Auto focus point

I usually shoot in manual mode and I noticed that the auto focus point I picked when taking the picture, wouldn't be displayed when I viewed the image using ZoomBrowser EX.
If I shoot using auto mode, the focus point would be displayed.
So, I just assumed that the AF point couldn't be displayed when shooting in manual.

Then purely by accident, I discovered that if I continued to hold the AF-on button(using back button focusing) while taking the picture, then the AF point is displayed when viewing the image with ZoomBrowser.
The AF point also appears on the camera's lcd screen.

Have I been doing something wrong by not keeping the AF-on button pressed while taking the picture as far as focus is concerned?
I thought by pressing the AF-on button and then releasing it when using back button focusing, the focus stayed locked until the picture was taken.

AF point display is a somewhat useless feature...it shows which AF point was active at the time of focus, but it fails to show which object in the scene was under that AF point. You could AF on your child standing on a boat launch in the center of the frame and then reframe the shot while holding down the shutter button half way, thereby putting your child at the edge of the frame -- but the center AF point will be somewhere over the lake.

canongear wrote in post #14520410
Have I been doing something wrong by not keeping the AF-on button pressed while taking the picture as far as focus is concerned?
I thought by pressing the AF-on button and then releasing it when using back button focusing, the focus stayed locked until the picture was taken.

No, you are doing fine, the focus does stay locked when you take your finger off. The key point here is, as Wilt pointed out, zoombrowser displays the active AF point. As soon as you take your finger off the back button, the point is no longer active. Focus has been locked in and the focus point is "turned off". When you switch to auto, the point will display in zoombrowser, because the focus point activation returns to the shutter button, and is therefore still active when the shot is taken.

As Wilt also points out though, AF point display isn't much use as it doesn't show where you focused, just which point you used. As soon as you recompose it will show the point as being over something completely different to where you focused.

Example of false display of AF point. Red squares indicate active AF points...but they were on top of the toy bike (pretend there is a child sitting there!) at the time of focus, yet the indicated AF point would lead you to believe that you focused on the yellow flower!

Au contraire, on my 40D I have back button AF (AF-on button) and the shutter button only locks exposure and fires shutter, yet Zoombrowser was able to display the active AF points (I had enabled all AF points for this example shot).

Wilt wrote in post #14521164
Au contraire, on my 40D I have back button AF (AF-on button) and the shutter button only locks exposure and fires shutter, yet Zoombrowser was able to display the active AF points (I had enabled all AF points for this example shot).

Ah, OK.

I was going with that as I couldn't see any other reason why zoombrowser would show the point when he held the AF-ON button down and not when he didn't. The logic would be that it doesn't show it because the point isn't activated when the button is released. In that case, I don't know why his doesn't, and yours does, show the points when the button isn't depressed.

Maybe different bodies work that way, I don't know.

Wilt wrote in post #14521068
AF point display is a somewhat useless feature...it shows which AF point was active at the time of focus, but it fails to show which object in the scene was under that AF point. You could AF on your child standing on a boat launch in the center of the frame and then reframe the shot while holding down the shutter button half way, thereby putting your child at the edge of the frame -- but the center AF point will be somewhere over the lake.

I understand most of what you are saying but, this is what I don't understand about the focus/recompose idea.
If the idea behind manually picking the AF point is to get a particular part of the image in better/sharper focus, wouldn't recomposing the shot once that is done, defeat the purpose?
I realize the focus has been locked once the AF ON button has been pressed but, the placement of the subject has now changed due to the recomposing of the shot.
Would this not change the focus quality in some way?
Focus/recompose is a common technique but, I just don't completely understand it I guess.

sandpiper wrote in post #14521405
Ah, OK.
Maybe different bodies work that way, I don't know.

I'm using a 40D as well.

canongear wrote in post #14522523
I understand most of what you are saying but, this is what I don't understand about the focus/recompose idea.
If the idea behind manually picking the AF point is to get a particular part of the image in better/sharper focus, wouldn't recomposing the shot once that is done, defeat the purpose?
I realize the focus has been locked once the AF ON button has been pressed but, the placement of the subject has now changed due to the recomposing of the shot.
Would this not change the focus quality in some way?
Focus/recompose is a common technique but, I just don't completely understand it I guess.

I'm using a 40D as well.

Yes recompose after focus does alter the true distance to the subject from the focused distance. However, I have calculated that DOF will easily contain any recompose focus error when the angular shift is modest (under about 15 degrees). The actual distance to the subject plane after recompose is closer than focused plane, and since a lot of lenses tend toward forward focus more than backfocus, that helps, too.
This chart shows the amount of recompose error vs. DOF for a FF camera...if the box is pink, DOF is exceeded by the recompose-induced error.

Some of the information provided in the chart, is a bit over my head.
However, I do appreciate your help in trying to show the effects of focus/recompose.
Especially if you did the work in creating the chart.

canongear wrote in post #14526907
Some of the information provided in the chart, is a bit over my head.
However, I do appreciate your help in trying to show the effects of focus/recompose.
Especially if you did the work in creating the chart.

Let's look at one row of the chart that I created, row 6...

1. Initial subject distance is 5', and
   
2. the recompose angle change is 15 degrees of arc.
   
3. That changes the subject in the frame by a horizontal distance of 1.34'.
   
4. The recompose angle causes the true subject distance to be 4.83'.
   
5. With a 50mm lens at f/2 on FF body, there is a distance of 4.85' to the near part of the DOF zone; since 4.83' is closer to the camera than 4.85', the DOF zone does not hide the shift in focus, so the box is pink
   
6. With a 50mm lens at f/4 on FF body, there is a distance of 4.72' to the near part of the DOF zone; since 4.72' DOF front is closer than subject at 4.83', the DOF zone does hide the shift in focus, so the box is not pink
   
7. With a 100mm lens at f/2 on FF body, a recompose shift of 15 degrees actually moves the subject out of the FOV of the frame, so this situation does not matter in terms of recompose focus shift!

Wilt wrote in post #14527164
Let's look at one row of the chart that I created, row 6...

1. Initial subject distance is 5', and
   
2. the recompose angle change is 15 degrees of arc.
   
3. That changes the subject in the frame by a horizontal distance of 1.34'.
   
4. The recompose angle causes the true subject distance to be 4.83'.
   
5. With a 50mm lens at f/2 on FF body, there is a distance of 4.85' to the near part of the DOF zone; since 4.83' is closer to the camera than 4.85', the DOF zone does not hide the shift in focus, so the box is pink
   
6. With a 50mm lens at f/4 on FF body, there is a distance of 4.72' to the near part of the DOF zone; since 4.72' DOF front is closer than subject at 4.83', the DOF zone does hide the shift in focus, so the box is not pink
   
7. With a 100mm lens at f/2 on FF body, a recompose shift of 15 degrees actually moves the subject out of the FOV of the frame, so this situation does not matter in terms of recompose focus shift!

I think part of the confusion may lie in the statement that the "initial subject distance" is 5' and then the recompose angle causing the "true subject distance" to be 4.83'. Surely the subject doesn't move, and one is the focused distance, presumably the first one? The way you have listed it suggests that the subject is at 5' and then you perform the focus / recompose action, rather than making it clear that you are focusing on a subject at 4.83' and the focus recompose action sets the focus plane at 5'.

It should be kept in mind though that whilst the point focused on is still within the DoF, that does not make it truly "in focus". Yes, the DoF will still mean that it is "acceptably sharp" but that is not quite the same thing. The sharpness still falls off anywhere away from the actual focus plane, in this case 5', so 4.83' is still missing focus by 0.17' and will be a little soft, being only 0.11' from where the sharpness drop off reaches "unacceptable". So, although still within the DoF it isn't completely hiding the shift in focus. Add in that DoF tables are usually slightly optimistic and are fairly vague to start with, relying on a certain size print at a certain distance and the persons eyesight being a little less than perfect, and that remaining 0.11' of "acceptable" sharpness can easily be lost. Certainly anybody pixel peeping would see a difference as they are looking at a much bigger image than the DoF tables calculate for.

When taking a head shot, the focus needs to be on the eyes (or the nearest eye), having the eyes merely "somewhere within the DoF" takes a lot away from the shot.

It also means that, by placing the focal plane some 0.17' behind the subject, you reduce the "wriggle room" to a little over an inch. If the photographer leans forward slightly as they recompose (or the subject does) and the shot ends up with the subject in front of the DoF zone, and so not acceptably sharp.

You may also need that extra bit of DoF at the front anyway. If photographing a typical dog fairly face on, you need the focal plane to be on the eyes even more, as if it is behind them, the eyes will be slightly soft (although within the DoF) but the nose will be well out of focus and unacceptably blurry, although the ears should look lovely. Getting the focal plane spot on will give lovely sharp eyes, and a slightly soft (but acceptable) nose and ears.

Your table is great, mathematically and theoretically, yet doesn't alter the fact that focus recompose can still cause issues in the real world when the table says you are still within the DoF zone.

Of course, you were very correct in your post where you stated "I think a contributing factor is the crazy preoccupation of so many today, for shooting with razor thin DOF using extra large aperture (f/1.2 and f/1.4) lenses." Focus recompose isn't a big issue when using a decent enough DoF to keep the focal plane well within the DoF zone, it is those who do it whilst using wide apertures for shallow DoF who will see the biggest problems.

sandpiper wrote in post #14528373
I think part of the confusion may lie in the statement that the "initial subject distance" is 5' and then the recompose angle causing the "true subject distance" to be 4.83'. Surely the subject doesn't move, and one is the focused distance, presumably the first one? The way you have listed it suggests that the subject is at 5' and then you perform the focus / recompose action, rather than making it clear that you are focusing on a subject at 4.83' and the focus recompose action sets the focus plane at 5'.

The statement in blue is totally wrong interpretation of the conditions that I was describing. The text in green does accurately reflect the true interpretion of what is happening. Here is an illustration of what happens to the plane of focus when a photographer introduces a recompose angle. I made this illustration quite a while ago, to depict the recompose angle, and have used it before on POTN.

sandpiper wrote in post #14528373
It should be kept in mind though that whilst the point focused on is still within the DoF, that does not make it truly "in focus". Yes, the DoF will still mean that it is "acceptably sharp" but that is not quite the same thing. The sharpness still falls off anywhere away from the actual focus plane, in this case 5', so 4.83' is still missing focus by 0.17' and will be a little soft, being only 0.11' from where the sharpness drop off reaches "unacceptable". So, although still within the DoF it isn't completely hiding the shift in focus. Add in that DoF tables are usually slightly optimistic and are fairly vague to start with, relying on a certain size print at a certain distance and the persons eyesight being a little less than perfect, and that remaining 0.11' of "acceptable" sharpness can easily be lost. Certainly anybody pixel peeping would see a difference as they are looking at a much bigger image than the DoF tables calculate for.

When taking a head shot, the focus needs to be on the eyes (or the nearest eye), having the eyes merely "somewhere within the DoF" takes a lot away from the shot.

It also means that, by placing the focal plane some 0.17' behind the subject, you reduce the "wriggle room" to a little over an inch. If the photographer leans forward slightly as they recompose (or the subject does) and the shot ends up with the subject in front of the DoF zone, and so not acceptably sharp.

You may also need that extra bit of DoF at the front anyway. If photographing a typical dog fairly face on, you need the focal plane to be on the eyes even more, as if it is behind them, the eyes will be slightly soft (although within the DoF) but the nose will be well out of focus and unacceptably blurry, although the ears should look lovely. Getting the focal plane spot on will give lovely sharp eyes, and a slightly soft (but acceptable) nose and ears.

Agree with your commentary.

I will make two counteracting points in response to the above...

• Yes, recompose can indeed cause issues, especially considering the fact that the usual DOF calculations assume lower human visual acuity than even the 20/20 vision standard that eye correction professionals try to achieve with eyeglass prescriptions. We typically see things out of focus even when the DOF calculators would lead us to believe 'in focus' on the print! My table used the less-than-reality lens manufacturer standard for DOF scales and tables that have been used for decades.
  
• How much did we ever hear about recompose focus issues in the days of manual focus cameras, before AF was invented and some people started to write about 'the problem'?!?! Not much, if at all.
  I had been reading photography magazines for most of my 45+ years of interest and involvement. Yet for over a half century film camera shooters have relied upon the precision focus aids (split image and microfocus) which are placed only at the very center of the viewfinder, to supplement the less precise surrounding groundglass screen, and the 'problem' did not warrant articles warning novices about an issue that needed to be considered.