A question about multiple AF points

[Olybirder]

How does the camera choose which AF point to use when 5 or 9 points are selected? For a long time I have noticed that the choice seems to be rather random and experiments today have confirmed this.

If I place the central circular marking in the viewfinder directly over the subject which I would like to be in focus (flower, bird's eye etc.) and half press the shutter button, I would expect the centre AF green rectangle to light up. However, in practice, the camera seems to choose any AF point apart from the centre one. If I keep lightly pressing and releasing the button the rectangle jumps about, apparently at random, but rarely choosing the central position.

This is really an issue when I am using C-AF with 5 or 9 points for BIF shots. By the time I get the green rectangle to light up on the bird, it is usually too late and the moment has passed.

Does anyone know the logic which the camera uses to select the AF point and if there is any way to make it choose the point which I want?

Ron

[MJ224]

Re: A question about multiple AF points

Can't help, but can confirm the focus point is not always on the point I want............

I use a 5 point focus set up when biffing, often the boids are against a clear sky and generally it works OK

[Walti]

Re: A question about multiple AF points

I've been playing with the camera and my finding suggests that if there is little difference in distance e.g. a flat surface it then seems totally random.

However if there is a significant difference in distance e.g. something perpendicular to the camera then it seems to select the nearest focus point each time. so on a BIF the nearest wingtip perhaps?

[Olybirder]

Re: A question about multiple AF points

It seems that it is not a common problem. I must be doing something wrong. Back to the drawing board.

Ron

[mik]

Re: A question about multiple AF points

I wonder if the only logic is the camera chooses what ever point is the best part for a focus lock.
Been trying 5 and 9 even tracking with the new camera and as you say its frustrating watching the focus confirm box jump about.

[GyRob]

Re: A question about multiple AF points

Yes it does jump about ,it would be nice if the centre point had priority and only changed if it lost focus but as the camera is looking for contrast and or nearest single point is the only way .
That said YOU then have to be spot on -not easy .
Rob.

[OM USer]

Re: A question about multiple AF points

I too have this problem. I use the zoom focus option on a function button to overcome it as you only get one focus point and can adjust the size of it to match your target (if you have the time). A long press on the OK button will central it if it ever wanders off.

[TonyR]

Re: A question about multiple AF points

How does the camera choose which AF point to use when 5 or 9 points are selected? For a long time I have noticed that the choice seems to be rather random and experiments today have confirmed this.

If I place the central circular marking in the viewfinder directly over the subject which I would like to be in focus (flower, bird's eye etc.) and half press the shutter button, I would expect the centre AF green rectangle to light up. However, in practice, the camera seems to choose any AF point apart from the centre one. If I keep lightly pressing and releasing the button the rectangle jumps about, apparently at random, but rarely choosing the central position.

This is really an issue when I am using C-AF with 5 or 9 points for BIF shots. By the time I get the green rectangle to light up on the bird, it is usually too late and the moment has passed.

Does anyone know the logic which the camera uses to select the AF point and if there is any way to make it choose the point which I want?

Ron

I don't know the answer to your question but I would like to.

However, I have another related question that I would also like to know the answer to. That is - how and when does the camera decide to change selected AF point during C-AF when 5, 9 or all points selected?

This is pretty important. What I would like it to do is switch to the point that has a subject at the expected distance. i.e. as the bird moves towards or away from you, a prediction method is used to estimate the expected distance to the subject and then find the AF point which matches. Then focus.

However, it might do any other number of things. It might stick with the first point chosen until that point doesn't have a subject. It might keep switching to the focus point with the nearest subject. etc.

In essence, a computer programmer might write either:

start loop
{
select "best" focus point
Focus useing selected point
} repeat

or

select initial focus point
start loop
{
Focus using selected point
If focus failed
{
Select another point
}
} repeat

or something else ...

Sorry if this is gibberish! I suspect that the camera does something more like the second method. In a burst of shots, I always get some soft images but I am beginning to think that my results are better when I am better at keeping the original point over the target. This may prevent the camera from switching AF points. Maybe the soft images are as the camera switches points. One thing I am sure of is that the green AF confirmation box is not illuminated all the time with 5, 9 or even all-points selected.

Anyway, once again, it would be helpful if Olympus would give us a clue about all of this stuff.