Canon autofocus information

FM Forum Rules
Canon SLRs, primes, and zooms lenses reviews

FM Forums \| Canon-mount SLRs \| Join Upload & Sell	Search Used
1 2 3 4 5 6 7 end

Archive 2005 · Canon autofocus information

neyoliv
Offline
Upload & Sell: Off

p.4 #1 · Canon autofocus information

Does anybody have a link to a image of the 20D sensor arrangement overlaid on top of the sensor indicators in the viewfinder?
I've found this image:

This image is copyrighted by the owner

from this page:
http://cweb.canon.jp/camera/eosd/20d/catalog/index06.html
but couldn't decipher what it means.

Thanks,

Ney Oliveira

Jul 18, 2005 at 03:30 AM

gsnethen
Offline
Upload & Sell: Off

p.4 #2 · Canon autofocus information

neyoliv wrote:
Does anybody have a link to a image of the 20D sensor arrangement overlaid on top of the sensor indicators in the viewfinder?
I've found this image:

This image is copyrighted by the owner

from this page:
http://cweb.canon.jp/camera/eosd/20d/catalog/index06.html
but couldn't decipher what it means.

Thanks,

Ney Oliveira

If you look at the rectangles in the viewfinder, you'll see that the top and bottom are horizontal rectangles, the center is a square and the rest (the three on the left and the three on the right) are vertical rectanges.

The shape of the rectangle represents the alignment of the sensor. The horizontal rectangles (top and bottom) represent a horizontal sensors which focus best on vertical lines.

The vertical rectangles (three on left and three on right) represent vertical sensors which focus best on horizontal lines.

The center square uses both sensors (depending on the aperature of the lens, as described earlier in the thread).

This arrangement makes sense to me, as the sensors along the center horizontal line are aligned to focus on the horizon or other horizontal features and the sensors along the center vertical line are aligned to focus on vertical features (people, poles, trees, tall buildings, etc.)

You can easily test each AF sensor by setting it as the active AF point and then aiming the camera at a horizontal or vertical boundary between two solid evenly-lit colors and trying to focus. You can even use the windows on your computer monitor.

---Gary Snethen

Jul 18, 2005 at 06:41 AM

TheJaff
Offline
Upload & Sell: Off

p.4 #3 · Canon autofocus information

That Sigma said that Canons AF stop at "the first part of focus" does not correspond with what Ive seen:

I recently got a Sigma 28-70mm 2.8 EX DG and a Sigma 70-300mm APO DG for my 350D. I jumped right in doing the ol' AF-test with chart, tripod and IR-remote. I took shots at apertures 2.8 up to 8.0 at 28, 38 and 70mm. Manual "defocusing" between each shot, both to infitity and "near-focus".
Distance to chart about 40cm (33cm is the minimum for the 28-70).

I pretty much instantly saw that the lenses didnt focus properly with AF on my camera. The focus locked on quite fast and without chasing (but with some noise as expected). But it is as if the AF-motor didn't stop fast enough and focus ran past the correct point. Eg, when starting from infinity I always ended up frontfocused about 1.5-2cm at ALL apertures. When starting from "near-focus" I always ended up back-focused about 1.5cm at ALL apertures.
This clearly results in a soft image in the plane where the focus shold be.

I havent tested the 70-300 as thoroughly (yet) but it seems to exhibit the same problems but, help me with the math here, 1.5-2cm error from 2m distance, at ~200mm 4.5 maybe isn't very much?

What do you think I should do? Is there a fix for this, some sort of calibration? Im thinking I should do this test with the kit-lens as well, as it might be the camera that's faulty.
Im grateful for any thoughts and suggestions.

(Im from Sweden so forgive any errors in text).

Jul 18, 2005 at 06:59 AM

freelancer
Offline
Upload & Sell: Off

p.4 #4 · Canon autofocus information

neyoliv wrote :http://cweb.canon.jp/camera/eosd/20d/catalog/index06.html
but couldn't decipher what it means.

Google translation from Japanese to English
http://translate.google.com/translate?u=http%3A%2F%2Fcweb.canon.jp%2Fcamera%2Feosd%2F20d%2Fcatalog%2Findex06.html&langpair=ja%7Cen&hl=en&ie=UTF-8&oe=UTF-8&prev=%2Flanguage_tools

Jul 18, 2005 at 09:18 AM

RDKirk
Offline
Dedicated FM
Upload & Sell: Off

p.4 #5 · Canon autofocus information

Aha! That image from Japan fills in a puzzle piece. I'll have to fool with the babelfish translation later, but that image gives an important datum.

Notice the color coding of the sensors and the legend.

At f5.6 (blue colored on the diagram), we see the two linear pixel sensors of each array (a pair of sensors make up an array to sense the phase difference for a line of contrast). So we see two sensors arranged horizontally (the array that senses vertical lines) and two sensors arranged vertically (the array that senses horizontal lines).

At f2.8 (both green and red on the diagram) we see the same two vertically aligned sensors active, but a different set of horiztonally aligned sensors have been activated. These are farther apart, making the baseline wider, I'd suspect to provide the greater precision Canon claims with f2.8 lenses.

This is different from the way I understood it, but it does correspond to Rob Galbraith's description in his review of the 20D.

The photgraph of the sensor itself creates a lot more questions, though. I'm wondering about which sensors are which (I sure picked a lousy time to give up learning Japanese).

This is my guess at the sensors that aren't marked in the drawing:

Jul 18, 2005 at 12:07 PM

gsnethen
Offline
Upload & Sell: Off

p.4 #6 · Canon autofocus information

RDKirk wrote:
This is my guess at the sensors that aren't marked in the drawing:

I think I may have better guesses at the sensor assignments.

Based on the center sensors, it looks like the sensor pairs tend to be aligned vertically (for vertical sensors) or horizontally (for horizontal sensors).

Also, if the spacing of the sensors affects their accuracy, you would expect that each sensor pair should have equal spacing.

Therefore, I think the sensors on the left side are as follows:

10
.......9
8

10
.......9
8

(where 8 = 8 o'clock position AF point, 9 = 9 o'clock position AF point, 10 = 10 o'clock position AF point)

And mirror image for the right.

This would keep the vertical spacing of the vertical sensors constant and would require horizontal deflection only for the diagonol AF points (i.e., the 9 and 3 o'clock positions wouldn't require horizontal deflection.)

---Gary Snethen

Jul 19, 2005 at 01:47 AM

neyoliv
Offline
Upload & Sell: Off

p.4 #7 · Canon autofocus information

I've made a tentative translation of the legends in the image, comparing the kanji on the image with the ones on the page and translating with Google. It goes as follows:
Uppermost, left and right (they are the same): "diagonal range point"
Second to the left: "left range point"
Second to the right: "right tange point"
Top of the box: "central range point"
First line in the box: "F2.8 corresponding central horizontal sensor"
Second line in box: "F5.6 corresponding central vertical sensor (2 line ranges/ ?? arrangement" (?? for 2 unidentified kanji)
Third line in box: "F5.6 corresponding central horizontal sensor"

Also, in the Google translation of the page there's this line:
"Furthermore, precision is raised with the diagonal 4 range point and the left and right range point which expanded former compared to base length."
It could be reordered to this:
"Furthermore, precision is raised with the 4 diagonal range points and the left and right range points which base length was expanded compared to former."
I think "former" refers to the 10D.

Ney Oliveira

Jul 19, 2005 at 05:02 AM

neyoliv
Offline
Upload & Sell: Off

p.4 #8 · Canon autofocus information

Sorry, not "tange". It's "range".

Ney Oliveira

Jul 19, 2005 at 05:04 AM

gsnethen
Offline
Upload & Sell: Off

p.4 #9 · Canon autofocus information

neyoliv wrote:
I've made a tentative translation of the legends in the image, comparing the kanji on the image with the ones on the page and translating with Google. It goes as follows:
Uppermost, left and right (they are the same): "diagonal range point"
Second to the left: "left range point"
Second to the right: "right tange point"

Hmm... that seems strange.

If I understand your translation, the diagram claims that the four interior sensors map to the four diagonol AF points and the eight exterior sensors map to the two left/right AF points.

Is that what the diagram says? If so, the diagram doesn't make sense to me. We know the AF points require two sensors each, so four sensors is too few to represent the four diagonol AF points. And eight is twice as many as are needed to represent the two exterior AF points.

The layout I proposed makes more sense to me, because I can imagine how it could actually be implemented with a minimum number of light paths.

Has anyone seen a higher resolution version of the sensor chip itself? Or more importantly, has anyone seen the actual lens component that separates the image into multiple parts for projection onto the sensor chip?

---Gary Snethen

Jul 19, 2005 at 06:06 AM

RDKirk
Offline
Dedicated FM
Upload & Sell: Off

p.4 #10 · Canon autofocus information

gsnethen wrote:
RDKirk wrote:
This is my guess at the sensors that aren't marked in the drawing:

I think I may have better guesses at the sensor assignments.

Based on the center sensors, it looks like the sensor pairs tend to be aligned vertically (for vertical sensors) or horizontally (for horizontal sensors).

Also, if the spacing of the sensors affects their accuracy, you would expect that each sensor pair should have equal spacing.

I think you are completely correct.

It kind of sux that we have to apply forensic techniques to figure out how our cameras work. Canon ought to do some white papers on this stuff.

Edited by RDKirk on Jul 20, 2005 at 04:15 AM GMT

Jul 19, 2005 at 05:09 PM

RDKirk
Offline
Dedicated FM
Upload & Sell: Off

p.4 #11 · Canon autofocus information

TheJaff wrote:
That Sigma said that Canons AF stop at "the first part of focus" does not correspond with what Ive seen:

But it is as if the AF-motor didn't stop fast enough and focus ran past the correct point. .

I believe that is what happens. The same thing happens with Tamron and Tokina lenses when focusing on difficult subjects (dim light or low contrast) when the lens must move a long distance. It seems that it doesn't get the "stop" command before it's gone too far. But under the same conditions, if the lens starts from a point close to the actual focus point, it stops properly at the correct focus.

My theory is that each 3rd party company uses one common proprietary language in the drive cpus for all their lenses, but they add a "translator" chip that is dedicated to that particular camera mount. So all of Sigma's lenses have drive chips that speak "Sigma" but will have a translator chip that translates from Sigma to Canon or Nikon.

Compared to the camera manufacturer's lenses, this need for translation would create just a bit more delay in following any instruction from the camera. That's my theory, anyway. I never fall in love with my own theories, so I'll ditch it the moment a better looking theory comes along.

Jul 19, 2005 at 05:17 PM

Stoffer
Offline
Upload & Sell: Off

p.4 #12 · Canon autofocus information

Sigh.

I have been trying to get a reliable focus lock with my 20D and 135 f/2L at f/2 using those four new (over 10D) focus point near the rule of thirds. But focus is all over the place.

Jul 20, 2005 at 07:47 AM

gsnethen
Offline
Upload & Sell: Off

p.4 #13 · Canon autofocus information

Stoffer wrote:
Sigh.

I have been trying to get a reliable focus lock with my 20D and 135 f/2L at f/2 using those four new (over 10D) focus point near the rule of thirds. But focus is all over the place.

Right now I'm back at using the center point AF and doing a recompose with much more succes. This is of course still prone to error since the plane of in-focus will shift a bit with the recompose, but it is more reliable in order to achieve the right focus.

I guess that I just doesn't get along too well with those four "new" focus points. They also have trouble getting focus lock with my 50 f/1.4 in low light and/or constrast. These AF points are the only "issue" I have with my 20D where I could see why a Series 1 camera would give me a benefit.

The four diagonal points detect horizontal lines. Have you tried targeting horizontal lines (object boundaries, etc.) with them when you have trouble focusing with those points?

While I've noticed that they don't work as well as the center AF point (which is a higher resolution vertical and horizontal AF point), I haven't noticed them behaving any differently from the left/right (3 o'clock and 9 o'clock) AF points on my 20D.

What I'm about to say is wild speculation, so please accept it as such. If you are only having problems with the diagonal AF points, and my theory about these points being the only ones that are deflected horizontally is true, it could be that the element that performs this horizontal deflection is misaligned in your particular camera. Have you tried your lenses on a different 20D to see if you can replicate the problem?

---Gary Snethen

Jul 20, 2005 at 08:33 AM

Stoffer
Offline
Upload & Sell: Off

p.4 #14 · Canon autofocus information

Hi Gary...

gsnethen wrote:
The four diagonal points detect horizontal lines. Have you tried targeting horizontal lines (object boundaries, etc.) with them when you have trouble focusing with those points?

Well, I have tried to look for any contrast to get AF from (like the eyes of a face) but I cannot predict where it locks focus. I might get a higher succes rate from going after only horizontal lines, but when you need focus on the eyes, you cannot start to hunt elsewhere for contrast. Besides the center point AF does the job.

gsnethen wrote:
While I've noticed that they don't work as well as the center AF point (which is a higher resolution vertical and horizontal AF point), I haven't noticed them behaving any differently from the left/right (3 o'clock and 9 o'clock) AF points on my 20D.

That is probably true, but the thing about those four diagonal is that they lay fairly close to the rule of thirds intersections so they would be very useful if only...

gsnethen wrote:
What I'm about to say is wild speculation, so please accept it as such. If you are only having problems with the diagonal AF points, and my theory about these points being the only ones that are deflected horizontally is true, it could be that the element that performs this horizontal deflection is misaligned in your particular camera. Have you tried your lenses on a different 20D to see if you can replicate the problem?

---Gary Snethen

Maybe, but I haven't got easy access to another 20D. First I thought that my new 135 f/2L had an issue with slight back-focus but under the right circumstance the focus is spot on (using the center AF point of course).

Jul 20, 2005 at 09:49 AM

TheJaff
Offline
Upload & Sell: Off

p.4 #15 · Canon autofocus information

Does anyone know if this High Precision Mode AF is incorporated in the 350D?

Jul 20, 2005 at 01:03 PM

HeyLeeMR2
Offline
Account Locked

p.4 #16 · Canon autofocus information

Wow. Amazing post. That's all I can say.

Jul 20, 2005 at 05:10 PM

freelancer
Offline
Upload & Sell: Off

p.4 #17 · Canon autofocus information

TheJaff wrote:
Does anyone know if this High Precision Mode AF is incorporated in the 350D?
No. The 350D has the same AF design as the 300D/10D.

Jul 21, 2005 at 08:20 AM

TheJaff
Offline
Upload & Sell: Off

p.4 #18 · Canon autofocus information

freelancer wrote:
TheJaff wrote:
Does anyone know if this High Precision Mode AF is incorporated in the 350D?
No. The 350D has the same AF design as the 300D/10D.
Any reference?

Jul 21, 2005 at 10:47 AM

RDKirk
Offline
Dedicated FM
Upload & Sell: Off

p.4 #19 · Canon autofocus information

TheJaff wrote:
freelancer wrote:
TheJaff wrote:
Does anyone know if this High Precision Mode AF is incorporated in the 350D?
No. The 350D has the same AF design as the 300D/10D.
Any reference?

There is only the fact that Canon doesn't claim a high precision mode for it. But it's not exactly the same system as in the 10D or 300D.

"The EOS Digital Rebel XT has a newly developed high-precision 7-point AF system for speedy and accurate focusing in any situation. You can easily choose your own point of focus or direct the camera to choose it automatically. You can even use predictive AF to ensure that moving subjects stay focused while you compose your shot. With a built-in dioptric adjustment mechanism and a new precision matte screen, achieving sharp focus is fast and easy with the Digital Rebel XT, no matter where or when you're taking pictures. "

http://consumer.usa.canon.com/ir/controller?act=ModelFeaturesAct&fcategoryid=139&modelid=11154&pageno=3

Jul 21, 2005 at 05:43 PM

freelancer
Offline
Upload & Sell: Off

p.4 #20 · Canon autofocus information

RDKirk wrote:But it's not exactly the same system as in the 10D or 300D.
We know that both the 300D and 10D share a number of components, for example the sensor, metering and AF systems.

It could be more than a coincidence that the placement of all 7 AF points on both the 350D and 300D are identical and they also happen to be all normal precision AF. I would be surprised that Canon were to develop a completely new AF system for the 350D. Correct me please but I am under the impression that the AF system (not counting AF algorithm tweaks) are shared across cameras as happened in the past

The 20D is a different camera altogether with the new 9 point AF and the high precision center AF point.

Jul 22, 2005 at 08:46 AM

RDKirk
Offline
Dedicated FM
Upload & Sell: Off

p.4 #21 · Canon autofocus information

freelancer wrote:
RDKirk wrote:But it's not exactly the same system as in the 10D or 300D.
We know that both the 300D and 10D share a number of components, for example the sensor, metering and AF systems.

It could be more than a coincidence that the placement of all 7 AF points on both the 350D and 300D are identical and they also happen to be all normal precision AF. I would be surprised that Canon were to develop a completely new AF system for the 350D. Correct me please but I am under the impression that the AF system (not counting AF algorithm tweaks) are shared across cameras as happened in the past

The 20D is a different camera altogether with the new 9 point AF and the high precision center AF point.

Well, Canon says the 350XT focusing system is a new system: "The EOS Digital Rebel XT has a newly developed high-precision 7-point AF system for speedy and accurate focusing in any situation."

Canon both shares systems and develops new systems for cameras. For instance, the sensor of the 350XT is new--they didn't even use the 20D sensor. The mirror escapement and viewfinder systems on the 350XT are also new--they didn't use those systems from either the 300D or the 20D.

To some degree, these design changes appear to have been for no reason other than having developed a cheaper assembly process (that was definitely a factor in a lot of the changes from the 10D to the 20D).

Jul 22, 2005 at 12:40 PM

kotya
Offline
Upload & Sell: Off

p.4 #22 · Canon autofocus information

jmaio wrote:
Just to put a point on this - its worth spending the extra money for f/2, f/1.8, f/1.4, and even f/1.2 max aperture lenses for use with the 20D - not just for low light situations, but even when I shoot at, say, f/8.

Is that right?

This might be a silly answer, but anyway. I always miss the statement about "faster lens for low light work". To me the aperture is the way to control DOF, not to keep the shutter speed faster. Yes, you may be able to hand hold the camera at f/1.4 but imagine this: your subject is not further then 3 meters and... You will get a very shallow DOF, literally, you may get an eyebrow in focus and the eye ball OOF. You may have taken a picture without flash, there is something sharp in it, but is everything you need to be sharp is actually sharp?

Of course, there is real low-light work and I saw some really good photos, but why faster lenses always refered as "low light lens" is beyond me.

Jul 26, 2005 at 05:56 PM

RDKirk
Offline
Dedicated FM
Upload & Sell: Off

p.4 #23 · Canon autofocus information

kotya wrote:
jmaio wrote:
Just to put a point on this - its worth spending the extra money for f/2, f/1.8, f/1.4, and even f/1.2 max aperture lenses for use with the 20D - not just for low light situations, but even when I shoot at, say, f/8.

Is that right?

This might be a silly answer, but anyway. I always miss the statement about "faster lens for low light work". To me the aperture is the way to control DOF, not to keep the shutter speed faster. Yes, you may be able to hand hold the camera at f/1.4 but imagine this: your subject is not further then 3 meters and... You will get a very shallow DOF, literally, you may get an eyebrow in focus and the eye ball OOF. You may have taken a picture without flash, there is something sharp in it, but is everything you need to be sharp is actually sharp?

Of course, there is real low-light work and I saw some really good photos, but why faster lenses always refered as "low light lens" is beyond me.

In my experience, f2.8 is about the balance point between speed and sufficient depth of field for lenses of 135 and below for available light work. I find there is a really big practical threshold between f4--which becomes extremely limiting--and f2.8, which is quite workable.

At longer lengths, it took some serious practice to follow focus with f2.8 lenses above 135mm, but it was do-able, even up to the f2.8 300mm cannon (notice "cannon" not "Canon"). But AF made that a piece of pie.

I would agree that faster lenses are more in the realm of DoF control, but still, when I'm at ISO 3200 and the shutter speeds are around 1/15 at f2.8, I find that there is another threshold in practice between 1/15 and 1/30. A shutter speed of 1/30 can capture a lot of human movement that's lost at 1/15.

Jul 26, 2005 at 11:58 PM

ISO1600
Offline
Dedicated FM
Upload & Sell: On

p.4 #24 · Canon autofocus information

However, perusing the complaints at DPReview, a possible problem has occured to me. A number of people are complaining that their 20D cameras focus perfectly well with lenses that have a maximum aperture of 2.8, but poorly with faster lenses.

i have encountered this with my 20D....

at concerts, in horrible lighting, i could sometimes achieve focus more accurately with my 70-200 IS than with my 50 1.4....

that never made sense to me, until now. more reason to be excited about my 1V HS!

Jul 31, 2005 at 01:00 PM

mcjeong
Offline
Upload & Sell: Off

p.4 #25 · Canon autofocus information

This is really great and thanks to all!

I recently purchased a 20D with the following lens Tamron 28-75 2.8, Canon 17-40L, Canon 50 1.8, Canon 70-200L 2.8, and Canon 10-22.

Took them to Acapulco where it's around 100F and 50% humidity. Got off the car took some great pictures! But my problem came around the second day. None of my pictures were in focus. Funny thing was that all pictures taken vertically came reasonably in focus, but the horizontal ones were most (8 out of 10) out of focus.

Started thinking why? Only three things happened between my first day (great photos) and the second day:

1. 20D with latest firmware froze and had to take out the battery
2. My hotel room was confi at 75F and less than 10% humidty, so when I took the lens out of the room in the morning I saw some condensation on the lens (don't see how this can affect horizontal picture focus)
3. Did lots of lens changing to test all my new toys (all lens changes were done quickly in controlled environment to leave out the elements - but never turned off the camera to do so)

I need some help cause my current investment only takes vertical pictures!

Aug 04, 2005 at 05:35 AM


FM Forums \| Canon-mount SLRs \| Join Upload & Sell 1 2 3 4 5 6 7 end