. In order to assure sharp images at that focal length, I have to manually adjust the focus collar about 1/4 of an inch closer. It is consistant! The shots I found this out with were infinity focus shots (shots of El Capitan, Half Dome, whatever from across the valley or from another mountain ridge). After making this discovery, I tested with closer subject matter and...Show more →

Jul 16, 2003 at 06:06 PM

edamron
Offline
•
Upload & Sell: Off

p.2 #20 · Best 10D Bad AF Example Ever

This is a little off-topic I suppose, but when I show the focus points, none are ever highlighted (i.e., red) like the images in some of the other posts.

Is there something else I need to do to see the active AF points when I made the shot?

Thanks

Jul 16, 2003 at 08:05 PM

1 2 3 end
FM Forums \| Canon Forum \| Join Upload & Sell 1 2 3 end