Re: WTF?

Sat, 04 Apr 2009 09:57:06 -0700 

On a second tought...
Some screens are brighter - so if you have a small underexposure at some point (2.8 and brighter) and you change to a screen that's just half that amount brighter than the original, you get a full scale adjustment and the problem wide open is almost solved. Price, you get a small overexposure at the other end. 

Well, I'm rather thick today. Maybe some tequila sunrises solve that ;-)

LF

Luiz Felipe escreveu:

William, that would make sense if the focus screen would change it's 
behavior as you changed the exposure mode or transmitted more light as 
you changed the aperture. I think... shitty week, here.



As I was raised believing, there is no better exposure than the stopped 
down metering just before the actual exposure of the film/ sensor. Those 
days there were some lenses that would not close their apertures 
properly, most of them cheapo samples with poor tolerances.



Issues were delays (lens still closing while the film was being exposed) 
and real aperture not equal to the ring values (either incorrect 
adjustment so the indicated 2.8 was actually closer to 2.0 and the 
difference to the lens wide open would be less than it should be, or low 
quality lenses promising fast apertures they didn't really have - so 
again the difference from the 2.8 to the real 2.0 wide open would be 
inferior to the promised but not delivered 1.7). But those issues would 
be prevented by stopped down metering just before the exposure - and 
later by real time metering from the film as in the Lx and some OM 
Olympus cameras.



Facing over or underexposure with open lens metering I would blame the 
wrong aperture as closed or the wrong assumption of the maximum aperture 
of the lens. I can't think of focusing screens with active, light level 
adjustable transmission index.


LF

William Robb escreveu:

In response to this question:


"I was informed that the camera is a little bit under-exposed, but I 
want to

know how much. That is, if I set the EV compensation, but how much (1/3?
1/2?) should I set so that I can get normal exposure?"

This answer was given:

"Well, regarding the exposure problem, let's say it's related to the lens
used.
For f/3.5-5.6 lens, exposure will be quite good.
But with fast lenses, you'll have a variable amount of underexposure...
- f/1.4 needs about +1 1/2ev
- f/2.8 needs about +1ev

and so on...


This is due to the fact that the focus screen does not respond in a 
linear

way to the aperture of the lens...

You should have a direct relation between shutter speed and aperture 
(i.e.

going from f/1.4 to f/2 should halve the shutter speed, going from say
1/1000 to 1/500).
But with the original k20 focus screen, this relation is not respected
(going from f/1.4 to f/2 won't halve the speed).

Note : I'm talking about using actual, stop-down apertures.

For A lenses, the exposure is mesured wide-open, then calculated for the

given aperture you want to use, so selecting f/2 will always gives 
half the

speed of f/1.4.
BUT the wide-open reading will be underexposed by the 1 1/2ev mentionned

above, so the lens will always underexpose, whatever aperture you're 
using.



Using an old lens (with aperture blades actually moving when you 
select an

aperture on the lens), you'll see that closing one stop won't halve the
speed.

You can lessen this behavior by using a LL60 screen (from the *ist line).

They are fully compatible with the k10/k20, and lessen the 
discrepancies to
the point where you won't really need to compensate the exposure 
anymore."


--------------------------------------------


I'm not seeing where the answer could possibly be correct, but I 
thought I'd

post the whole thing here, hoping to generate some discussion.

This comes from that forum thang, where I managed to get a 2 point 
demerit yesterday for telling a guy to fuck off.

Godders, I must just have been having a shitty day yesterday...

William Robb



--
PDML Pentax-Discuss Mail List
PDML@pdml.net
http://pdml.net/mailman/listinfo/pdml_pdml.net

to UNSUBSCRIBE from the PDML, please visit the link directly above and 
follow the directions.







--
Luiz Felipe
luiz.felipe at techmit.com.br
http://techmit.com.br/luizfelipe/

--
PDML Pentax-Discuss Mail List
PDML@pdml.net
http://pdml.net/mailman/listinfo/pdml_pdml.net
to UNSUBSCRIBE from the PDML, please visit the link directly above and follow 
the directions.






















					Reply via email to