Better 10D focus test for Pearl Rider

I do not know

i called Canon and told them the whole story. This is before i got the second new body which is a differnt issue. I expalined to Canon about the 50mm lens working ok after repair and now the new 17-40L not working. I expressed concern about sending in my camera with the 17-40L lens since by the time they were done fixing it it would be past the 7 day return policy(of the new lens) of the store i bought the 17-40L from and i was nervous about Canon's ability to fix my problem. The person i talked to at Canon a higher level than the initial person who answered the phone understood my concern and recommended i return the 17-40L to the store to be safe an resend my camera to canon for adjustment with the 50mm lens. WHAT? thats what i said why do you want my 50mm lens if it works fine with the camera remeber it is the 17-40L that was having the problem she answered the problem was likely secondary to my camera. I proclaimed my camera had already been serviced and fixed as evident by my 50mm lens working she said "i am pretty sure your camera is still the problem and that there was a new software released a few weeks ago that we didn't have the first time you sent the camera in"i have already posted this story. and people on this forum think the new software she was talking about is probably some type of calibration software?

What's interesting is; i asked her why this time they wanted me to send in the 50mm lens and not my camera by itself ? she had no good answer for this.

Hellow DavidP,

I have seen some facts:

1. D30, D60, 1D, 1DS have no problems with autofocus
3. D30, D60, 1D, 1DS ETTL don't work very well and you need FEL.

2. 10D has autofocus problems
3. 10D ETTL don't need FEL.
Why? As far i know is the same algorithm (it's not called ETTL+)
But works in a different way. What different way?

4. 300D ETTL seems to work like D60 ETTL
5. 300D has autofocus problems? we will see. I think not.

6. Phil has not detected any problem with 10D autofocus system. The camera come from Canon.
7. Any "official" test has showed any problems with 10D autofocus.
I supose cameras come from Canon.
8. In this tests ETTL was not tested.

9. You, me and other people have evident problems with 10D autofocus.

What is happening? Any ideas?

Dr. Watson

Thank you. I think he's a bit nuts sometimes. Highly intelligent but nuts.

M

rlindabury said:

Nothing against you David but you gotta admit that's a real
stretch. Your original statement was that all the AF sensor points
were 4x the size of what's etched in the viewfinder. (stated as
FACT)

In reality, not a single one is 4x the etched size unless you start
doing some kinda funky major rationalizations.. like the one below.
lol

Go figure..

DavidP said:

Now, if you take an AF square defined by a square the covers the
extent of the AF sensors, it's 4x as large as the one etched in the
viewfinder.

Mark Gillett said:

Just looked again, make that 200% and 80% for the small ones.

Click to expand...

--
http://www.neonlightsimaging.com/artshow/final.htm
Extrapolation from few solid data points is best left to those with
years of training and experience in such things.

Click to expand...

--
Bob Lindabury

Click to expand...

Ok.. I'll go along with that. Based on the images of the etched points vs. the actual sensors, the 4x didn't fly. If what you're saying is true, then fine, I can believe the AF areas are 4x the size of the what's etched.

No problem here on that front. I'm not looking to knock you down David, just trying to get clear on a couple of points.

My only problem with Bill is in his tests. He's done a ton of tests and not a single one of them are the test you, Mishkin and Canon recommend. Why is that? Why do I only see doorknobs and diagonal rulers?

It was pretty much proven the ruler test was bogus months ago and that EVERYONE who even thought they might have a AF issue use the Canon test. But do we see that, nope.

If I saw that test at the correct distance, with the correct light, on the test chart, I wouldn't doubt Bill for a second and I doubt anyone else would. But what do we get, two pieces of paper taped to a window making a dig at 'ol Pearly.

Matter of fact, I barely doubt Bill at this point. I just want him to get his damn camera fixed or to get a refund. LOL

DavidP said:

And while you're busy contemplating that, go read what Chuck
Westfall had to say on the issue.

The AF will look outside what's shown in that image as the AF
sensor. When will it do this? Chuck didn't / can't reveal that.

But, the actual AF area is much larger than what's inscribed on the
viewfinder. And certainly occupies more area than the skinny
little sensor sketches indicate, too. Maybe not so much along the
long axis, but perpendicular to that axis.

Click to expand...

--
Bob Lindabury

I'm sure not. I have never had the privelage of a 3 or higher. I would love to, though. I just know that I am happy (ish) with mne. I took some falcons today, in flight also, and they came out really well. I used all 7 points on the AI Servo for the first time. This seemed to work better that just one point.

M

weirdrob said:

Mark Gillett said:

M

weirdrob said:

bill jones said:

An experienced at least by time criteria semi-pro (have owned my
own photography buisness) buys a 10D to replace his aging nikon
F-100

Click to expand...

The F100 is very similar to my EOS-3...though the F-100 is uaually
considered to be a better body. The comment I have to make is that
my EOS-3 focuses better (quicker and with greater accuaracy) than
does my 10D. I would expect that the F-100 would also focus much
better than the 10D.

Could your focus issue be a case of unreasonable expectations...?
This is not an accusation...just trying to expore the situation.

I've had to learn how to compensate for the differences between the
two bodies (3 and 10D). Sometimes this does require a touch of MF
on the 10D...with some subjects. This was , of course, more
pronounced with the 70-200/2.8 that I borrowed than with the lenses
I own.

Reading these threads has made me realize that the main difference
might well be the width and height of the center + AF sensor as
compared to the size of the image sensor. The AF sensor is larger
than the red square and it will "reach out" and grab something else
as a focus point. I shoot MF, a little 4x5, and use a "D" screen
(plain grid) on my T90...so I have learned to focus without the
normal aids...just the fresnal screen. You can learn to judge focus
and then fine tune it with a touch on the focus ring.
--
Rob Wierman
http://www.pbase.com/weirdrob/001_10d

Click to expand...

Click to expand...

My 3 focuses quicker and locks in focus better than the 10D...just
as it does a better job than my other EOS bodies. I really notice
the difference with my 300/4 IS L that I have used with my 10s, 10D
and 3. I actually bought the 3 because I was so unhappy with the
10s AF. I was shooting better with my T90 and FDn 300/4 L.

I imagine that the 5 and 50E (A2 & Elan IIE on this side of the
pond) probably don't AF much better than my 10s. Neither hold a
candle to the 3.
--
Rob Wierman
http://www.pbase.com/weirdrob/001_10d

Click to expand...

This indeed may be the case and i may be a better canidate for the 1Ds but i am a little nevous to invest that much money after this ordeal with the 10D. One lesson learned is until one is sure there are no possible widespread flaws in a camera design( and i am not necessarly saying there is one with the 10D) you should probably buy your camera from a local dealer with a good return policy rather than on line although PC Nation has really been outstanding so far in trying to work out this problem with me they exchanged the first camera even after 5 months since purchase after they listened to my story and spoke to their Canon rep.

weirdrob said:

bill jones said:

An experienced at least by time criteria semi-pro (have owned my
own photography buisness) buys a 10D to replace his aging nikon
F-100

Click to expand...

The F100 is very similar to my EOS-3...though the F-100 is uaually
considered to be a better body. The comment I have to make is that
my EOS-3 focuses better (quicker and with greater accuaracy) than
does my 10D. I would expect that the F-100 would also focus much
better than the 10D.

Could your focus issue be a case of unreasonable expectations...?
This is not an accusation...just trying to expore the situation.

I've had to learn how to compensate for the differences between the
two bodies (3 and 10D). Sometimes this does require a touch of MF
on the 10D...with some subjects. This was , of course, more
pronounced with the 70-200/2.8 that I borrowed than with the lenses
I own.

Reading these threads has made me realize that the main difference
might well be the width and height of the center + AF sensor as
compared to the size of the image sensor. The AF sensor is larger
than the red square and it will "reach out" and grab something else
as a focus point. I shoot MF, a little 4x5, and use a "D" screen
(plain grid) on my T90...so I have learned to focus without the
normal aids...just the fresnal screen. You can learn to judge focus
and then fine tune it with a touch on the focus ring.
--
Rob Wierman
http://www.pbase.com/weirdrob/001_10d

Click to expand...

hehe.. as far as I know, DavidP doesn't have any evident problems with 10D autofocus because he uses a 1D I believe.

yqs said:

Hellow DavidP,

I have seen some facts:
9. You, me and other people have evident problems with 10D autofocus.

Click to expand...

--
Bob Lindabury

..Its Sunday evening, I have done all my processing for the day, there is nothing on any of the 26 stations on TV, I have eaten, so what is wrong with a bit of forum banter??

M

BryanS said:

Good God, men. You seriously don't have anything better to do than
to carry on this childish bantering back and forth? This has
gotten to a point where most of us just PITY you.

Give it a freakin' rest already. Yeesh.

--
BryanS

Click to expand...

They should've had you try another 17-40/4 first.

At any rate, have you sent anything else back yet? At the end of your post, you left off where you'd sent your body and 50/1.8 (the first time for both).

Have you sent anything else back yet? You have a second body now, as I recall? And they took the first one back?

I'm gonna have to start taking notes. I don't know if you've told the whole story yet in one concise post. I'm having to piece together things from various posts now.

I'd have to suggest (agian) just sending it ALL back in to Canon. Assuming you haven't already done so.

--
http://www.neonlightsimaging.com/artshow/final.htm

Extrapolation from few solid data points is best left to those with years of training and experience in such things.

The reason this is so fustarting is because the 10D is such a good camera
I have taken many pictures that are amazing(at least to me)
Great color, low noise with the24-70L amazing contrast.
It was a dream when the focus works or the depth of field is large.

Imagine the fustartion when you have such a capable tool but it is flawed by what appears to me(stresssed ME, the truth i could not possibly knows only Canon Knows)
a quality control issue not a defect in design.

DavidP said:

I think you'd find that the outer sensors on the -3 and -1 series
would be much closer in performance to the center 10D focus point.

weirdrob said:

My 3 focuses quicker and locks in focus better than the 10D...just
as it does a better job than my other EOS bodies. I really notice
the difference with my 300/4 IS L that I have used with my 10s, 10D
and 3. I actually bought the 3 because I was so unhappy with the
10s AF. I was shooting better with my T90 and FDn 300/4 L.

I imagine that the 5 and 50E (A2 & Elan IIE on this side of the
pond) probably don't AF much better than my 10s. Neither hold a
candle to the 3.

Click to expand...

--
http://www.neonlightsimaging.com/artshow/final.htm
Extrapolation from few solid data points is best left to those with
years of training and experience in such things.

Click to expand...

My post wasn't at issue to the specifications of the sensors...just the ability of the various cameras to function in comparison with each other. I could care less if the sensors in my camera were the worst on the planet...if they could AF faster and lock better than any of the others. Kind of a "bottom line" attitude I have about most things.

No doubt about it...the sensors need to be better in DSLR's. They will be better in time. It's only a cost factor. Look at the price vrs. technology jumps we've seen in the past few years.

I'm still waiting for a 3D...
--
Rob Wierman
http://www.pbase.com/weirdrob/001_10d

Yes, the bottom line is what counts.

But if the camera functioned better than most others, I would find it highly doubtful that the camera used the worst sensors on the planet.

Anyway, just trying to explain WHY the -3 and -1 focus better/faster than the other cameras.

weirdrob said:

My post wasn't at issue to the specifications of the sensors...just
the ability of the various cameras to function in comparison with
each other. I could care less if the sensors in my camera were the
worst on the planet...if they could AF faster and lock better than
any of the others. Kind of a "bottom line" attitude I have about
most things.

Click to expand...

--
http://www.neonlightsimaging.com/artshow/final.htm

Extrapolation from few solid data points is best left to those with years of training and experience in such things.

If I follow you correctly, you're trying to make a case for the fact that the improved 10D's improvement in E-TTL is having an adverse affect on its AF performance.

I'm not sure there'd be any reason this has to be true, because the E-TTL is really tied to the metering system (number and size of meter points) and not the AF system, per se.

yqs said:

I have seen some facts:

1. D30, D60, 1D, 1DS have no problems with autofocus
3. D30, D60, 1D, 1DS ETTL don't work very well and you need FEL.

2. 10D has autofocus problems
3. 10D ETTL don't need FEL.
Why? As far i know is the same algorithm (it's not called ETTL+)
But works in a different way. What different way?

4. 300D ETTL seems to work like D60 ETTL
5. 300D has autofocus problems? we will see. I think not.

6. Phil has not detected any problem with 10D autofocus system. The
camera come from Canon.
7. Any "official" test has showed any problems with 10D autofocus.
I supose cameras come from Canon.
8. In this tests ETTL was not tested.

9. You, me and other people have evident problems with 10D autofocus.

What is happening? Any ideas?

Click to expand...

--
http://www.neonlightsimaging.com/artshow/final.htm

Extrapolation from few solid data points is best left to those with years of training and experience in such things.

And I'm formatting and doing ScanDisk on a bunch of Jaz cartridges I'm about to put on Ebay.

Yuck, this is worse than scanning film! LOL

Mark Gillett said:

..Its Sunday evening, I have done all my processing for the day,
there is nothing on any of the 26 stations on TV, I have eaten, so
what is wrong with a bit of forum banter??

Click to expand...

--
http://www.neonlightsimaging.com/artshow/final.htm

Extrapolation from few solid data points is best left to those with years of training and experience in such things.

DavidP said:

Do you really think you could've seen the difference through the
viewfinder on the 10D?

As bad as the AF is on some of these cameras in certain situations,
it still beats unaided MF for me (and most others).

weirdrob said:

Reading these threads has made me realize that the main difference
might well be the width and height of the center + AF sensor as
compared to the size of the image sensor. The AF sensor is larger
than the red square and it will "reach out" and grab something else
as a focus point. I shoot MF, a little 4x5, and use a "D" screen
(plain grid) on my T90...so I have learned to focus without the
normal aids...just the fresnal screen. You can learn to judge focus
and then fine tune it with a touch on the focus ring.

Click to expand...

Click to expand...

Better at f/1.8 than at f/4...and this was f/1.8 in bright light. Siince I haven't tried the shot, I'm not sure if I could MF that scene or not...but, I would expect that I could.

My 50/1.8 arrives later this week and then I'm off to Egypt for two weeks. I'll try it when I get back. Maybe Dr. Jones can provide the point size so I can print up the test targets and duplicate the shoot.
--
Rob Wierman
http://www.pbase.com/weirdrob/001_10d

Hellow DavidP,

I have a theory. I am very interested about your opinion about it.
Please be patient, my english is very poor.

My point of view is

1. The "AF big sensor area" is the same area that ETTL system uses to metering the flash output around the autofocus point.

2. The camera focus in the MOST contrast point inside the "AF big sensor area", not inside the "AF square" point.

3. I think Canon CAN change the size of the "AF big sensor area". In fact, i think they do this when they calibrate a camera.

Some personal conclusions

1. ettl vs autofocus

If you make the "AF big sensor area" bigger then ETTL metering is like a "partial metering" and it is more acurate. You don't need to FEL when you recompose.

But, in this case, the autofocus is not working properly because there are more probabilities that the camera find a point inside the "AF big sensor area" with more contrast than the point inside the "AF square" point.

If you make this area more little then the autofocus is better, but the ETTL metering is like a "spot metering" and is less acurate. If you recompose, then it doesn't work.

2. Contrast

Altought we focus in a point with enough contrast, we have always the probability that there's a point with more contrast inside the "AF big sensor area". And we can't see this big area.

3. Wide and tele
For example, with a 70-200

• With 70 the objects are more little. Then we have more probability to find a more contrasty object inside the "AF big sensor area" and more probability that autofocus fails.
• With 200 the objects are (in proportion) bigger. Then we have more probability that the object in the square fills the "AF big sensor area" and autofocus doesn't fails.

4. D60 vs 10D vs 300D
I remember all the past post about D60 ETTL.

I supose Canon has tried to fix it in 10D, but... then the autofocus system is not too acurate.

I think the 300D system (autofocus and ETTL) is going back to D60 system.

5. Official tests

With test about autofocus, Canon change the "AF big sensor area" size to be nearly the AF square area. Then the autofocus is very very acurate. And in this test nobody think about ETTL.
I think, Phil tests this kind of cameras.

Please, what do you think about this theory?

DavidP said:

I'm not sure there'd be any reason this has to be true, because the
E-TTL is really tied to the metering system (number and size of
meter points) and not the AF system, per se.

yqs said:

I have seen some facts:

1. D30, D60, 1D, 1DS have no problems with autofocus
3. D30, D60, 1D, 1DS ETTL don't work very well and you need FEL.

2. 10D has autofocus problems
3. 10D ETTL don't need FEL.
Why? As far i know is the same algorithm (it's not called ETTL+)
But works in a different way. What different way?

4. 300D ETTL seems to work like D60 ETTL
5. 300D has autofocus problems? we will see. I think not.

6. Phil has not detected any problem with 10D autofocus system. The
camera come from Canon.
7. Any "official" test has showed any problems with 10D autofocus.
I supose cameras come from Canon.
8. In this tests ETTL was not tested.

9. You, me and other people have evident problems with 10D autofocus.

What is happening? Any ideas?

Click to expand...

--
http://www.neonlightsimaging.com/artshow/final.htm
Extrapolation from few solid data points is best left to those with
years of training and experience in such things.

Click to expand...

many of us are over looking an important point regarding the focus issues i have been having.

My camera is outstanding at locking in on the target i aim it at . It almost never hunts for the target like my nikon did in low light or low contrast situations. In fact the focus on this camera is amazing on how fast it locks on the target and how well it works in low light. My cameras error must somehow be related to calibration since it always front focuses the same amount. I can shoot the ruler 100 times and every time it front focuses. The last one back focused consistently. This to me indicates its some type of quality control issue or some issue between the lenses and the camera and not the inherent quality of the autofocus system. This is what is called a systematic error.
I don't no anything about engineering or at what level this occurs..

I think this is why people who have a functioning 10D are so critical when any of us compalin of a focus problem because i could only imagine how amazing this camera must be if it is calibrated correctly based on the speed in which my camera focuses incorrectly consistently

DavidP said:

Hmm, Jim hasn't returned HIS 10D, either. Is it possible that
those who bought from Houston Camera Exchange either haven't
noticed a problem, or did so AFTER they could exchange it, or sent
it off to Canon for repair, etc, etc, etc.

Click to expand...

If they had a problem and called THCE, they would've volunteered to send it in for them this saving them the shipping costs. You think they WOULDN'T want to do that?

DavidP #28649 said:

I doubt they sell that many of them, anyway.

Click to expand...

Well, they've sold a LOT of them and could sell much more if they could get them. They had been getting them in lots of 10 and selling them usually within a week. When I got the call about them being in, they'd gotten 10 in that day, before I could call back at 1:30, they were down to 3.

DavidP said:

It has been the experience of many that you can't talk to a service
tech UNLESS you scream at the appropriate people.

Click to expand...

And of course, you omit the experiences of many others who didn't have any issue with Canon service. False impressions given by omission.

I realize this new tets is a little rough since i shot indoors lthough the light is pretty good
i will repeat in am outside but all in all same results of front focus
the red line is at the depth of the paper and at 18.5 notice the front focus