Go Back   Steve's Digicams Forums > Digital SLR and Interchangeable Lens Cameras > Canon EOS dSLR

Reply
 
Thread Tools Search this Thread
Old Sep 8, 2004, 10:04 AM   #11
Senior Member
 
geoffs's Avatar
 
Join Date: Jun 2004
Posts: 2,025
Default

NHL, couldn't they have designed a new focusing module array for the 20D that incorporates different sensor types for the different focusing points? Perhaps the logic to handle diverse sensor types on one array module makes this too expensive, but then again perhaps not, maybe it makes it cheaper!
geoffs is offline   Reply With Quote
Old Sep 8, 2004, 1:32 PM   #12
NHL
Senior Member
 
NHL's Avatar
 
Join Date: Jun 2002
Location: 39.18776, -77.311353333333
Posts: 11,547
Default

The 20D's 9 AF sensors are arranged differently than the 10D:




Which is more similar to the Minolta 7D http://konicaminolta.com/products/co...ax7/ope02.html



than the elliptical arrangement of the 1D:





Cost is usually associated with volume, ie the more one makes the cheaper it get which is not the case with the 1D's.
NHL is offline   Reply With Quote
Old Sep 8, 2004, 1:47 PM   #13
Senior Member
 
Join Date: Dec 2002
Posts: 5,803
Default

So many questions.

I agree, NHL, they couldn't directly "borrow" part of the AF sensor of the 1D. But they could use it as the basis for making part of it worse. Clearly they haven't changed the aperture limitation (it's the same as the 10D, not the 1D.) If they could find a way to use some of the same parts as the 1D's AF (or the 1DMkII's) it would be cheaper for them inventory managment wise.

geoffs,

Yes, the 1D's AF system is more "Accurate". The 10D's AF system only guaranties that it will put the subject within 100% of the DOF. The 1D and the other pro bodies say they will put the focus point within the middle 1/3 of the DOF. It doesn't say where within the DOF it will be, so at some focal lengths it will be in the middle and in others shifted slightly one way or the other. I interpreter this to mean that they are better at detecting critical focus and so the edges of the DOF aren't in focus enough to satisfy it, so it does finer adjustments to get the subject into the sharper part of the DOF.

This is the difference between normal precision and high precision AF sensors. Canon Pro camera bodies have more high precision AF points and they work at smaller apertures (once the max aperture is exceeded they work like normal precision sensors.) The Pro bodies are still high precision with f4 lenses, while the amateur bodies are only high precision at F2.8. The idea of my 600mm f4 lens focusing at "high precision" makes me giddy.

It is also faster at making adjustments (within the limits of the lens's physical attributes) to get it into focus. And I believe the Pro bodies also don't hunt as much. It detects the direction that it is out of focus better and starts off in the right direction. Too often, my 10D will hunt out to the limit of its range and then start focusing in the other direction. Drives me nuts and costs me pictures (I've actually wondered if something isn't just wrong with my camera, but I need to find someone with a 10D that I can meet up with and do some tests.)

It almost sounds like the 20D has a slightly better high precision AF sensor than the 10D. It still has the aperture requirement but it focuses better with the right lenses. This is just speculation, though.

I agree completely with that link that NHL posted about how AF works on the Minolta cameras. As far as I know basically all cameras do it that way now (probably not all, come on, someone correct me!) To my mind, it makes a lot of sense. If you have a really small DOF (smaller f-stop), then when you place the subject in it and stop down the subject will be closer to the center of the DOF. This is what people usually want. Also, with a smaller DOF, I'ts easier to detect when you're not in it.

Most of what I said above comes from here:

http://www.bobatkins.com/photography/eosfaq/eos3af.html

which is a great page about Canon AF systems in their film bodies. Most (if not all) apply to the digital bodies too. Now I haven't read that page in awhile, so if I'm made any mistakes, please correct me!

I believe ALL of canon cameras have a mix of normal and high precision AF points. In the 10D, only the middle AF point is high-precision (and only at f2.8.) I don't recall what it is on the 20D. The 1D has the middle... 7? points as high precision. Something like that.

Eric
eric s is offline   Reply With Quote
Old Sep 8, 2004, 2:37 PM   #14
NHL
Senior Member
 
NHL's Avatar
 
Join Date: Jun 2002
Location: 39.18776, -77.311353333333
Posts: 11,547
Default

There's a "slight" nuance:
The Canon center AF sensors are the + cross type
The Nikon are combined -\- slighly slanted...
The Minolta are both + and x cross type ! :-)

http://konicaminolta.com/products/co...x7/ope01a.html
NHL is offline   Reply With Quote
Old Sep 8, 2004, 3:25 PM   #15
Senior Member
 
geoffs's Avatar
 
Join Date: Jun 2004
Posts: 2,025
Default

Ok, the net result of all of this discussion is that I now understand the whys and hows of the 3x more precise AF statement. BTW, Eric, that article on Bob Atkins site is very good and he also has his own review of the 20D posted (he likes it and is going to get one even though he already has a 10D).

We can debate the pros and cons of which camera manufacturer's AF sensor arrangement and types will yield the best AF result. I bet on dpreview such a discussion could become quite heated. So, enough of that.

I've got one more question that's arisen since reading some of the articles pointed to here. When an AF scheme is said to be 3x more precise when the lens used has a max aperture of, let's say, f/2.8 or greater, does that mean the more precise AF will still be utilized even though that lens is stopped down to a much smaller aperture for the actual photograph?
geoffs is offline   Reply With Quote
Old Sep 8, 2004, 3:33 PM   #16
NHL
Senior Member
 
NHL's Avatar
 
Join Date: Jun 2002
Location: 39.18776, -77.311353333333
Posts: 11,547
Default

geoffs wrote:
Quote:
When an AF scheme is said to be 3x more precise when the lens used has a max aperture of, let's say, f/2.8 or greater, does that mean the more precise AF will still be utilized even though that lens is stopped down to a much smaller aperture for the actual photograph?
Correct the AF focusing is always done prior to the camera stopping the lens down for the actual shot when the mirror is up and you can't see anyway...

Unless you press the DOF pre-view button on the side of the camera (or mount a polarizer up front!).
NHL is offline   Reply With Quote
Old Sep 8, 2004, 3:36 PM   #17
Senior Member
 
geoffs's Avatar
 
Join Date: Jun 2004
Posts: 2,025
Default

Thank you, NHL! That is precisely (pun intended) what I wanted to know!!!

Uh oh, one more question... Does this work for 3rd party lenses with Canon AF capability and that meet the requirements (ie: f/2.8 or greater max aperture)?
geoffs is offline   Reply With Quote
Old Sep 8, 2004, 4:24 PM   #18
Senior Member
 
Join Date: May 2003
Posts: 577
Default

NHL wrote:
Quote:

Correct the AF focusing is always done prior to the camera stopping the lens down for the actual shot when the mirror is up and you can't see anyway...

Unless you press the DOF pre-view button on the side of the camera (or mount a polarizer up front!).
What has a polarizer to do with this? I don't get it.

Barthold
barthold is offline   Reply With Quote
Old Sep 8, 2004, 5:49 PM   #19
NHL
Senior Member
 
NHL's Avatar
 
Join Date: Jun 2002
Location: 39.18776, -77.311353333333
Posts: 11,547
Default

Quote:
Uh oh, one more question... Does this work for 3rd party lenses with Canon AF capability and that meet the requirements (ie: f/2.8 or greater max aperture)?
All optical principles work the same no matter who build the lenses (Canon or 3rd parties)



Quote:
What has a polarizer to do with this? I don't get it.
There's a filter factor associated with a polarizer (or an ND filter) which will reduce the amount of light reaching the AF sensors (affecting the exposure and not the f-stop) -> this will affect the sensitivity of the phase detectors, similar to using the same lens in a darker environment...

This is different from the effect of attaching a 1.4x or 2x TC which effectively change the ratio of the barrel against the "new" multiplied focal lenght which will feed the "corrected" largest f-stop to the camera.
NHL is offline   Reply With Quote
 
Reply


Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off



All times are GMT -5. The time now is 5:29 PM.