Re: dc.watch about K10D exposure errors with K/M lenses


|
| Forum | Pentax SLR Talk |
| Subject | Re: dc.watch about K10D exposure errors with K/M lenses [SIMILAR] |
| Posted by | Jan Moren [PROFILE] [GALLERY] |
| Date/Time | 09:30:14, 05 February 2007 (GMT) |
 |  |
 |
blende8 wrote: > > > URL: > dc.watch.impress.co.jp/cda/longterm/2007/02/02/5509.html > > I would like to read a good translation! > Bablefish just isn't good enough here. > Even A-lenses show errors. > > Could someone please extract the essence for us?
Basically, what they seem to be saying is that as you stop down, light comes from different angles. The focusing screen diffuses that light to some extent, but bright focusing screens diffuse the light less than older screens (that's why they're brighter). So the light that gets to the sensor is not actually directly proportional to the light that will hit the sensor anymore.
The camera needs to compensate for this, and it can do so as long as it knows the actual aperture and focal length but with manual lenses, or lenses off the "A" setting it has no clue, and thus the measurement comes out wrong. In addition, modern lenses tell the camera much more detail about themselves, like where in the light path the aperture is located, which improves precision further. They note that other makers tend to deal with this by disabling metering altogether.
In addition, this seems to be the reason you can't use any other metering mode than center weighted with manual lenses; the off-center sensors get even larger errors than the central one, and basically become useless.
-- Japan: http://www.lucs.lu.se/people/jan.moren/log/current.html Images: http://www.flickr.com/photos/jannem/
|
 |
| Message | |
|  |
| Navigation | |
|
Below is the navigator for this thread, you can use this to view
other messages in this thread. You can use the previous and next buttons to scroll
through the messages in this thread. Or the 'Next New' button to jump to the next
newly posted message.

Shortcut keys:
fForum
1First message
pPrevious
nNext
wNext new
rReply
qQuote