To what extent does AI servo depend on the lens ?

[Shabok]

Hi all,

I'm going to buy a 70-200 lens (2nd hand Canon 70-200 f2.8L, new Canon 70-200 f4L or new Sigma 70-200 f2.8 HSM, still undecided) in order to shoot mainly mountain biking. I read contradictory comments on photographyreview.com concerning the difference between Canon USM and Sigma HSM autofocusing, some were saying that AI servo performance using the aforementioned Sigma lens was noticeably inferior compared to what we get using one of those Canon lenses ; others said that AF is virtually as fast on USM and HSM lenses.

Since this point is very important in my decision process, could those that have tested both lenses (Sigma 70-200 2.8 HSM and Canon 70-200 f2.8 or f4 L) share their impressions ?

Thanks in advance,
Shabok

 

[jonlg]

Same question/same lens, as I like to shoot motocross.
On the weekend my 75-300 and D60 and me had only a few good ones.
What we were shooting was elementry, so it's got me buggered.
I figure the AI can't keep up, but hope this is not the case.
Does panning IS make a difference to the end results?
John

Hi all,

I'm going to buy a 70-200 lens (2nd hand Canon 70-200 f2.8L, new
Canon 70-200 f4L or new Sigma 70-200 f2.8 HSM, still undecided) in
order to shoot mainly mountain biking. I read contradictory
comments on photographyreview.com concerning the difference between
Canon USM and Sigma HSM autofocusing, some were saying that AI
servo performance using the aforementioned Sigma lens was
noticeably inferior compared to what we get using one of those
Canon lenses ; others said that AF is virtually as fast on USM and
HSM lenses.

Since this point is very important in my decision process, could
those that have tested both lenses (Sigma 70-200 2.8 HSM and Canon
70-200 f2.8 or f4 L) share their impressions ?

Thanks in advance,
Shabok

--
~

 

[Mike Long]

I have not compared them but I promise you that you will never be disappointed with the images from the 70-200L 2.8 non-IS. It is a wonderful lens for portraits, landscapes and whatever. If you have the $$$, buy the IS version. If not, buy the non-IS and don't look back.

Mike

Hi all,

I'm going to buy a 70-200 lens (2nd hand Canon 70-200 f2.8L, new
Canon 70-200 f4L or new Sigma 70-200 f2.8 HSM, still undecided) in
order to shoot mainly mountain biking. I read contradictory
comments on photographyreview.com concerning the difference between
Canon USM and Sigma HSM autofocusing, some were saying that AI
servo performance using the aforementioned Sigma lens was
noticeably inferior compared to what we get using one of those
Canon lenses ; others said that AF is virtually as fast on USM and
HSM lenses.

Since this point is very important in my decision process, could
those that have tested both lenses (Sigma 70-200 2.8 HSM and Canon
70-200 f2.8 or f4 L) share their impressions ?

Thanks in advance,
Shabok

 

[Darren Moore]

Took some bird photos last weekend with the birds gliding past right beside me and flying straight towards me. the AI kept up fine on my new 28-135 IS USM lense, but my old 75-300 couldn't handle it. I think the AI works best with the latest version of USM lenses

Darren

Hi all,

I'm going to buy a 70-200 lens (2nd hand Canon 70-200 f2.8L, new
Canon 70-200 f4L or new Sigma 70-200 f2.8 HSM, still undecided) in
order to shoot mainly mountain biking. I read contradictory
comments on photographyreview.com concerning the difference between
Canon USM and Sigma HSM autofocusing, some were saying that AI
servo performance using the aforementioned Sigma lens was
noticeably inferior compared to what we get using one of those
Canon lenses ; others said that AF is virtually as fast on USM and
HSM lenses.

Since this point is very important in my decision process, could
those that have tested both lenses (Sigma 70-200 2.8 HSM and Canon
70-200 f2.8 or f4 L) share their impressions ?

Thanks in advance,
Shabok

--
~

 

[hyslopc]

I took some D60 AI-Servo photos of my 2-year-old daughter running towards me with my 28-135, and many of them were out of focus.She runs a little slower than I walk, so I was pretty disappointed. It was late evening, and I was zoomed in to around 110mm - could it be the aperture? They were 1/60 F5.6.

Darren

Hi all,

I'm going to buy a 70-200 lens (2nd hand Canon 70-200 f2.8L, new
Canon 70-200 f4L or new Sigma 70-200 f2.8 HSM, still undecided) in
order to shoot mainly mountain biking. I read contradictory
comments on photographyreview.com concerning the difference between
Canon USM and Sigma HSM autofocusing, some were saying that AI
servo performance using the aforementioned Sigma lens was
noticeably inferior compared to what we get using one of those
Canon lenses ; others said that AF is virtually as fast on USM and
HSM lenses.

Since this point is very important in my decision process, could
those that have tested both lenses (Sigma 70-200 2.8 HSM and Canon
70-200 f2.8 or f4 L) share their impressions ?

Thanks in advance,
Shabok

--
~

 

[Inka]

I just shoot over 1000 pics about athletics with D60, and what I noticed was: Canon L200 2.8 works brilliantly with AI servo, of course there was some unsharp, but sigma 28-70 2.8 and 20 1.8 was not fast enough for AI servo and moving objekts and the pics quality was bad against canons L 200.
And now I am looking only canons L-series lensess.

Hi all,

I'm going to buy a 70-200 lens (2nd hand Canon 70-200 f2.8L, new
Canon 70-200 f4L or new Sigma 70-200 f2.8 HSM, still undecided) in
order to shoot mainly mountain biking. I read contradictory
comments on photographyreview.com concerning the difference between
Canon USM and Sigma HSM autofocusing, some were saying that AI
servo performance using the aforementioned Sigma lens was
noticeably inferior compared to what we get using one of those
Canon lenses ; others said that AF is virtually as fast on USM and
HSM lenses.

Since this point is very important in my decision process, could
those that have tested both lenses (Sigma 70-200 2.8 HSM and Canon
70-200 f2.8 or f4 L) share their impressions ?

Thanks in advance,
Shabok

 

[Inka]

Your speed value was not fast enough, I think It should be about 500-or more if you wana get sharp pictures in that situation.

Darren

Hi all,

I'm going to buy a 70-200 lens (2nd hand Canon 70-200 f2.8L, new
Canon 70-200 f4L or new Sigma 70-200 f2.8 HSM, still undecided) in
order to shoot mainly mountain biking. I read contradictory
comments on photographyreview.com concerning the difference between
Canon USM and Sigma HSM autofocusing, some were saying that AI
servo performance using the aforementioned Sigma lens was
noticeably inferior compared to what we get using one of those
Canon lenses ; others said that AF is virtually as fast on USM and
HSM lenses.

Since this point is very important in my decision process, could
those that have tested both lenses (Sigma 70-200 2.8 HSM and Canon
70-200 f2.8 or f4 L) share their impressions ?

Thanks in advance,
Shabok

--
~

 

[hyslopc]

Do I feel like an idiot or what?

Darren

Hi all,

I'm going to buy a 70-200 lens (2nd hand Canon 70-200 f2.8L, new
Canon 70-200 f4L or new Sigma 70-200 f2.8 HSM, still undecided) in
order to shoot mainly mountain biking. I read contradictory
comments on photographyreview.com concerning the difference between
Canon USM and Sigma HSM autofocusing, some were saying that AI
servo performance using the aforementioned Sigma lens was
noticeably inferior compared to what we get using one of those
Canon lenses ; others said that AF is virtually as fast on USM and
HSM lenses.

Since this point is very important in my decision process, could
those that have tested both lenses (Sigma 70-200 2.8 HSM and Canon
70-200 f2.8 or f4 L) share their impressions ?

Thanks in advance,
Shabok

--
~

 

[Darren Moore]

I can't see shutter speed having any direct effect on this.

Look at time taken to focus and whether the camera hunts at all while focusing. Best results for focusing are always bright light and high contrast. Seeing as the camera has to track a moving object it must also refocus very quickly. So bright light is an ideal situation.

Hence indirectly shutter speed is another consequence of good light. Thus if you have a high shutter speed you either have plenty of light and AI should work fine, a high ISO or very fast lense in which case you takes your chances.

So if you really want to see what it'll do, try the running daughter thing again outside on a bright sunny day.

Darren

Darren

Hi all,

I'm going to buy a 70-200 lens (2nd hand Canon 70-200 f2.8L, new
Canon 70-200 f4L or new Sigma 70-200 f2.8 HSM, still undecided) in
order to shoot mainly mountain biking. I read contradictory
comments on photographyreview.com concerning the difference between
Canon USM and Sigma HSM autofocusing, some were saying that AI
servo performance using the aforementioned Sigma lens was
noticeably inferior compared to what we get using one of those
Canon lenses ; others said that AF is virtually as fast on USM and
HSM lenses.

Since this point is very important in my decision process, could
those that have tested both lenses (Sigma 70-200 2.8 HSM and Canon
70-200 f2.8 or f4 L) share their impressions ?

Thanks in advance,
Shabok

--
~

 

[androm]

I would go for the IS version

I'll post a mode 2 panning hand held pic using the 100-400 IS (have 70-200 IS as well) from the
RIAT last weekend in a seperate thread

I'm going to buy a 70-200 lens (2nd hand Canon 70-200 f2.8L, new
Canon 70-200 f4L or new Sigma 70-200 f2.8 HSM, still undecided) in
order to shoot mainly mountain biking. I read contradictory
comments on photographyreview.com concerning the difference between
Canon USM and Sigma HSM autofocusing, some were saying that AI
servo performance using the aforementioned Sigma lens was
noticeably inferior compared to what we get using one of those
Canon lenses ; others said that AF is virtually as fast on USM and
HSM lenses.

Since this point is very important in my decision process, could
those that have tested both lenses (Sigma 70-200 2.8 HSM and Canon
70-200 f2.8 or f4 L) share their impressions ?

Thanks in advance,
Shabok

 

[hyslopc]

In fact the sun's coming out right now, so I might head off to the park with her. This is one of the things I love about photography - it gets me outdoors!

 

[Joo]

Shutter speed might not have a direct affect on the AI Servo performance, but it probably explains why his shots of his daughter were out of focus. 1/60 second shutter speed is a little slow and he might be mistaking motion blur for out of focus shots. When shooting a moving target, you really should have a faster shutter speed.

With regards to more light, definitely more light should help AI Servo performance since AI Servo is an AF Mode and more light improves AF. Also, the center AF point is supposed to be more sensitive, so if you want the best AI Servo performance on a D30/D60, you hsould use the center AF point.

Joo

So if you really want to see what it'll do, try the running
daughter thing again outside on a bright sunny day.

Darren

Darren

Hi all,

I'm going to buy a 70-200 lens (2nd hand Canon 70-200 f2.8L, new
Canon 70-200 f4L or new Sigma 70-200 f2.8 HSM, still undecided) in
order to shoot mainly mountain biking. I read contradictory
comments on photographyreview.com concerning the difference between
Canon USM and Sigma HSM autofocusing, some were saying that AI
servo performance using the aforementioned Sigma lens was
noticeably inferior compared to what we get using one of those
Canon lenses ; others said that AF is virtually as fast on USM and
HSM lenses.

Since this point is very important in my decision process, could
those that have tested both lenses (Sigma 70-200 2.8 HSM and Canon
70-200 f2.8 or f4 L) share their impressions ?

Thanks in advance,
Shabok

--
~

--

• Canon EOS D-30 & PowerShot S100
• Maybe one day I'll take a decent picture. In the meantime, I'll blame the equipment.

 

[Mark J.]

I can't see shutter speed having any direct effect on this.

Of course it has a direct effect. If the object is moving enough so that within the time the shutter is open (in this case 1/60th of a sec) there is any significant change in the object's position, the moving object will be blurred.

It won't matter whether the camera is focusing accurately or not.

[SNIP!]

So if you really want to see what it'll do, try the running
daughter thing again outside on a bright sunny day.

On a bright sunny day at 1/60th of a sec, she will still be blurry.

In this sort of situation, using Shutter Priority mode (Tv) and a higher shutter speed will freeze the motion and let you evaluate better whether your lens is focusing properly or not. You may not need to go as high as 1/500th, but 1/60th is pretty slow.

Darren

Your speed value was not fast enough, I think It should be about
500-or more if you wana get sharp pictures in that situation.

[SNIP!]

Regards,

Mark J.

--
EOS 1D ~ 28-70mm f/2.8 L ~ 100-400mm f/4.5-5.6 L IS ~ 20-35mm f/3.5-4.5 ~ 550EX

 

[Gordon Berg]

My understanding of AF is that the maximum apeture of a lens comes into play by not the Fstop you set for the picture... I presume that means that it focuses wide open then stops down.

If you were shooting at ISO 100, that's not very low light - EV 11+. A closing rate of 4 ft/sec shouldn't challenge the tracking system. I don't know what the internal delay from focus lock to shutter release is but I wouldn't think it would be over 15 milliseconds... time enough for her to travel less than an inch - two inches by the time the shutter closed. At F5.6 you couldn't see if the focus were off that little. I think that exonerates the tracking in this case. Sounds more like just a failure to focus accurately.

You know I could imagine AI Servo tracking working against you in case where the focussing is somewhat inaccurate and the speed is low. It has to focus twice to compute a closing rate and the error may indicate a significant speed in the wrong direction, in which case it would adjust so the focus point was further off than either of the actual focus distance computations. When the speed is high, the errors are likely to be small relative to the change in distance and while the projection of future position will amplify the errors, you're still better off to project.

Darren

Hi all,

I'm going to buy a 70-200 lens (2nd hand Canon 70-200 f2.8L, new
Canon 70-200 f4L or new Sigma 70-200 f2.8 HSM, still undecided) in
order to shoot mainly mountain biking. I read contradictory
comments on photographyreview.com concerning the difference between
Canon USM and Sigma HSM autofocusing, some were saying that AI
servo performance using the aforementioned Sigma lens was
noticeably inferior compared to what we get using one of those
Canon lenses ; others said that AF is virtually as fast on USM and
HSM lenses.

Since this point is very important in my decision process, could
those that have tested both lenses (Sigma 70-200 2.8 HSM and Canon
70-200 f2.8 or f4 L) share their impressions ?

Thanks in advance,
Shabok

--
~

 

[jonlg]

Shutter speed might not have a direct affect on the AI Servo
performance, but it probably explains why his shots of his daughter
were out of focus. 1/60 second shutter speed is a little slow and
he might be mistaking motion blur for out of focus shots. When
shooting a moving target, you really should have a faster shutter
speed.

These photos http://www.pbase.com/jonlg/thumper_nats_mxcc were shot with AI servo and 125 to 250 sec.

Either my skills are going or AI servo is poor, but I really expected better than I shot.

Admittedly I was using the slow 75-300 IS (yes IS was off), but it was a bright sunshiny day.

Thus my earlier questions in regards to "panning IS mode" and the faster "RING USM" if that will be of any benefit.
Yes I know the 1D is the answer but I'm stuck with the D60.
Any more thought
John

 

[hockeyshooter]

Same question/same lens, as I like to shoot motocross.
On the weekend my 75-300 and D60 and me had only a few good ones.
What we were shooting was elementry, so it's got me buggered.
I figure the AI can't keep up, but hope this is not the case.

I shot about 150 frames at a motocross event last Sunday. It was bright and sunny weather and I was using the EF 70-200/2.8 L on my D30 on AI Servo mode. I adjusted the ISO setting as necessary to keep the shutter speed over 1/1000 sec and picked the central AF sensor.

I noticed that a good deal of the images are not as sharp as they really should be. In some instances where one bike is following another, even though I was following the closer bike with the focus sensor, its the following bike that is sharp.

But its not a straightforward front-focussing problem because some shots taken on the same settings are sharp. I think its basically down to the poor AF performance of the D30. If the EOS1D wasn't so absurdly expensive, I'd buy one in a flash - the vast majority of my work is sports.

Chris.
--
http://www.wildphotos.org/motoX/

 

[hockeyshooter]

These photos http://www.pbase.com/jonlg/thumper_nats_mxcc were
shot with AI servo and 125 to 250 sec.

Actually the shots you have posted look fairly sharp and I would say they are sharper than my results, but I guess it rather depends on how much processing you've done to them.

Incidentally I thnk your watermarking is a little too subtle - it could be edited out quite easily.

Chris.
--
http://www.hockeyphotos.com/

 

[David Knight]

Inka

Glad you're getting on ok with the athletics shooting, do you have any on the web to share?

David

Hi all,

I'm going to buy a 70-200 lens (2nd hand Canon 70-200 f2.8L, new
Canon 70-200 f4L or new Sigma 70-200 f2.8 HSM, still undecided) in
order to shoot mainly mountain biking. I read contradictory
comments on photographyreview.com concerning the difference between
Canon USM and Sigma HSM autofocusing, some were saying that AI
servo performance using the aforementioned Sigma lens was
noticeably inferior compared to what we get using one of those
Canon lenses ; others said that AF is virtually as fast on USM and
HSM lenses.

Since this point is very important in my decision process, could
those that have tested both lenses (Sigma 70-200 2.8 HSM and Canon
70-200 f2.8 or f4 L) share their impressions ?

Thanks in advance,
Shabok

 

[hyslopc]

On a bright sunny day, I wouldn't have to use 1/60. The problem is that my 28-135 has a max aperture of 5.6 at 110mm, and it was late in the evening, so even at 1/60 the shots were a bit dim. At 1/250 the shots would have been way underexposed.

Incidentally, I was using the center focus point (someone else asked about this).

I will try this again as soon as the sun comes out, and then we'll know for sure how good the D60's AI Servo is. It looks like rain for now so I'm stuck indoors.

On a bright sunny day at 1/60th of a sec, she will still be blurry.

In this sort of situation, using Shutter Priority mode (Tv) and a
higher shutter speed will freeze the motion and let you evaluate
better whether your lens is focusing properly or not. You may not
need to go as high as 1/500th, but 1/60th is pretty slow.

 

[David Knight]

Or you could up the ISO. Living in London as I do you could wait for ever for that yellow thing in the sky.

Incidentally, I was using the center focus point (someone else
asked about this).

I will try this again as soon as the sun comes out, and then we'll
know for sure how good the D60's AI Servo is. It looks like rain
for now so I'm stuck indoors.

On a bright sunny day at 1/60th of a sec, she will still be blurry.

In this sort of situation, using Shutter Priority mode (Tv) and a
higher shutter speed will freeze the motion and let you evaluate
better whether your lens is focusing properly or not. You may not
need to go as high as 1/500th, but 1/60th is pretty slow.