difference in selected AF point

jcnz

New member
Messages
9
Reaction score
0
3bff45e81bfd4f04bf22cd8b8ff0cc59.jpg

6fae2b018efd430f9623d3e3948430b0.jpg

I used my 60D with 18-200 lens to shot these.

Both are at 1/160 sec, F/10, ISO 100, focal length 80mm.

In the top one, I chose the central focus point to auto-focus; (at the butterfly)

In the bottom one, I picked the top-right focus point to auto-focus.

I was expecting that they should look as sharp as each other but the side focus point doesn't seem as sharp.

Is that normal for 60D or is it to do with the lens? (should I send them back to Canon to check?)
 
Last edited:
Solution
jcnz wrote:
Wei Kelun wrote:

Did you recompose after focus lock? Or did you crop the photos so that the butterfly is in the center in each one?

One other note: like I said above, there are a ton of contrasty areas in your scene, all tight within a plane. So it could be that you'd have more success ensuring that the focus is on your butterfly (and not the wall) by using Spot AF instead of Single-point AF.
No, I didn't recompose after focus lock, but I did crop both photos, the original photos are much larger ones, I cropped them just around the butterfly to show the difference.

How do you mean by Spot AF?

As far as I can see, 60D lets me do 2 things;

1 to light up all 9 AF points and it will choose the AF point(s) for me...
I think everything's fine with your gear. It looks like the first shot focussed on the butterfly, and the second shot focussed on the wall behind it. The wall has lots of contrasty parts in the texture of the brick, and that will attract the attention of your autofocus points.
 
The top right focus point will be somewhere on the wall, behind the butterfly, so the wall would be sharper and the butterfly blurred on the second shot. At 1/160 sec I am not sure that there is not some motion blur involved, but since the pictures are quite small it is really not possible to say.

You did not use a tripod, did you?
 
Wei Kelun wrote:

I think everything's fine with your gear. It looks like the first shot focussed on the butterfly, and the second shot focussed on the wall behind it. The wall has lots of contrasty parts in the texture of the brick, and that will attract the attention of your autofocus points.
Thanks Wei.

I manually picked the focus point and pointed it at the butterfly, the AF point lights up and I shot.

Can that still cause the focus point go off to the wall?
 
Klaus dk wrote:

The top right focus point will be somewhere on the wall, behind the butterfly, so the wall would be sharper and the butterfly blurred on the second shot. At 1/160 sec I am not sure that there is not some motion blur involved, but since the pictures are quite small it is really not possible to say.

You did not use a tripod, did you?
Thanks Klaus.

No i didn't use a tripod.

I was thinking that since my shutter speed is twice of the focal length, I should be fine ... I will try again with a tripod.

My concern though is that, when I am out shooting with my kid, it's hard to ask a child to hold a position while I set up my tripod ... :-) but i guess that's not Canon's fault
 
jcnz wrote:
Wei Kelun wrote:

I think everything's fine with your gear. It looks like the first shot focussed on the butterfly, and the second shot focussed on the wall behind it. The wall has lots of contrasty parts in the texture of the brick, and that will attract the attention of your autofocus points.
Thanks Wei.

I manually picked the focus point and pointed it at the butterfly, the AF point lights up and I shot.

Can that still cause the focus point go off to the wall?
Did you recompose after focus lock? Or did you crop the photos so that the butterfly is in the center in each one?

Also, the center AF point is about twice as sensitive as the peripheral AF points, and that could be your issue here.

One other note: like I said above, there are a ton of contrasty areas in your scene, all tight within a plane. So it could be that you'd have more success ensuring that the focus is on your butterfly (and not the wall) by using Spot AF instead of Single-point AF.
 
Last edited:
Wei Kelun wrote:

Did you recompose after focus lock? Or did you crop the photos so that the butterfly is in the center in each one?

One other note: like I said above, there are a ton of contrasty areas in your scene, all tight within a plane. So it could be that you'd have more success ensuring that the focus is on your butterfly (and not the wall) by using Spot AF instead of Single-point AF.
No, I didn't recompose after focus lock, but I did crop both photos, the original photos are much larger ones, I cropped them just around the butterfly to show the difference.

How do you mean by Spot AF?

As far as I can see, 60D lets me do 2 things;

1 to light up all 9 AF points and it will choose the AF point(s) for me out of the 9 points; (did you refer to this setting?)

2 Or I manually pick one of the 9 points and 60D focuses at that (this is what I did).

Thanks
 
jcnz wrote:
Wei Kelun wrote:

Did you recompose after focus lock? Or did you crop the photos so that the butterfly is in the center in each one?

One other note: like I said above, there are a ton of contrasty areas in your scene, all tight within a plane. So it could be that you'd have more success ensuring that the focus is on your butterfly (and not the wall) by using Spot AF instead of Single-point AF.
No, I didn't recompose after focus lock, but I did crop both photos, the original photos are much larger ones, I cropped them just around the butterfly to show the difference.

How do you mean by Spot AF?

As far as I can see, 60D lets me do 2 things;

1 to light up all 9 AF points and it will choose the AF point(s) for me out of the 9 points; (did you refer to this setting?)

2 Or I manually pick one of the 9 points and 60D focuses at that (this is what I did).

Thanks
Sorry, I'm a newcomer here and I'm not used to all the different camera types on the forum. I thought we're talking about the 7D, where you can kind of "zoom in" on each autofocus point for more pinpoint precision. Also, what I said about the center autofocus point being twice as sensitive as the peripheral ones seems to be a 7D thing, too. I'll be more careful to make sure which camera we're talking about going forward.
 
Last edited:
Wei Kelun wrote:

Sorry, I'm a newcomer here and I'm not used to all the different camera types on the forum. I thought we're talking about the 7D, where you can kind of "zoom in" on each autofocus point for more pinpoint precision. Also, what I said about the center autofocus point being twice as sensitive as the peripheral ones seems to be a 7D thing, too. I'll be more careful to make sure which camera we're talking about going forward.
That's OK, thanks very much for your input.
 
jcnz wrote:
Klaus dk wrote:

The top right focus point will be somewhere on the wall, behind the butterfly, so the wall would be sharper and the butterfly blurred on the second shot. At 1/160 sec I am not sure that there is not some motion blur involved, but since the pictures are quite small it is really not possible to say.

You did not use a tripod, did you?
Thanks Klaus.

No i didn't use a tripod.

I was thinking that since my shutter speed is twice of the focal length, I should be fine ... I will try again with a tripod.

My concern though is that, when I am out shooting with my kid, it's hard to ask a child to hold a position while I set up my tripod ... :-) but i guess that's not Canon's fault
It is difficult to tell from your samples if motion blur is actually involved, but 1/160 is no insurance against it. My point is, that the subject you focus on when using the upper right focus point is further away from the camera than the butterfly, so it is no wonder the butterfly is not in focus. If you had taken a picture of a flat wall parallel to the sensor, it would make little difference which focus point you select, but if different parts of the subject are at different distances from the camera, the choice of focus point makes a difference.

I know a tripod is not always handy, but shooting macro without one is not easy.
 
I tried again with a tripod and I can't really tell the difference in sharpness using different focus points (although I don't have the same butterfly there again).

Lessons learnt, thank you very much.
 
jcnz wrote:
Klaus dk wrote:

The top right focus point will be somewhere on the wall, behind the butterfly, so the wall would be sharper and the butterfly blurred on the second shot. At 1/160 sec I am not sure that there is not some motion blur involved, but since the pictures are quite small it is really not possible to say.

You did not use a tripod, did you?
Thanks Klaus.

No i didn't use a tripod.

I was thinking that since my shutter speed is twice of the focal length, I should be fine ... I will try again with a tripod.

My concern though is that, when I am out shooting with my kid, it's hard to ask a child to hold a position while I set up my tripod ... :-) but i guess that's not Canon's fault
One remark about 1/160 being twice your used focallenght.

In my exif data it shows the focallenght of the lens not the focallenght including cropfactor of the sensor (in case of a 60D 1.6).

So in order using twice the focallenght you should use 1/250 or even faster.


Using a tripod it is advized to switch of IS of the lens.

Good luck.
 
jcnz, your images have dimensions 765 pixels by 510 pixels (unless I am having trouble understanding the new forum photo viewing widgets!). This is a really tiny crop from your photo, only showing about 1/8 of the full width.

This is probably as good as you can expect. If we're only looking at an area 1/8 as wide as your actual photo, you've effectively blown it up to 8 times as big. That means any focusing error or blur looks 8 times worse than it would if you were viewing the whole photo at the same output size.

I suspect the butterfly was too small for the AF sensor to focus on precisely (regardless of which AF point you used, and even if you were in spot AF mode), so the camera could have been slightly front or back focused from what you expected.

But more importantly, 2x focal length is not nearly enough to avoid blur due to camera shake for such a tiny crop enlarged to this size on our screens. Remember all those rules of thumb about handheld shutter speed are based on assumptions about typical output size and viewing distance for your photo. If you are enlarging the photo as much as you are here (at my monitor size, the entire photo would probably be 4-5 feet wide!), you need to adjust shutter speed accordingly, to compensate for camera shake.

My guess is that anyway, at a normal viewing size, since you shot at f/10 (a pretty small aperture), the butterfly was probably tack sharp, and everything around it well within the acceptable DOF.

So don't worry, be happy!
 
jcnz wrote:
Wei Kelun wrote:

Did you recompose after focus lock? Or did you crop the photos so that the butterfly is in the center in each one?

One other note: like I said above, there are a ton of contrasty areas in your scene, all tight within a plane. So it could be that you'd have more success ensuring that the focus is on your butterfly (and not the wall) by using Spot AF instead of Single-point AF.
No, I didn't recompose after focus lock, but I did crop both photos, the original photos are much larger ones, I cropped them just around the butterfly to show the difference.

How do you mean by Spot AF?

As far as I can see, 60D lets me do 2 things;

1 to light up all 9 AF points and it will choose the AF point(s) for me out of the 9 points; (did you refer to this setting?)

2 Or I manually pick one of the 9 points and 60D focuses at that (this is what I did).

Thanks
Ah, there is your explanation! There is a big difference between cropping from the center and from a corner, especially when you use a lens like the 18-200. This lens is know to be significantly less sharp off center:

http://www.photozone.de/canon-eos/400-canon_18200_3556is

Furthermore, when you crop, the effective focal length of the lens grows, so your shutter speed should be equally shorter.

Imagine a perfect 30mm lens. If you crop from it so the FOV is equal to 300mm, you enlarge the picture 10 times (linear) and throw away 99% (90% squared) of the picture, so motion blur would also be enlarged ten times.

The shutter speed rule just stipulates what shutter speed to use to get acceptable sharpness at a certain FOV.
 
Solution
Klaus dk wrote:
jcnz wrote:
Wei Kelun wrote:

Did you recompose after focus lock? Or did you crop the photos so that the butterfly is in the center in each one?

One other note: like I said above, there are a ton of contrasty areas in your scene, all tight within a plane. So it could be that you'd have more success ensuring that the focus is on your butterfly (and not the wall) by using Spot AF instead of Single-point AF.
No, I didn't recompose after focus lock, but I did crop both photos, the original photos are much larger ones, I cropped them just around the butterfly to show the difference.

How do you mean by Spot AF?

As far as I can see, 60D lets me do 2 things;

1 to light up all 9 AF points and it will choose the AF point(s) for me out of the 9 points; (did you refer to this setting?)

2 Or I manually pick one of the 9 points and 60D focuses at that (this is what I did).

Thanks
Ah, there is your explanation! There is a big difference between cropping from the center and from a corner, especially when you use a lens like the 18-200. This lens is know to be significantly less sharp off center:

http://www.photozone.de/canon-eos/400-canon_18200_3556is


Furthermore, when you crop, the effective focal length of the lens grows, so your shutter speed should be equally shorter.

Imagine a perfect 30mm lens. If you crop from it so the FOV is equal to 300mm, you enlarge the picture 10 times (linear) and throw away 99% (90% squared) of the picture, so motion blur would also be enlarged ten times.

The shutter speed rule just stipulates what shutter speed to use to get acceptable sharpness at a certain FOV.
Yes, there's your answer. This is a lens issue, not a problem with your 60D.

On a side note. If you have IS engaged and it's working properly, then your shutter speed was plenty fast enough.
 

Keyboard shortcuts

Back
Top