I read some info about the 51 point AF system for the new Nikon D300.
The dpreview showed a graphical presentation on the locations of those
51 points within the targeted digital frame (approximately covers 2/3
of the square, centered in the middle of the frame). Could someone
provide a little more detail about this 51 point system?. I may be a
little confused about this concept against the exposure metering in
some old digital cameras.... averaging or spot metering. I guess that
this is not about the light exposures, but instead it is about the
distance measurements from the object(s) to the sensor plate. If the
later is correct, then how do they manage to get a razor sharp picture
if you have to average the distances over 51 spot locations? Or, did
the camera actually take average distances from those 51 locations? I
guess you cannot have 51 different focused distances to the sensor and
each measured distance will be registered into a single 2D plate, can
you? Or, I may miss something completely and more fundamental here? I
think that taking average distances over 51 points will not definitely
work to produce a razor sharp image on the main focus/target area,
unless each of those 51 points have identical distances (or close to
identical distances) to the sensor.
Thanks for discussion and comments.
<aniramca@gmail.com> wrote in message
news:5b818771-9c68-48b0-8b3a-cb3f61760252@v32g2000hsa.googlegroups.com...
>I read some info about the 51 point AF system for the new Nikon D300.
> The dpreview showed a graphical presentation on the locations of those
> 51 points within the targeted digital frame (approximately covers 2/3
> of the square, centered in the middle of the frame). Could someone
> provide a little more detail about this 51 point system?. I may be a
> little confused about this concept against the exposure metering in
> some old digital cameras.... averaging or spot metering. I guess that
> this is not about the light exposures, but instead it is about the
> distance measurements from the object(s) to the sensor plate. If the
> later is correct, then how do they manage to get a razor sharp picture
> if you have to average the distances over 51 spot locations? Or, did
> the camera actually take average distances from those 51 locations? I
> guess you cannot have 51 different focused distances to the sensor and
> each measured distance will be registered into a single 2D plate, can
> you? Or, I may miss something completely and more fundamental here? I
> think that taking average distances over 51 points will not definitely
> work to produce a razor sharp image on the main focus/target area,
> unless each of those 51 points have identical distances (or close to
> identical distances) to the sensor.
> Thanks for discussion and comments.
I can't help with the D300. However, my D70 has several points, but only
one is in
use at any given time. The only connection between AE and AF is that some
cameras use
the distance to the focus point as a guide to set the exposure.
Jim
<aniramca@gmail.com> wrote in message
news:5b818771-9c68-48b0-8b3a-cb3f61760252@v32g2000hsa.googlegroups.com...
>I read some info about the 51 point AF system for the new Nikon D300.
> The dpreview showed a graphical presentation on the locations of those
> 51 points within the targeted digital frame (approximately covers 2/3
> of the square, centered in the middle of the frame). Could someone
> provide a little more detail about this 51 point system?. I may be a
> little confused about this concept against the exposure metering in
> some old digital cameras.... averaging or spot metering. I guess that
> this is not about the light exposures, but instead it is about the
> distance measurements from the object(s) to the sensor plate. If the
> later is correct, then how do they manage to get a razor sharp picture
> if you have to average the distances over 51 spot locations? Or, did
> the camera actually take average distances from those 51 locations? I
> guess you cannot have 51 different focused distances to the sensor and
> each measured distance will be registered into a single 2D plate, can
> you? Or, I may miss something completely and more fundamental here? I
> think that taking average distances over 51 points will not definitely
> work to produce a razor sharp image on the main focus/target area,
> unless each of those 51 points have identical distances (or close to
> identical distances) to the sensor.
> Thanks for discussion and comments.
More likely that the uP reads the various distances, decides which are
the near ones (and thus likely the subject(s)) and adjusts the focus
point for those, possibly adjusting exposure by aperature at the same
time to ensure sufficient depth of field to cover them all.
In article <C3A2C344.47B41%benmicklem@hotmail.com>, Ben Micklem
<benmicklem@hotmail.com> wrote:
> In Live View mode, it will use contrast detection AF on any desired point in
> the frame (i.e. change the focus, see if this increases contrast, if not, go
> the other way, etc.). All other AF modes won't work, as the focus sensors
> are in the bottom of the mirror box, and they get their image from a
> secondary mirror that reflects light from the middle of the frame (that has
> passed through the semi-silvered primary mirror). As the mirror is up during
> live view, this can't work. I think there is an Olympus live view method
> where the live view comes from a second colour CCD in the prism housing, and
> so the mirror is down.
the nikon d3/d300 live view has two modes. in handheld mode, the
mirror will drop and autofocus will use the usual 51 point focus
sensors, then the mirror goes back up to resume live view. in tripod
mode, contrast detection is used off the sensor itself and the mirror
does not drop to focus, with the focus point placed anywhere in the
frame.
Nikon 51 point AF system in D300 
Linear Mode
