1D IV Tracking Experiment

It would be helpful if you posted all your C. Fn III AF settings.
--
Blake in Vancouver
Panasonic Stuff, Canon Stuff. Mac Stuff & annoying PC & Windows stuff.
 
Remember the AF tracker is supposely using range data to also help decide what AF point should be active. The range is not changing but the tracker must make some assumptions regarding this. I am not sure why the bright bkgd should cause the tracker to fail. But something is happening. I might try the test again in daytime so the lights can be off.--
A bird in the viewfinder is worth...
 
I couldn't remember how the AF points work off the top of my head.

I actually do own a MkIV, and just got back from a tracking test of my own. Shot 45 frames and kept 25 of them. Pretty good average considering how long it's been since I last used the camera.

My tracking charts look like this, no strings attached:


Looks correct. I'm pretty sure that you can't get all 45 points to work once you select one. Only a user specified number of AF points around the selected point will work.
The 1D Mark IV has a "45 point expansion" mode which really will turn all 45 points on and will show in real time which point the AF system has switched to.
 
Remember the AF tracker is supposely using range data to also help decide what AF point should be active. The range is not changing but the tracker must make some assumptions regarding this. I am not sure why the bright bkgd should cause the tracker to fail. But something is happening.
Another thing to remember is that the AF sensor has to adjust its sensitivity to the amount of light it's receiving. That makes it tough for them to track dim objects in front of bright backgrounds. If some bright light causes the sensitivity to drop, your chart will effectively disappear to the AF sensor and it will start looking for something else.

This was a common problem with the 1D Mark III AF. When a subject would move out of shadow into bright sunlight (very common in football) often the AF would completely fail to track the subject. They've improved this with the Mark IV but there's no way to solve this completely. It's just another case where you need to anticipate the AF failing and prepare to restart it when it happens.
 
Just to followup. I had fully expected the 1DIV to successfully AF in the simple panning test based on Canons literature. I am surprised. But more surprised that others on this forum are not also surprised. i would suggest you try this simple test outdoors on an isolated object, mailbox, birdfeeder, etc and see what results you get.
--
A bird in the viewfinder is worth...
 
Just to followup. I had fully expected the 1DIV to successfully AF in the simple panning test based on Canons literature. I am surprised. But more surprised that others on this forum are not also surprised. i would suggest you try this simple test outdoors on an isolated object, mailbox, birdfeeder, etc and see what results you get.
--
A bird in the viewfinder is worth...
OK I did more research since noone seems to be able to account for the AF problem(?). Herte is a quote from Canon:

When using C.Fn III-8 AF Expansion with option 3: ‘Use all 45 points’ selected, during AI Servo AF, (if the center AF point is manually selected), 6 AF points in the rows above and below, and 3 AF points to the right and left (total 18 focus points) are available. The AF Expansion area follows the subject wherever it moves within the 45-point AF coverage area. All 45 AF points operate in this instance.

So it seems what I was trying to do is only possible if the manually selected AF point is the center AF point. I will try the test again with center point and pan right and left acros the FOV.
--
A bird in the viewfinder is worth...
 
Just to followup. I had fully expected the 1DIV to successfully AF in the simple panning test based on Canons literature. I am surprised. But more surprised that others on this forum are not also surprised. i would suggest you try this simple test outdoors on an isolated object, mailbox, birdfeeder, etc and see what results you get.
--
A bird in the viewfinder is worth...
OK I did more research since noone seems to be able to account for the AF problem(?). Herte is a quote from Canon:

When using C.Fn III-8 AF Expansion with option 3: ‘Use all 45 points’ selected, during AI Servo AF, (if the center AF point is manually selected), 6 AF points in the rows above and below, and 3 AF points to the right and left (total 18 focus points) are available. The AF Expansion area follows the subject wherever it moves within the 45-point AF coverage area. All 45 AF points operate in this instance.

So it seems what I was trying to do is only possible if the manually selected AF point is the center AF point. I will try the test again with center point and pan right and left acros the FOV.
--
A bird in the viewfinder is worth...
I tried this test on my patio with a hanging plant. Good light. I just held the shutter half closed and watched the active AF point box light up as I panned the camera. Not 100% successful but much better than before.

So my conclusions: Keep selected AF point on subject unless center point is selected and 45 point expansion is enabled.--
A bird in the viewfinder is worth...
 
Just to followup. I had fully expected the 1DIV to successfully AF in the simple panning test based on Canons literature. I am surprised. But more surprised that others on this forum are not also surprised. i would suggest you try this simple test outdoors on an isolated object, mailbox, birdfeeder, etc and see what results you get.
--
A bird in the viewfinder is worth...
OK I did more research since noone seems to be able to account for the AF problem(?). Herte is a quote from Canon:

When using C.Fn III-8 AF Expansion with option 3: ‘Use all 45 points’ selected, during AI Servo AF, (if the center AF point is manually selected), 6 AF points in the rows above and below, and 3 AF points to the right and left (total 18 focus points) are available. The AF Expansion area follows the subject wherever it moves within the 45-point AF coverage area. All 45 AF points operate in this instance.

So it seems what I was trying to do is only possible if the manually selected AF point is the center AF point. I will try the test again with center point and pan right and left acros the FOV.
--
A bird in the viewfinder is worth...
I tried this test on my patio with a hanging plant. Good light. I just held the shutter half closed and watched the active AF point box light up as I panned the camera. Not 100% successful but much better than before.

So my conclusions: Keep selected AF point on subject unless center point is selected and 45 point expansion is enabled.--
A bird in the viewfinder is worth...
Further update. Rereading the manual regarding the 45 point expansion option:
"All 45 points area
In the AI Servo AF mode, the autofocus will center on the manuallyselected
AF point with the row of six AF points right above and below the
AF point and the three AF points on the immediate left and right will also
become active (total of eighteen AF points). If the subject moves to
another AF point within the expanded AF points, the expanded AF will
center on that AF point. Since all 45 AF points will operate, as long as
the subject is within the Area AF frame, it will be focused at all times.
In the One-Shot AF mode, if focus cannot be achieved with the manuallyselected
AF point, focusing will be done with the remaining 44 AF points."

This would seem to indicate that the subject should be trackable even if the center AF point is not selected. So my original test should had been valid.
--
A bird in the viewfinder is worth...
 
I don't know if it is the same, but with the 1d3 AF servo auto AF point selection always starts tracking with the center focus point. (according to the manufacturer)
 
I'm not defending 1D MkIV. I could really care less if people like it or not. I would expect this test to fail though, especially with a wide angle lens. How often does anyone pan across stationary subjects in a real world scenario? Good for testing I guess, but it doesn't seem like a good photographic technique to me.

I would think a photographer would really be better off limiting the active AF points, and improving their panning technique in order to keep that small group of AF points on the (moving) subject, than relying on the camera to do it for them with all 45 points active. The camera is dumb as a rock without somebody behind the viewfinder to guide it down the right path.

Are you thinking about changing the way you use it, or are you going to return it now that the test has shown that it is not that reliable with all 45 points active?
I tried this test on my patio with a hanging plant. Good light. I just held the shutter half closed and watched the active AF point box light up as I panned the camera. Not 100% successful but much better than before.

So my conclusions: Keep selected AF point on subject unless center point is selected and 45 point expansion is enabled.--
Further update. Rereading the manual regarding the 45 point expansion option:
"All 45 points area
In the AI Servo AF mode, the autofocus will center on the manuallyselected
AF point with the row of six AF points right above and below the
AF point and the three AF points on the immediate left and right will also
become active (total of eighteen AF points). If the subject moves to
another AF point within the expanded AF points, the expanded AF will
center on that AF point. Since all 45 AF points will operate, as long as
the subject is within the Area AF frame, it will be focused at all times.
In the One-Shot AF mode, if focus cannot be achieved with the manuallyselected
AF point, focusing will be done with the remaining 44 AF points."

This would seem to indicate that the subject should be trackable even if the center AF point is not selected. So my original test should had been valid.
 
I'm not defending 1D MkIV. I could really care less if people like it or not. I would expect this test to fail though, especially with a wide angle lens. How often does anyone pan across stationary subjects in a real world scenario? Good for testing I guess, but it doesn't seem like a good photographic technique to me.

I would think a photographer would really be better off limiting the active AF points, and improving their panning technique in order to keep that small group of AF points on the (moving) subject, than relying on the camera to do it for them with all 45 points active. The camera is dumb as a rock without somebody behind the viewfinder to guide it down the right path.

Are you thinking about changing the way you use it, or are you going to return it now that the test has shown that it is not that reliable with all 45 points active?
No No!! I am not complaining at all about the camera. Background. I shot a couple of basketball games, my first, and it was difficult to determine where to aim after track had started because the active AF point does not light up when tracking with six assist AF points, only initially. I had read that using 45 point expansion would allow the active AF point to light up. So my test was to determine if this was usable. Apparently only if the center AF point is manually selected.
I do not suggest that this is a useful photepraphic technique, but if an AF point is selected at an extreme edge of the FOV and the subject moves across the FOV it would seem that the camera should be able to track the subject without panning. My test was a stationary subject but the camera panning.

I have zero concerns about the camera. I am more than happy with it. Just trying to get more familar with how it works and the limitations.--
A bird in the viewfinder is worth...
 
I see glitches when using the 1D MkIV, but most of them only crop up when I have a TC on a lens. It hiccups on AIServo focus when panning. I just tap the rear focus button again while tracking and it wakes up most of the time.

I've never shot team sports. It would probably drive me to taking up knitting. ;-)
 
I don't know if it is the same, but with the 1d3 AF servo auto AF point selection always starts tracking with the center focus point. (according to the manufacturer)
clk_walker is not using APS mode. He (or she) is using the new 45 point expansion mode that's new with the Mark IV. You can select any point to start with and the viewfinder will show in real time when it has switched to another focus point.
 
I do not suggest that this is a useful photepraphic technique, but if an AF point is selected at an extreme edge of the FOV and the subject moves across the FOV it would seem that the camera should be able to track the subject without panning. My test was a stationary subject but the camera panning.
In some cases it can do that but as I've said in previous posts, there are many conditions that will make that not happen every time. Some people have unrealistic expectations of AI-Servo being able to track their subject around the viewfinder like a laser-guided missile. Why else would Canon call it AI -Servo which stands for "Artificial Intelligence? It must able to recognize your subject!

But it doesn't. It's little more than a sensor with 45 black and white pixels. It doesn't see anything but contrast and distance.
 
Herte is a quote from Canon:

When using C.Fn III-8 AF Expansion with option 3: ‘Use all 45 points’ selected, during AI Servo AF, (if the center AF point is manually selected), 6 AF points in the rows above and below, and 3 AF points to the right and left (total 18 focus points) are available. The AF Expansion area follows the subject wherever it moves within the 45-point AF coverage area. All 45 AF points operate in this instance.

So it seems what I was trying to do is only possible if the manually selected AF point is the center AF point.
No, that's just another case of Canon being unable to write clear English.

They're using the center point selection as an example. If you select another point you will still get up to 18 focus points active and fewer if you select a point close to the edge of the 45 point grid. I've done this many times.

Why didn't they call it 18 point expansion? Because 45 point expansion sounds better and is far more confusing.

The Mark III brought many new custom AF functions that were so poorly explained in the manual that Canon had to write a completely separate AF manual to explain them. Then with the Mark IV came out with the same functions, Canon did not change one single word in their descriptions in the manual to make them any clearer. Canon is just terrible at documentation.
 
I don't know if it is the same, but with the 1d3 AF servo auto AF point selection always starts tracking with the center focus point. (according to the manufacturer)
clk_walker is not using APS mode. He (or she) is using the new 45 point expansion mode that's new with the Mark IV. You can select any point to start with and the viewfinder will show in real time when it has switched to another focus point.
Thanks for the clarification and I hope I didn't add to the confusion. :D
 

Keyboard shortcuts

Back
Top