7D AF Sensitivity Setting with 500mm plus 1.xTC for BIF

[David Stephens]

I've had great success with my 7D/500mm combination, with and without the 1.4x TC. I've got the AI servo tracking sensitivity set to Slow, per Artie's guide and this has worked well for static birds and BIF in all situation, but when I add the 1.4x TC I have trouble with birds circling fairly near me.

For example, a couple of weeks ago, two eagles that had been sitting on a log took off and started to circle relatively low. I felt like the AF was functioning well and the shots looked good in the viewfinder, but when I got home, almost all were slightly out of focus. The shots I wanted were with the eagles banking toward me with the light on their backs and wings fully extended.

My static shots from this outing were all sharp and a pelican flying by parrallel to the shore ended up looking sharp. I haven't had any problems with the TC off, using the bare 500/f4 IS.

I'm thinking that the 1.4TC was slowing down the servo enough that it wasn't keeping up with the eagles flying at me, even though they were thirty to sixty yards away. I'm thinking that the solution, other than taking the TC off, is to speed up the AI servo speed when I add the TC.

Over 40 bald eagles have been reported at a nearby lake and I plan to spend the next two mornings seeing what I can capture.

[John Guastella]

Dave, it's my understanding that the 7D AF sensitivity setting you refer to has nothing to do with speed of focus acquistion, but rather with how quickly the camera will switch focus from subject to background (or vice versa) if the focus point moves off of the desired subject.

Having said that, I recently acquired the 500 and have been using it almost exclusively with the 1.4X extender - with my 7D. With the sensitivity setting on "slow" I have experienced the same behavior you describe, with great results on static birds, but so-so results with BIF. I've tried setting the sensitivity to a faster value, but that didn't help. I have a feeling that the slowing of the AF system when an extender is attached is not something that can be cured by a camera setting, which is why many people are so excited about the claimed AF speed improvements of the newly-released extenders when combined with the soon-to-be-released new superteles.

John

[David Stephens]

Dave, it's my understanding that the 7D AF sensitivity setting you refer to has nothing to do with speed of focus acquistion, but rather with how quickly the camera will switch focus from subject to background (or vice versa) if the focus point moves off of the desired subject.

Having said that, I recently acquired the 500 and have been using it almost exclusively with the 1.4X extender - with my 7D. With the sensitivity setting on "slow" I have experienced the same behavior you describe, with great results on static birds, but so-so results with BIF. I've tried setting the sensitivity to a faster value, but that didn't help. I have a feeling that the slowing of the AF system when an extender is attached is not something that can be cured by a camera setting, which is why many people are so excited about the claimed AF speed improvements of the newly-released extenders when combined with the soon-to-be-released new superteles.

John, do you have this problem when the TC is not mounted? I don't. Those images are generally very sharp. When I need the reach of the TC, then I start having problems.

BTW, I usually use the expanded single-point for BIF and was doing that for these eagles. The background was blue sky, so I could have used 19-point, like Artie, but DPP shows AF points on the birds, so the problem wasn't that I was missing the birds. My next outing, if the BG is blue sky, then I'll try Arties advice about using the 19-point.

My lens is the Series I and my TC is Series II, so I'm doubting that the Series III TC will help me with servo speed, but I'll jump all over it if it will.

[William Malacarne]

Have you done a Micro Adjustment on the lens and then one with the lens and TC. It may help.

Bill

[David Stephens]

Have you done a Micro Adjustment on the lens and then one with the lens and TC. It may help.

The lens and the lens plus TC are sharp in static situations.

Dave

[Doug Brown]

It's unlikely that your problem is going to be solved by micro adjustment. More than likely it's a tracking issue or a shutter speed issue. You haven't given us enough information to diagnose the problem. Post a sample frame with EXIF data along with all relevant AF settings.

In addition to Artie's guide, you might want to have a look at my 7D setup guide that can be found here.

[Joerg Rockenberger]

Due to a failed internet connection I just lost a lengthy post about an exercise I did earlier this year re the different AF tracking sensitivity settings. Here is the scoop: I focused on a plank in a boardwalk (1DMkIV, 500F4IS, AI servo, central focus point etc.) and while continuing to focus I pointed the lens at a plank just in front of it. Then I watched the time it took for the system to refocus. I repeated this for the different tracking sensitivities. Turns out that at -2 it actually takes quite a moment. At first I thought this is bad but then I realized that in tracking BIF such time delay is beneficial if your AF point tends to wander off the subject. I didn't test the 500/1.4TC combo but it may be worthwhile to try to find out what the root cause of your observation is. JR

[Grant Atkinson]

Hi David, I have little experience with the 500f4 and tc combination, but lots of frames with my 7D and 300f2.8 has shown me that I almost always got my best results on moving subjects with that camera using extra fast shutter speeds. I have read several references to the fact that the small pixel pitch is more demanding of high speed to freeze movement. Perhaps the extra magnifaction from the TC is making this phenomenon more pronounced. I dont know how much light or extra speed/iso leeway you have
Just my thoughts
cheers
Grant

[David Stephens]

_MG_3910_DxO5Dvs3 by dcstep, on Flickr

Camera	Canon EOS 7D
Exposure	0.001 sec (1/1250)
Aperture	f/7.1
Focal Length	700 mm
ISO Speed	200
Exposure Bias	0 EV

I wanted to post a Screen Print here of Digital Photo Professional showing the AF point on the bird, but this forum program wouldn't allow me to Paste it from a document and I couldn't figure out how to reformat it and save it as a Image file. You'll have to trust me that a single AF is turned on on the eagles head and neck.

The AI servo is set at Slow.

This is the very best image, but the others are similar, with shutter speed between 1/1200 and 1/2000-second. I was using Aperture Preferred mode.

Here's the image at 100%

http://www.flickr.com/photos/dcstep/...n/photostream/

[David Stephens]

Grant, what do you mean by "extra fast shutter speeds?"

[David Stephens]

Joerg, thanks for sharing your efforts. (I hate when a great, long post gets lost to the mists).

I really you've hit on the root of my problem. I'm going to try a variety of servo speeds during my next outing. If I figure out how to adjust for the TC, then I might set up a Custom mode to quickly switch as I put on the TC. What have you done?

[Doug Brown]

I recommend using focus point expansion around your central AF point (the diamond-shaped AF pattern. Bump your ISO to 400 and choose a faster shutter speed. Don't worry about the other stuff for now. Give us a field report and show us a sample frame with the new settings. Good luck!

[John Guastella]

John, do you have this problem when the TC is not mounted?

My BIF keeper rate with the bare 500 is much higher than with 1.4X extender mounted.

Re: Joerg's post on trying the slowest sensitivity setting: I've tried all of the sensitivity settings (from slowest to highest) when using the 500 + 1.4X, and didn't observe an increase in my BIF keeper rate. I'm not saying you shouldn't try this, but I don't think it will markedly improve your results. I think the problem is the loss of light when the extender is attached: the AF system just doesn't have enough light to function quickly and accurately.

John

[William Malacarne]

There is another post on the forum about the new TC III's....According to Cuck Westfall, the article states....

http://usa.canon.com/dlc/controller?...508&fromTips=1

"As with previous EF Extenders, usage of Series III EF Extenders lowers AF drive speed to improve AF performance. When Extender EF 1.4X III is used, AF drive speed is reduced by 50%. When Extender EF 2X III is used, AF drive speed is reduced by 75%. This may seem like a drawback, but in reality subject tracking performance remains quite high when Series III Extenders are used with IS II lenses. This is due to improvements in AF precision made possible by the new microcomputer in the extenders."

I made the red hi-light.....

and also as John stated below....less light.

Bill

[Grant Atkinson]

Hi David, as a rough guideline, when I am trying to capture action like BIF with the 7D I feel most confident when speeds are faster than 1/1600s..so at least double the old rule of s/speed equal to focal length. As Doug suggested, going to ISO400 even at 7.1 on the image you posted would have given you the 1/2500sec. Also if you note some of the shutter speeds on Art Morris blog or his images taken with that camera on BPN, you will see faster speeds than are typical..check out his in flight turkey vulture last week.
I agree with John above about the AF accuracy being negatively affected by adding the converter, however if you have good bright light, which it seems you do, then there are enough folk posting very sharp BIF images taken with similar gear on the forum to show it can be done..

Let us know how your next session turns out...
cheers
Grant

[David Stephens]

There is another post on the forum about the new TC III's....According to Cuck Westfall, the article states....

http://usa.canon.com/dlc/controller?...508&fromTips=1

"As with previous EF Extenders, usage of Series III EF Extenders lowers AF drive speed to improve AF performance. When Extender EF 1.4X III is used, AF drive speed is reduced by 50%. When Extender EF 2X III is used, AF drive speed is reduced by 75%. This may seem like a drawback, but in reality subject tracking performance remains quite high when Series III Extenders are used with IS II lenses. This is due to improvements in AF precision made possible by the new microcomputer in the extenders."

I made the red hi-light.....

and also as John stated below....less light.

Well duh, no offense, but this is no news. The 1.4x TC takes the maximum f-stop from f/4 to f/5.6.

My point is that combining a TC with the AI servo sensitivity set to Slow is a recipe for even slower response.

The field test is coming tomorrow. I'll try both higher shutter speeds and, separately, speeding up the AI servo when the TC is on. If I'm lucky to get close enough to the bald eagles, I'll even pull off the TC and shoot with the bare lens.

[David Stephens]

Hi David, as a rough guideline, when I am trying to capture action like BIF with the 7D I feel most confident when speeds are faster than 1/1600s..so at least double the old rule of s/speed equal to focal length. As Doug suggested, going to ISO400 even at 7.1 on the image you posted would have given you the 1/2500sec. Also if you note some of the shutter speeds on Art Morris blog or his images taken with that camera on BPN, you will see faster speeds than are typical..check out his in flight turkey vulture last week.
I agree with John above about the AF accuracy being negatively affected by adding the converter, however if you have good bright light, which it seems you do, then there are enough folk posting very sharp BIF images taken with similar gear on the forum to show it can be done..

Let us know how your next session turns out...
cheers
Grant

Thanks Grant. I was shooting aperture-preferred and my SSs varied from 1/1200th to 1/2500th and all the images were poor, in fact, this particular one was the very best out of maybe 60 or 100 shots. I should have mentioned this earlier when Doug suggested more SS.

The forecast for tomorrow a.m. is more brilliant Colorado sun, so, once the sun's fully up, I'll be shooting-shutter preferred at 1/2000-second to eliminate that variable. I'll adjust the ISO to shoot for f/8 to f/11 to give the lens/TC a good chance to focus well, even with a slight miss by the AF.

Oh, btw, I am using a heavy tripod, heavy ballhead and highly respected side-mounted gimbal. I won't mention brands so that Artie can feel free to join in on the camera setup question, if he feels like it.

With luck I'll post results tomorrow, but it could be as late as Sunday night, given the holiday.

[William Malacarne]

David

I know the light is less.....just like you do.....but the slow down is due to electronics to slow down the focus so it can keep up with the less amount of light. At least that is how most people are treating it.

Bill

[David Stephens]

David

I know the light is less.....just like you do.....but the slow down is due to electronics to slow down the focus so it can keep up with the less amount of light. At least that is how most people are treating it.

Right and we, apparently, have no idea how setting the AI servo to "Slow" interacts with that condition. Does it make it even slower or does it just maintain the same level of slowness no matter what the setting? Hopefully I can raise the speed and enjoy a positive impact on my AF response.

We'll see...

[William Malacarne]

In the fine Article that Doug Brown wrote and referenced to you he stated this about "Slow" in regards to "C.Fn III: Autofocus/Drive"

"Next we’ll cover ‘C.Fn III: Autofocus/Drive.’ Some of these settings are among the most important of all 7D settings when it comes to getting the best possible images. I have ‘C.Fn III 1 AI Servo tracking sensitivity’ set to ‘Slow.’ If I’m tracking a flying bird and it flies behind a tree, slow tracking sensitivity means the camera will not give up focus on the bird and grab focus on the tree. This setting makes the AF much less twitchy."

This means the slow has no bearing on the initial focus speed but it gives a delay when it comes off of focus so you have time to regain proper focus.

Bill

[David Stephens]

In the fine Article that Doug Brown wrote and referenced to you he stated this about "Slow" in regards to "C.Fn III: Autofocus/Drive"

"Next we’ll cover ‘C.Fn III: Autofocus/Drive.’ Some of these settings are among the most important of all 7D settings when it comes to getting the best possible images. I have ‘C.Fn III 1 AI Servo tracking sensitivity’ set to ‘Slow.’ If I’m tracking a flying bird and it flies behind a tree, slow tracking sensitivity means the camera will not give up focus on the bird and grab focus on the tree. This setting makes the AF much less twitchy."

This means the slow has no bearing on the initial focus speed but it gives a delay when it comes off of focus so you have time to regain proper focus.

Thanks Bill.

My problem is not with initial focus, but tracking and holding focus, particularly with birds moving toward me and away. I may need a little more "twitch" when the TC is attached, not less.

I suspect that if Doug knew the answer he would have specifically responded.

I seems to me, if the AF delays focusing on background, then it's going to delay its adjustment as a bird flies toward me. If the TC already slows AF and the AI servo sensitivity slows it further, then that may cause what I'm seeing with my rig.

I'm going over all my settings again right now to make certain that I haven't overlooked something before my outing in the morning.

[Arthur Morris]

Interesting thread. I will try to give some overview comments here in spite of my lack of technical knowledge on the subject....

1-Though several folks have stated this clearly, it seems that some folks want the reality to be otherwise: With the AF tracking sensitivity set to Slow there is zero effect on either the speed of initial AF acquisition or on AI Servo AF tracking accuracy.

2-Because teleconverters rob the AF system of light I try to avoid using a TC with the 7D for flight (yet I will use the 1.4X with the MIV without reservation. Not sure why....)

3- And I do not believe that anyone has mentioned this: images created with the 1.4X TC in place are about 14% less sharp than images created with the prime lens alone... Major fact.

4- Lastly, the old shutter speed rule of thumb would suggest 1/1120 as minimum shutter speed based on the effective focal length. I do not purposely try for faster shutter speeds when using the 7D.....

5- And really lastly, doing comparative "tests" on birds in flight is pushing the limits of credibility; there are just too, too many variables. And photographing cars coming down the highway is not the same as photographing birds in flight...

[Roger Clark]

Interesting thread. I will try to give some overview comments here in spite of my lack of technical knowledge on the subject....

1-Though several folks have stated this clearly, it seems that some folks want the reality to be otherwise: With the AF tracking sensitivity set to Slow there is zero effect on either the speed of initial AF acquisition or on AI Servo AF tracking accuracy.

Artie,
As I understand it, the tracking speed is really misnamed. It should be something like
AF re-acquisition delay, and labeled short to long.

2-Because teleconverters rob the AF system of light I try to avoid using a TC with the 7D for flight (yet I will use the 1.4X with the MIV without reservation. Not sure why....)

The major factor is not light level but the light cone. As the light cone changes from f/2.8, to f/4, to f/5.6, the difference in the phase detect autofocus sensors becomes more difficult. After all you can AF in all kinds of light, from bright sun, to sunset light, even in moonlight and on stars. So light level is secondary to the light cone from the f/ratio.

The 1D series cameras have finer AF sensors, so can do a better job at slower f/ratios. Also, the AF circle of confusion is tighter on the 1D series. I read this in a canon white paper (probably from Chuck).

3- And I do not believe that anyone has mentioned this: images created with the 1.4X TC in place are about 14% less sharp than images created with the prime lens alone... Major fact.

There is also a loss in contrast due to diffraction starting to impact the image.

4- Lastly, the old shutter speed rule of thumb would suggest 1/1120 as minimum shutter speed based on the effective focal length. I do not purposely try for faster shutter speeds when using the 7D.....

Speed rule should be a function of the angular size of a pixel. For example, the 500 mm on a 7D has about half the angular size of a pixel than a Nikon D3. So shutter speed should be about double on a 7D over a D3 (or canon 1D Mark II). That illustrates the wall one approaches in trying to get more detail: smaller pixels collect less light and for action
are recording finer detail which requires faster shutter speed and that means even less light.

5- And really lastly, doing comparative "tests" on birds in flight is pushing the limits of credibility; there are just too, too many variables. And photographing cars coming down the highway is not the same as photographing birds in flight...

I agree. If the bird is flapping its wings, the velocity is changing with each pump of the wings. That makes it a challenge for any AF system.

I posted in another thread some AF speed tests I did on several cameras. The non 1D cameras were all about the same, but the 1D IV (I might have doen the 1DII also) was significantly faster with the same lens.

OK, found it:

I did a little experiment with tracking speeds. I took a canon 300 mm f/2.8
L IS lens with canon 2x TC. I manually moved the focus to minimum distance,
and then pointed at a bright distant source (the Moon in a dark sky), I half
pressed the focus and counted how long to come into focus. Here are some
results:

1D Mark II: 4 to 5 seconds
5D Mark II: 4 to 5 seconds
7D: 4 to 5 seconds
1D Mark IV: 2 to 2.5 seconds.

Then 300 f/2.8 with no TCs:

1D Mark IV: about 2/3 second or a little less.

Also, as expected, changing "tracking speed" in the custom functions did not
change the time to achieve initial focus.

Roger

[Arthur Morris]

Well duh, no offense, but this is no news. The 1.4x TC takes the maximum f-stop from f/4 to f/5.6.

No "Duh" necessary. Several folks have already assumed (incorrectly as you noted) that the "lowering of AF drive speed" was only caused by the new Series II TCs.

In fact, In another thread, I highlighted the exact same portion of the text in red that Bill did.

[Arthur Morris]

Thanks a stack Roger. Interesting test there at the end with the o-of moon.

It is gratifying when your technical explanations jibe perfectly with my gut reactions (in spite of my admitted lack of technical expertise (" in spite of my lack of technical knowledge on the subject....") And obviously my use of incorrect scientific terms (at times).

I did understand about half of what you said :) And I accept the rest also as fact.

I am gonna put up a new post: "TC or No TC for Better Images?" and I am sure that you will be the perfect person to comment accurately there.

Thanks again for your help and your expert technical explanations.

[Grant Atkinson]

Hi David, one important thing i forgot to add about my own use of very fast shutter speeds with the 7D, is I am almost always hand holding in a vehicle that usually has other folk in it creating movement.
cheers
Grant

[Roger Clark]

David,
Where are you photographing eagles in Colorado?
Roger

[Alan Lillich]

The major factor is not light level but the light cone. As the light cone changes from f/2.8, to f/4, to f/5.6, the difference in the phase detect autofocus sensors becomes more difficult. After all you can AF in all kinds of light, from bright sun, to sunset light, even in moonlight and on stars. So light level is secondary to the light cone from the f/ratio.

The best explanation of this I've seen related it to binocular vision. The further apart your eyes are, the better your depth perception. Same thing for phase detect AF sensors. They look at light beams coming from opposite sides of the lens and figure out if they cross in front of or behind the sensor plane. An f/2.8 lens has a wider opening than an f/5.6 lens, hence the beams are at wider angles, and the sensor's job is easier.

Alan

[David Stephens]

David,
Where are you photographing eagles in Colorado?
Roger

Yes indeed, Rog. The image above is from Cherry Creek State Park. This morning I was at Barr Lake State Park above the airport and the prairie roads North of DIA.

I'm still analyzing my results from today, but I did raise the AI servo sensitivity to the max and did increase my keeper ration when using the 1.4x TC. Light was similar at both shoots, but I think that the birds were a little closer today, so it makes it hard to attribute my improvement to the faster AI servo speed.

I'll look at the images in more depth and see if I can conclude anything.

[Arthur Morris]

...I did raise the AI servo sensitivity to the max and did increase my keeper ratio when using the 1.4x TC. Light was similar at both shoots, but I think that the birds were a little closer today, so it makes it hard to attribute my improvement to the faster AI servo speed.

David, I do not mean to be argumentative but raising the sensitivity has no effect at all on either the speed of initial focusing acquisition or AI Servo tracking accuracy. No matter how many times you state your belief that setting the tracking sensitivity to High will improve the AI Servo speed it will never be true.

[Roger Clark]

Yes indeed, Rog. The image above is from Cherry Creek State Park. This morning I was at Barr Lake State Park above the airport and the prairie roads North of DIA.

I'm still analyzing my results from today, but I did raise the AI servo sensitivity to the max and did increase my keeper ration when using the 1.4x TC. Light was similar at both shoots, but I think that the birds were a little closer today, so it makes it hard to attribute my improvement to the faster AI servo speed.

I'll look at the images in more depth and see if I can conclude anything.

David,
Thanks for the info. I've never gone to Cherry Creek State Park, but it is nice to know that there are eagles there. I've always found that the eagles at Bar Lake are too distant. Have you tried the Union reservoir area?

While I agree with Artie that technically, the servo sensitivity does not change acquisition speed, it does change how fast the camera decides to change and lock onto a new subject. In my experience, with slow tracking, when I have lost the subject, it also takes longer to reacquire it. So it is a two edged sword. I often prefer fast mode for the AI servo. For little birds which move very fast, there may be a fraction of a second where one can image, so a fast decision by the camera to lock onto the subject helps me. For larger slower birds it doesn't seem to make as much difference, except at takeoff. If I'm pointing on a bird at rest, and the bird takes off straight at me, I want the camera to decide the focus has changed and to start tracking it. It seems that servo tracking on fast helps in this situation too. But if I can remember to lift my finger off the shutter button and then half press again, it doesn't matter what the servo setting is. I just need more experience doing that, but that too takes time and on a little fast bird, that can be a significant fraction of the time one has to image it. No easy solution, except experience. Artie and others here clearly have it. I don't get enough time away from work (retirement looks good).:)

Roger

[David Stephens]

David, I do not mean to be argumentative but raising the sensitivity has no effect at all on either the speed of initial focusing acquisition or AI Servo tracking accuracy. No matter how many times you state your belief that setting the tracking sensitivity to High will improve the AI Servo speed it will never be true.

I've never said that AI servo sensitivity has any impact on initial focus acquisition. This thread is about tracking.

My belief is that AI servo sensitivity has some impact on the predictive focus function of the AI/AF servo. Canon does a VERY poor job of explaining it. On their own site they use the same words used by many in this thread, but they demonstrate it by showing pictures of a boy riding a bicycle toward the camera, with nothing between the boy and the camera. Is that a mistake, or does AI servo sensitivity have some impact on tracking a subject moving toward the camera?

It's been mentioned here that a bird flying doesn't really fly at a constant speed and that its speed varies as it pumps its wings. Perhaps, when the depth change is different than predicted, the fast AI servo speed will adjust more quickly to correct for the error. I don't know, but that's why I tried some field tests.

By the way Artie, I had great success with the 19-point scheme. It was very, very good with blue sky and surprisingly good with distant backgrounds. With relatively close backgrounds I had to go to single-point.

Here's one from a series that really surprised me, using 19 point and AI servo set to fast:


Eagle excitement by dcstep, on Flickr

These eagles are 300-yards away and the tree line a good deal further, so that perhaps explains why the camera didn't latch onto the BG. As Doug said, the eagles are generally too far away at Barr Lake, where this was taken.

[David Stephens]

Doug, I haven't tried Union Reservoire, but it IS now on my list.

I too found the fast AI servo useful. I was shooting a hawk flying in back of power lines and I seemed to reacquire more quickly when I accidentally let it focus on the lines. It just seemed to me that the systems reactions were quicker in all respects.

What I think I've learned is that there may be situations where I want slow AI servo and others where I want fast AI servo. The closer the BG is, the slower the speed I want. With blue sky I want everything fast and all 19 points active. I may set up custom settings, but for now I'm just keeping the AI servo function cued up so that I can quickly go there and move it up and down. I'm switching a lot between 19 point and single point.

[David Stephens]

...



The major factor is not light level but the light cone. As the light cone changes from f/2.8, to f/4, to f/5.6, the difference in the phase detect autofocus sensors becomes more difficult. After all you can AF in all kinds of light, from bright sun, to sunset light, even in moonlight and on stars. So light level is secondary to the light cone from the f/ratio.

...

Speed rule should be a function of the angular size of a pixel. For example, the 500 mm on a 7D has about half the angular size of a pixel than a Nikon D3. So shutter speed should be about double on a 7D over a D3 (or canon 1D Mark II). That illustrates the wall one approaches in trying to get more detail: smaller pixels collect less light and for action
are recording finer detail which requires faster shutter speed and that means even less light.

...

These two points were very helpful to me. I was wondering why my 500mm was plenty fast at f/5.6, but f/5.6 with the TC was clearly not the same thing.

On the second point, this explains some of my dissapointment with the 7D in less than stellar light.

[David Stephens]

Doug, I haven't tried Union Reservoire, but it IS now on my list.
...

ROGER, sorry about calling you Doug.

BTW, which part of Union do you like to shoot from in the a.m.?

[David Stephens]

Thanks to everyone that contributed.

I got out again yesterday and shot another couple of hundred shots, some in similar bright light and some at dusk at high ISO. After a total of around 700 shots, in field conditions, I'm keeping my AI servo sensitivity set at fast when I have the 1.4xTC on.

Per Artie's 7D guide, I'm using 19-point AF more often and switching to single-point or expanded single-point when the BG might be problematic. With blue sky or distant, contrasty background, I stay with 19-point, but with a close, contrasty BG I go to single point or expanded. Of course, in brush and with limbs and stuff in the FG I also go to single-point.

Without the TC I'll go to medium AI servo sensitivity and use the same criteria for AF point selection. Given my subjects this weekend, I didn't really do much testing w/o the TC. Since I had no problem in the past keep the sensitivity down, but not all the way. If testing shows that to be problematic, I'll let you know.

[Arthur Morris]

Here is a follow-up to THE question in the form of the answer from Rudy Winston (Canon Advisor, Technical Information, CIG Professional Products Marketing Division):

According to everything we've ever heard from the engineers at Canon Inc. in Japan, the FAST vs. SLOW AI Servo AF tracking settings on the EOS-1D, EOS-1Ds and now EOS 7D cameras ONLY influences how the system behaves if you are tracking some sort of moving subject, and then the active AF point (or points) lose sight of that subject momentarily. You thought the bird was going to zig, and it zagged instead, and now you're looking for a moment on a bunch of distant trees in the background as you struggle to get the lens and the active AF point back on that pesky bird. As I'm sure you're aware, the tracking settings dictate whether the system will try to immediately re-focus on whatever the active AF point(s) now see, or whether there will be a distinct pause before the camera tries to re-focus on whatever new subject is at the AF point in question.

Again, according to what we've been told, as well as my own personal experience (not really with birds, but I've shot enough sports in my day to have a sense of its operation), as long as you can keep the active AF point solidly and consistently on your moving subject, and it doesn't drift away momentarily as you're tracking it, your choice of tracking settings should have no difference in AF consistency or accuracy. I wish I had a nickel for every time I've told a shooter (including some pros) that if you're photographing Indy cars coming straight at you at the Indianapolis Speedway with a big tele lens, and you keep your AF point upon the car in question and that point sees adequate detail the whole time, you will NOT get better or "faster" AF tracking by selecting the FAST setting. All it will do is insure that if you lose the car for a moment, the camera will very quickly try to re-focus on whatever new subject it's now seeing with that AF point.

There are perfectly valid reasons for any pro who shoots challenging moving subjects to vary these settings, choosing the slow options for one type of subject (presumably, subjects that move enough that you feel you'll often lose the subject for a moment or two when tracking them), and switch to the fast options for other moving subjects, where you either know you deliberately want to instantly re-focus on new subjects in AI Servo AF, or perhaps if you think your subject's movement may be so erratic that you want an edge in getting the system to react to sudden changes in its direction and movement. I'm sure someone with all the experience you have in this mightily challenging field would have other situations and priorities where you tend to prefer one or the other. Absolutely no problem, as long as you don't go into it simply assuming that if you set it to "FAST", you'll ipso-facto get better AI Servo AF performance in general. As I said, as far as we know, you won't.

In any case, if you experiment with faster than the SLOW that I always use and it works for you, go with that.

[Arthur Morris]

One thing that I should have mentioned here as it is closely related to the main question here is the option of turning Focus Search On off. In our User's Guides I recommend doing that with the super-telephotos when the birds are all taking off or landing at a relatively consistent distance..... There are lots of other ramifications that are discussed in detail in the UGs.

[Alan Lillich]

A lot of what I think Dave is experiencing is the ambiguity of "slow" versus "fast" and the lack of proper explanations of camera behavior. In this and IS behavior it is really aggravating that the camera companies don't give us facts. What exactly are the delays in milliseconds? The bird is moving the whole time, so the focus is drifting the whole time the sensor is off of it. When the sensor comes back on the bird it won't magically still be in focus, the system has to somehow decide when it is back on the subject or when to give up. How is this decision made?

(Or is the focus drifting? Another lack of facts! Does the focus continue to move according to recent motion, just as it is supposed to do to account for mirror blackout?)

Suppose the slow delay is X milliseconds and the fast delay is X/4. If Dave can get the sensor back on the bird within the fast window then the camera will start to reacquire focus in no more than X/4 milliseconds at the fast setting. He has to wait an additional 3/4 X at the slow setting.

In other words, each individual and situation needs a delay that is long enough to not start focusing off the subject, but short enough to not let the subject get too far out of focus before reacquiring at the new subject location.

[Arthur Morris]

Hi Alan, I hear you. The manufacturer's out has always been the same when the question of AF tracking accuracy comes up (as seem in the following excerpt from Rudy Winston's note above): " (as long as ...) that point sees adequate detail the whole time..." In other words, the AF system always need sufficient contrast in order to see. If the sensor is on the bird's all white or all black breast all AF bets are off; there is no contrast so AF is searching.....

I am thinking that you are well aware of this but wanted to post it for the less technically oriented :)

In addition, there are so, so many variables when photographing birds in flight that crunching the numbers does not make sense to me. But I am not a scientist :)

[David Stephens]

Yeah, Alan, why no facts and specs?

Even Rudy Winston is vague because even he hasn't gotten straight talk from the engineers, evidently.

Adding the TC makes things even harder to interpret. We all know that initial AF acquisition is slower, but I have the feeling that tracking is slower and selecting the slow AI servo setting slows the responsiveness to irregular behaviour.

Oh well, for me, I've got a system of settings that has improved my keeper ratio. I'd just encourage anyone having trouble to experiment with various settings.

[Arthur Morris]

Even Rudy Winston is vague because even he hasn't gotten straight talk from the engineers, evidently.

David, I am not sure why you have trouble understanding these principles: with birds in flight, there cannot be any objective testing. There cannot be any consistent accurate results. And in addition, with moving subjects there is always the skill of the operator and operator error to take into account.

Yes, it would be great if we have science-fiction-like AF tracking accuracy: get the bird anywhere in the frame and the eye is razor sharp. The advances of the past two decades have been astounding: from manual focus where I would be lucky to get one sharp image on a roll of 36, to today's amazing AI Servo and Continuous AF where even beginners can create 50% sharp on the eye images.

No matter how open and frank the engineers in Japan might be, it would not change any of the above.

[David Stephens]

David, I am not sure why you have trouble understanding these principles: with birds in flight, there cannot be any objective testing. There cannot be any consistent accurate results. And in addition, with moving subjects there is always the skill of the operator and operator error to take into account.

Yes, it would be great if we have science-fiction-like AF tracking accuracy: get the bird anywhere in the frame and the eye is razor sharp. The advances of the past two decades have been astounding: from manual focus where I would be lucky to get one sharp image on a roll of 36, to today's amazing AI Servo and Continuous AF where even beginners can create 50% sharp on the eye images.

No matter how open and frank the engineers in Japan might be, it would not change any of the above.

Artie, I'm not sure why you're having trouble understanding. I never said that bird in flight would yield reliable testing results. I said what I concluded for myself and suggested that others try for themselves. I merely focused on trying differing AI servo speeds over several hundred shoots in varying light and with different subjects, concluding that I needed different settings for different conditions, just as you have suggested. The only difference from your suggested settings and what I'm doing is that when I add the TC and I'm shooting BIF, I speed up the AI servo. I feel that that helps me and I'm certain that it's causing no harm.

If the engineers in Japan would provide us quantitative data we might be able to validate my conclusion. It'd be nice to know exactly what happens with AI servo speed when you add a 1.4x TC on a 500mm lens. Does it stay exactly the same or, like the focus acquisition, does it also slow some.

Why can't they say exactly what "fast" and "slow" means? Now that we've been discussing it a week (thanks to all the contributors), I've reexamined the Canon video and descriptions and reread all the attempts to explain it in the camera guides, I'm thinking that it's really the frequency at which the AI servo compares the actual focus to the predicted focus and adjusts to the new target. "Slow" would mean x-times per second and "Fast" might mean 5x-times per second, for instance. To work at all, it has to compare actual results to its predicted results at some frequency.

I'm not expecting science fiction AF, at least not yet. I just want to understand how it works. I don't think that anyone will argue that Canon's done a poor job of explaining it.

[Danny J Brown]

David,
Thanks for the info. I've never gone to Cherry Creek State Park, but it is nice to know that there are eagles there. I've always found that the eagles at Bar Lake are too distant. Have you tried the Union reservoir area?

While I agree with Artie that technically, the servo sensitivity does not change acquisition speed, it does change how fast the camera decides to change and lock onto a new subject. In my experience, with slow tracking, when I have lost the subject, it also takes longer to reacquire it. So it is a two edged sword. I often prefer fast mode for the AI servo. For little birds which move very fast, there may be a fraction of a second where one can image, so a fast decision by the camera to lock onto the subject helps me. For larger slower birds it doesn't seem to make as much difference, except at takeoff. If I'm pointing on a bird at rest, and the bird takes off straight at me, I want the camera to decide the focus has changed and to start tracking it. It seems that servo tracking on fast helps in this situation too. But if I can remember to lift my finger off the shutter button and then half press again, it doesn't matter what the servo setting is. I just need more experience doing that, but that too takes time and on a little fast bird, that can be a significant fraction of the time one has to image it. No easy solution, except experience. Artie and others here clearly have it. I don't get enough time away from work (retirement looks good).:)

Roger

I agree with Roger. I've been photographing eagles, geese and ducks in flight for the past few weeks with my new 1D IV and when I track with the slow setting I feel like I disconnected six of the twelve spark plugs on my Ferarri and I start missing my 40D. On the fast setting, which I love, the camera might stray for a microsecond if I get a little behind but finds the bird back in the same microsecond. I'm like everybody else in that I don't really understand why certain things work for me; I just know they do. Artie is right. If it feels right for you...do it! Don't use settings just because they work for somebody else. Happy shooting!

DB

[David Stephens]

I agree with Roger. I've been photographing eagles, geese and ducks in flight for the past few weeks with my new 1D IV and when I track with the slow setting I feel like I disconnected six of the twelve spark plugs on my Ferarri and I start missing my 40D. On the fast setting, which I love, the camera might stray for a microsecond if I get a little behind but finds the bird back in the same microsecond. I'm like everybody else in that I don't really understand why certain things work for me; I just know they do. Artie is right. If it feels right for you...do it! Don't use settings just because they work for somebody else. Happy shooting!

Thanks Danny, your experience is consistant with mine with my 7D (maybe I've only got 8-cylinders to start with, but they're from a BMW M3 with an 8600 rpm redline) AND I was wondering about the MkIV. A friend has offered to let me try his MkIV for a comparison to my 7D in this regard. I really don't Jones for the larger body and bigger investment, BUT if the AF were leagues ahead for BIF, then I'd be very tempted. I suspect that the 7D, properly set up, is very close, at least with a 1.4x TC.

Danny, do you use the 2.0x TC with your MkIV? How does that work for BIF?

[Danny J Brown]

Thanks Danny, your experience is consistant with mine with my 7D (maybe I've only got 8-cylinders to start with, but they're from a BMW M3 with an 8600 rpm redline) AND I was wondering about the MkIV. A friend has offered to let me try his MkIV for a comparison to my 7D in this regard. I really don't Jones for the larger body and bigger investment, BUT if the AF were leagues ahead for BIF, then I'd be very tempted. I suspect that the 7D, properly set up, is very close, at least with a 1.4x TC.

Danny, do you use the 2.0x TC with your MkIV? How does that work for BIF?

Whoops, I forgot to say that all of my BIF work has been with the 1.4TC. I haven't even tried it without the converter because I'm always looking for more reach. Actually, I wouldn't mind having Artie's 800! Regarding weight, I need that extra to balance out the Wimberley so I'd be throwing a battery pack on the 7D just to get it. I think you are right that the AF systems are probably about the same on the two cameras. Some say they even like the 7D a little better or at least as much as the Mark IV. Take care.

DB

P.S. I don't own a 2.0 TC.

.

[Danny J Brown]

David,

One of my workmates is a full time wildlife photographer for the Missouri Department of Conservation and when I shot eagles with him a while back he never took the 2.0 TC off of his 600mm lens. He was using a Mark III at the time but uses a Mark IV now. Hope that helps.

DB

[David Stephens]

David,

One of my workmates is a full time wildlife photographer for the Missouri Department of Conservation and when I shot eagles with him a while back he never took the 2.0 TC off of his 600mm lens. He was using a Mark III at the time but uses a Mark IV now. Hope that helps.

As the Tim the toolman would say, "Mmmm, more power." :cool:

[David Stephens]

Whoops, I forgot to say that all of my BIF work has been with the 1.4TC. I haven't even tried it without the converter because I'm always looking for more reach. Actually, I wouldn't mind having Artie's 800! Regarding weight, I need that extra to balance out the Wimberley so I'd be throwing a battery pack on the 7D just to get it. I think you are right that the AF systems are probably about the same on the two cameras. Some say they even like the 7D a little better or at least as much as the Mark IV. Take care.

.

So you own both the 7D and MkIV? When does a person that owns the MkIV use his 7D?

BTW, my 7D balances perfectly on my Wimberley Sidekick, with and without the TC and with no battery pack. (I chimp very little and don't use live view a lot, so I'm easy on my batteries.)

[Danny J Brown]

So you own both the 7D and MkIV? When does a person that owns the MkIV use his 7D?

BTW, my 7D balances perfectly on my Wimberley Sidekick, with and without the TC and with no battery pack. (I chimp very little and don't use live view a lot, so I'm easy on my batteries.)

Sorry....I wasn't clear. I don't own a 7D. I meant that if I did, I'd probably end up putting a battery pack on it anyways, which would make it as bulky as the 1D. Sounds like your's is doing fine for your set-up though. Having said all that I sure love the feel of the 1D. I also loved the way my 40D handled and felt with the battery pack attached.

DB