Pages: [1]   Go Down

Author Topic: 20D focus problem identified, solution not  (Read 4508 times)

barriecourtenay

  • Newbie
  • *
  • Offline Offline
  • Posts: 2
20D focus problem identified, solution not
« on: August 27, 2005, 08:08:14 am »

I suggest you are expecting too much from a digital camera,just get taking photo's and enjoy,its probally all in your mind
Logged

Ronny Nilsen

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 361
    • The Quiet Landscape
20D focus problem identified, solution not
« Reply #1 on: August 29, 2005, 08:16:57 am »

I have noticed the same on my 20D, but the sensor seems to be a bit above  
the marking, not to the side.
Logged
Ronny A. Nilsen
www.ronnynilsen.com

Ronny Nilsen

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 361
    • The Quiet Landscape
20D focus problem identified, solution not
« Reply #2 on: September 01, 2005, 02:47:16 am »

I guess it's the way things are in a consumer body. The viewfinder have about 95% coverage -> makes the camere cheeper to produce. Since the viewfinder is not 100% it's possible to to be more "sloppy" when bulding the viewfinder and still get what you see in the picture + something. I guess this means that one should not ecpext the AF points to match up 100% in a consumer body.

I have a EOS 1 body that I have used for 14 years now, and I still miss the viewfinder from that camera. 100% what you see is what you get! But the price of a 1D MkII compared to the 20D was to much considering that the viewfinder was the most important thing to me that was missing from the 20D. I'll probably buy a 1D something in the future when there is a version that is the same size as the 5D with optional battery grip.

But as you say, if it's to much off, it should be sent back for adjustments.
Logged
Ronny A. Nilsen
www.ronnynilsen.com

drh681

  • Newbie
  • *
  • Offline Offline
  • Posts: 39
20D focus problem identified, solution not
« Reply #3 on: September 01, 2005, 08:28:58 pm »

yeah.
 all it takes is one guy doing a wrong assembly adjustment for a couple weeks to really foul up the quality control system.
especially if there is a real hard push to get units out the door by a specific date!

I find it amusing and a bit annoying that so many people dont think those things can happen to Canon too.

appearantly Canon has decided that they will "just fix it when it comes back"
Logged

Husnu

  • Newbie
  • *
  • Offline Offline
  • Posts: 14
20D focus problem identified, solution not
« Reply #4 on: August 25, 2005, 10:08:19 am »

I have a problem with my Canon 20D’s focusing. It’s been in the local Canon technical service for a few days. They confirm the problem, and they say they checked some of the new  bodies they have in stock, and all was acting the same. They say they will investigate what to do and let me know. I’d appreciate if you’d share your thoughts and a  possible solution if you know any. (or someone to write to about this at Canon).

The problem is basically this; the focus sensors seems to be offset with regard to the focus marks by about one center focus width to the left.

The camera is about 10 months old. I was getting some back focused photos from time to time, but casual focus tests (like shooting a newspaper on the wall) showed no problems. I was blaming my self for reframing after focus lock and the like. Recently it started to get on my nerves and I looked deeper into the problem.

The problem didn’t occur when the target subject was relatively big compared with the focus marks, say as big as the partial metering circle. It happened when a single focus point is selected and the target is covering a small area , maybe a little bigger then the focus mark, say you are shooting the photo of someone from a distance, and you target your subject’s right eye, so that the center focus square fully covers the eye and the surrounding area. This should mean that even if the sensor size is bigger than the focus mark, most of the sensor is seeing the target and maybe a little is seeing the background, and should focus on the closer subject, which is the targeted subject. NO, it chooses the focus on the background, even if the focus mark is totally on the target and the target has enough detail to focus on.  

So I ended up doing the following test to see what was going on.
1-I selected a pencil which I managed to place upright on the table. Adjusted camera to subject distance so that the width of the pencil just filled the width of the center focus mark.  
2- First I positioned everything so that the background was just an empty wall.
3- I press the shutter button and the focus locks right on the target filling the focus mark. Good
4- I turn the camera to the right a little so that the target is just outside the center focus mark to its left. I pres the button, and focus locks. Good. Also proves that the sensor is actually bigger than the mark.
5-.Try step 4 with the subject just to the right of the center focus mark. I expected it to react like it did in step 4, since the sensor is bigger. I pres the button. Miss. Try again. Miss. Miss, miss. No good.
6- Try to place the target about one center focus mark width away from the mark to the left (like step 4 with more distance) .Press the button , and it focuses right on.

Conclusion;
1-   Center sensor width is about 3 times the width of the center mark.
2-   Center focus mark is not centered and only covers the right hand side 1/3 of the sensor.

Test with detailed background to cross check.
1-   When the small target is filling the center focus mark; target missed, background is in focus, every time.
2-   When target is just to the left of the center mark; focus is locked on target, no back focus. Sometimes the focus is missed, but this is normal I guess because the target is so small, but the success rate is very high.
3-   When the target is just to the right of the center mark; focus is missed, background is in focus every time.
4-   When target is to the left of the center focus mark about 1 center mark width; Target missed, background in focus, as expected.
 

I also checked with the camera turned upside down, to eliminate any inconsistency in the results due to light positions. The results are not changed.




I live outside the US and the local Canon people could not help at the moment. I am reluctant to send the camera to the US at the moment. If someone here had a similar problem and could have the camera fixed, I’ll be happy to know.


Thanks for reading.
Regards,
Husnu
Logged

Husnu

  • Newbie
  • *
  • Offline Offline
  • Posts: 14
20D focus problem identified, solution not
« Reply #5 on: August 27, 2005, 09:25:59 am »

The local technical service has already confirmed that there is a problem with my camera. They don't yet know how to fix it. In the meantime I am enjoying my EOS3 as I have my first EOS in 1991, and my EOS 1, and EOS 1N, none of which had this sensor-mark misalignment. I even enjoy the 20D as long as I remember to use the shifted imaginary focus point with certain subjects. With greater subject-background distances, the problem can be observed through the viewfinder without the need to take a photo, so the camera's being digital is irrelevant.
Logged

drh681

  • Newbie
  • *
  • Offline Offline
  • Posts: 39
20D focus problem identified, solution not
« Reply #6 on: August 31, 2005, 07:12:18 pm »

dont expect a lot of support for your "auto focus issues" around here. (nor on most other boards)

the most common remark will be along the lines of "you aint doing it right!".
This is in spite of a responder's total lack of knowledge of the poster's experience with Canon cameras!
the next most common will be "you need to do more post processing; shoot in RAW; use a tripod...  and other sophomoric pseudo advice.

I'm here to tell you that there are problems in Canon's assembly line for 20D bodies. Misaligned sensors, miscalibrated AF...

it all comes with meeting shipping commitments before quality control samples are tested.

My only suggestion is; send it for adjustment.
Logged

ARD

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 296
    • http://
20D focus problem identified, solution not
« Reply #7 on: September 01, 2005, 03:43:05 am »

There is a problem with the front focusing on the 20D, and I suspect it is a lot more common than many people think. I had a 20D until a few weeks ago, but it was suffering the same. After looking at a few in my local shop, there were all showing the same problem.

My supplier somehow managed to strike a deal with Canon whereby they exchanged the unit for a 1D MKII. A shame because I really liked the 20D.

Like previously said, in the mass production market, I would imagine that if a fault exists it gets passed down the line to many units.

It might be a good idea if you contact some of the photography magazines about this. If they do a write up on it, Canon may feel obliged to look into this problem properly, and if needed issue a product recall for affected serial numbers............................................but then again it might not ???
Logged
Pages: [1]   Go Up