camear canon 70d auto focus problem

[junqi]

i did some of the research on google to re compose your position meaning that you press the button half way down the focus point will be lock and you can move around. how come when i press my canon 70 button half way down the red color focus gone. when i half press and hold the button and on the first subject. the first subject is in focus mode but when i move to 2nd subject the subject is out of focus any idea ?

 

[Overread]

The cameras AF system works by finding the closest point of contrast difference it can detect in the scene under the AF point(s) that are active.

So if you are using a single AF point only that point will look for a contrast difference underneath itself in the scene. However because its only looking for a difference in contrast it has no idea of the "subject" that its focusing upon*.

In addition the focal plane is rather like a sheet of paper parallel to the front of the lens. When the focus moves this plane moves back and forth until it stops over the point of contrast difference that the AF sensors pick up upon. Anything covered by the paper is then in-focus (the thickness of the paper defined by the aperture and counts as the depth of field - ergo thin depth of field would be a thin sheet - whilst a wide depth of field would be very thick).


In focus and recompose when you focus the focus is locked, its not locked to the subject, but rather the position of the focal plane relative to the lens is locked- that sheet of paper is locked at a certain distance from the lens. Thus if you then move the lens and camera (recompose) whilst the focal plane is fixed there is a potential that whilst you're moving it the plane of focus could be moved off your original target subject.

In addition it sounds like you were hoping you could point at two different subjects and have the camera then focus upon both at the same time; this is not possible because a camera only has one focal plane to work with, so it can't focus on two separate areas. The way you'd cover this is to use a bigger f number (smaller aperture) to give a greater depth of field - ergo make the paper thicker so that the depth of field can cover both subjects. There are also a few other things you can do such as using a tilt shift lens or stacking multiple shots together - but for now stick with the simple idea of a single focal plane setup.



*some cameras do now have facial recognition software so the camera can start to have some "idea" as to what it is focusing upon; although I've not yet used that tech in a DSLR to know the ins and outs of how it features and functions

 

[junqi]

The cameras AF system works by finding the closest point of contrast difference it can detect in the scene under the AF point(s) that are active.

So if you are using a single AF point only that point will look for a contrast difference underneath itself in the scene. However because its only looking for a difference in contrast it has no idea of the "subject" that its focusing upon*.

In addition the focal plane is rather like a sheet of paper parallel to the front of the lens. When the focus moves this plane moves back and forth until it stops over the point of contrast difference that the AF sensors pick up upon. Anything covered by the paper is then in-focus (the thickness of the paper defined by the aperture and counts as the depth of field - ergo thin depth of field would be a thin sheet - whilst a wide depth of field would be very thick).


In focus and recompose when you focus the focus is locked, its not locked to the subject, but rather the position of the focal plane relative to the lens is locked- that sheet of paper is locked at a certain distance from the lens. Thus if you then move the lens and camera (recompose) whilst the focal plane is fixed there is a potential that whilst you're moving it the plane of focus could be moved off your original target subject.

In addition it sounds like you were hoping you could point at two different subjects and have the camera then focus upon both at the same time; this is not possible because a camera only has one focal plane to work with, so it can't focus on two separate areas. The way you'd cover this is to use a bigger f number (smaller aperture) to give a greater depth of field - ergo make the paper thicker so that the depth of field can cover both subjects. There are also a few other things you can do such as using a tilt shift lens or stacking multiple shots together - but for now stick with the simple idea of a single focal plane setup.



*some cameras do now have facial recognition software so the camera can start to have some "idea" as to what it is focusing upon; although I've not yet used that tech in a DSLR to know the ins and outs of how it features and functions

thanks for the information provided.