I haven't noticed any problems, but then, I use center weighted and
manual.  I meter just about like I was using an MX.  Pick something
that has a medium tone for metering and set the meter accordingly,
then compose, focus and shoot.

-- 
Bruce


Sunday, February 18, 2007, 8:24:30 AM, you wrote:

PS> I haven't experienced overexposure problems, but I never shoot in
PS> straight program mode. In hyperprogram, TAv, Av and Tv modes, I get
PS> good exposure results. My exposure comp is usually at zero or plus .3.
PS> Paul
PS> On Feb 18, 2007, at 9:42 AM, Dario Bonazza wrote:

>> I find my K10D to be adjuted for slight overexposure, let's say  
>> around +0.3
>> stops, as opposed to the *istD tipically showing around -0.7 to -1
>> underexposure.
>> Maybe the *istD was trying to protect again burnt-out highlights,  
>> while the
>> K10D is mostly trying to fight noise. Anyway, I consider those  
>> above being
>> just different designers' choice, and not a big problem when you  
>> are aware
>> of that.
>>
>> However, very often (10-30% of shots, according to different shooting
>> situations) my K10D gives badly overexposed readings, far  
>> exceeeding that
>> +0.3 reported above. It's not uncommon I get say +2 to +3 stops
>> overexposure, which means pure crap.
>> Not sure, but apparently the bad behavior occurs much more in bright
>> sunlight. The camera was set to P and auto ISO setting, hence no  
>> risk to
>> fall outside hardware capabilities for getting proper exposure. I  
>> tried both
>> multi-pattern and center-balanced metering, with little difference.
>> Maybe
>> multi-pattern is more prone to wild overexposure than center- 
>> balanced, but
>> I'm not yet sure of that.
>>
>> Another K10D owned by a friend of mine does exactly the same. Has  
>> anyone
>> else seen such a behavior?
>>
>> Dario
>>
>>
>> -- 
>> PDML Pentax-Discuss Mail List
>> [email protected]
>> http://pdml.net/mailman/listinfo/pdml_pdml.net





-- 
PDML Pentax-Discuss Mail List
[email protected]
http://pdml.net/mailman/listinfo/pdml_pdml.net

Reply via email to