Forgot Password
Pentax Camera Forums Home
 

Reply
Show Printable Version Search this Thread
10-13-2008, 09:24 AM   #16
Veteran Member
jgredline's Avatar

Join Date: Oct 2007
Location: LosAngeles, Ca.
Photos: Albums
Posts: 10,628
QuoteOriginally posted by stevebrot Quote
As noted above, the K10D (and K20D for that matter) have well-documented metering issues with M-42 and non-A K-mount lenses. I have heard all sorts of reasoning on this matter as to causes. The focus screen has been singled out, but there have been posts that indicate no improvement with replacement screen. Pentax has been pretty mum about workarounds.

Here are my observations as the owner of a K10D and a few non-A lenses:
  • There is no general bias to apply to exposure
  • The pattern varies according to lens. Some overexpose at larger apertures, others at smaller
  • Most will meter correctly at f/5.6-f/8.0

My general approach is to either
Meter at f/5.6 and manually calculate shutter speed for the desired aperture.
-- OR --
Do the green button at the desired aperture and chimp the histogram to determine the best shutter speed.
Steve

BTW...The accuracy at f/5.6 noted above is a generality. My Zenitar 16/2.8 with green-button overexposes 2-3 stops regardless of aperture.
Thanks Steve, I too have seen what you described.

10-13-2008, 09:43 AM   #17
Senior Member




Join Date: May 2008
Location: stockholm
Photos: Albums
Posts: 111
Could it be that you only get center weight or spot metering on M lenses?? I just have the feeling I read it somewhere. On some Pentax film bodies this was the case anyway...
10-13-2008, 09:46 AM   #18
Loyal Site Supporter
Loyal Site Supporter
Canada_Rockies's Avatar

Join Date: Nov 2007
Location: Sparwood, BC, Canada
Photos: Gallery | Albums
Posts: 12,385
QuoteOriginally posted by debbie Quote
Could it be that you only get center weight or spot metering on M lenses?? I just have the feeling I read it somewhere. On some Pentax film bodies this was the case anyway...
Any lens without an A setting, or any lens set to an aperture number other than the A setting has the problem. Even FA lenses with aperture rings, when the ring is off the A setting, will display the same problems.
10-13-2008, 10:16 AM   #19
Otis Memorial Pentaxian
stevebrot's Avatar

Join Date: Mar 2007
Location: Vancouver (USA)
Photos: Gallery | Albums
Posts: 42,007
QuoteOriginally posted by debbie Quote
Could it be that you only get center weight or spot metering on M lenses?? I just have the feeling I read it somewhere. On some Pentax film bodies this was the case anyway...
Yes...

Steve

10-13-2008, 11:09 AM   #20
Veteran Member
Big Dave's Avatar

Join Date: Jan 2008
Location: San Antonio, Texas
Photos: Gallery
Posts: 547
M42 lenses work great in AV mode, because you have exposure compensation available to correct the exposure.

Dave

QuoteOriginally posted by dlacouture Quote
Unfortunately, no, m42s also suffer from the same problem : underexposure above f/5.6, overexposure below f/8...

I'll try the LL60 in a little while, just to see how it handles things...

For those interested, I tried a split-screen from jinfinance (chinese vendor on ebay), and it shows the same behavior as the original screen, but with a slight +2/3ev overexposure (which is not so bad as it sort of solves my problem of -1ev underexposure with my fast manual lenses)...
10-13-2008, 11:29 AM   #21
Veteran Member
Marc Sabatella's Avatar

Join Date: Sep 2006
Location: Denver, CO
Photos: Gallery
Posts: 10,685
QuoteOriginally posted by debbie Quote
Could it be that you only get center weight or spot metering on M lenses??
This is true, and it definitely explains why exposure is different on an M lens than a more modern lens which viewing the same scene. If you otherwise use multi-segment - I have switched to center-weighted all the time, primarily to get more consistency.

But in any case, that doesn't explain why a given M lens would produce different exposure at different apertures. That is apparently mostly an issue of some sort with the focus screen. Although I do also suspect there could be a timing issue at play that might cause the DOF preview to be more accurate than the green button - but still be subject to whatever problems are caused by the focus screen.

For the record, this issue seems unique to the K10D, although I hear mixed things about the K20D. But the DS, DL, K100D, and K200D seem to be immune.
10-13-2008, 06:13 PM   #22
Site Supporter
Site Supporter
Lowell Goudge's Avatar

Join Date: Jan 2007
Location: Toronto
Photos: Gallery | Albums
Posts: 17,869
QuoteOriginally posted by Marc Sabatella Quote
This is true, and it definitely explains why exposure is different on an M lens than a more modern lens which viewing the same scene. If you otherwise use multi-segment - I have switched to center-weighted all the time, primarily to get more consistency.

But in any case, that doesn't explain why a given M lens would produce different exposure at different apertures. That is apparently mostly an issue of some sort with the focus screen. Although I do also suspect there could be a timing issue at play that might cause the DOF preview to be more accurate than the green button - but still be subject to whatever problems are caused by the focus screen.

For the record, this issue seems unique to the K10D, although I hear mixed things about the K20D. But the DS, DL, K100D, and K200D seem to be immune.
Its funny you mention the DOF vs green button. In one of my many talks with pentax they indicated that the DOPF preview might be more reliable, and I also reported that I noticed the time for stop down when metering was much longer on my *istD than K10D. Funny thing, next software the lens stopped down more with the K10D than previously, but this did not help metering.

As far as the focusing screen issue. I changed out mine and put in the *istD screen, and the metering worked just as it did on the *istD.

Also, if you take an A lens out of A the metering is just as bad as an old K lens.

As for the K20D, I have heard it has exactly the same screen as the K10D, therefore unless they protected the sensor from scattered light off the focusing screen, I would suspect a similar problem/

10-14-2008, 01:46 PM   #23
Otis Memorial Pentaxian
stevebrot's Avatar

Join Date: Mar 2007
Location: Vancouver (USA)
Photos: Gallery | Albums
Posts: 42,007
QuoteOriginally posted by Marc Sabatella Quote
.

..That is apparently mostly an issue of some sort with the focus screen. Although I do also suspect there could be a timing issue at play that might cause the DOF preview to be more accurate than the green button - but still be subject to whatever problems are caused by the focus screen.
I have heard the focus screen theory tossed around quite a bit but have some doubts based on the mixed results some folks have reported. I would like to propose another theory. I was told by my local camera repairman that the meters in older cameras (my Ricoh XR7 to be specific) require multiple trim pots to allow correction for non-linear response from the SPD detector. It my understanding that this function is accomplished in modern cameras using mathematical splines based on the reported aperture from the lens (think "crippled" mount...). Much cheaper to build and easier to modify between model. Lacking this information in stop-down mode, the user is at the mercy of the detector raw response or at least a more primitive set of correction factors. The observed difference between the Kx00 models and the Kx0 models can be accounted for by the different detectors and metering systems between the two lines.

Just a thought...

Steve

BTW...the timing hypothesis is not too far afield either except that the problem still exists with manual aperture lenses (like my J-9). However, I suspect that slow iris response when using green button metering may be yet another factor in this big mess.
10-14-2008, 02:52 PM   #24
Site Supporter
Site Supporter
Lowell Goudge's Avatar

Join Date: Jan 2007
Location: Toronto
Photos: Gallery | Albums
Posts: 17,869
QuoteOriginally posted by stevebrot Quote
I have heard the focus screen theory tossed around quite a bit but have some doubts based on the mixed results some folks have reported. I would like to propose another theory. I was told by my local camera repairman that the meters in older cameras (my Ricoh XR7 to be specific) require multiple trim pots to allow correction for non-linear response from the SPD detector. It my understanding that this function is accomplished in modern cameras using mathematical splines based on the reported aperture from the lens (think "crippled" mount...). Much cheaper to build and easier to modify between model. Lacking this information in stop-down mode, the user is at the mercy of the detector raw response or at least a more primitive set of correction factors. The observed difference between the Kx00 models and the Kx0 models can be accounted for by the different detectors and metering systems between the two lines.

Just a thought...

Steve

BTW...the timing hypothesis is not too far afield either except that the problem still exists with manual aperture lenses (like my J-9). However, I suspect that slow iris response when using green button metering may be yet another factor in this big mess.
Steve

Not to throw your theory out but I actually did change focusing screen between an *istD and K10D and found that with the *istD screen, the K10D metered exactly as the *istD metered using the same lens both times.

Also, Taking any new lens out of A results in the same metering problem.

As a result, I am inclined to believe the screen is the total issue.

As for trimming the metering for non linearity, that is one thing but once set it is constant, which is why old cameras work, what is a problem in the K10D is that the metering changes non linearily as a function of total light, but as a function of F Stop. For example, I can get teh same curve that I plotted, using 100 ISO and bright day light or 1600 ISO and very dim lighting. if the sensor was non linear with light in totality, it would not meter the same way over such a wide range of absolute light.

The fact that it meters differently as a function of F Stop, and that this is consistent, (i.e. the same behavior and type of error) with a wide variety of lenses suggests the screen is the total problem.

Slo Iris is a different issue, and I believe pentax addressed this point, as I reported between V1.2 and 1.3 because the stop down time of the green button changed.
10-14-2008, 03:25 PM   #25
Otis Memorial Pentaxian
stevebrot's Avatar

Join Date: Mar 2007
Location: Vancouver (USA)
Photos: Gallery | Albums
Posts: 42,007
QuoteOriginally posted by Lowell Goudge Quote

...I actually did change focusing screen between an *istD and K10D and found that with the *istD screen, the K10D metered exactly as the *istD metered using the same lens both times...
Glad to hear that you had good success using that method. I don't have access to a *istD screen and can't comment. I have read reports of people having good luck with the Katz Eye screen, though other people say that changing to that screen had no affect.

QuoteOriginally posted by Lowell Goudge Quote

...The fact that it meters differently as a function of F Stop, and that this is consistent, (i.e. the same behavior and type of error) with a wide variety of lenses suggests the screen is the total problem...
This has not been my experience. The type and degree of over/under exposure is related to aperture, but seems to be lens-specific. Case in point being my Tamron 70-150/3.5 with PK-M mount. That lens meters fairly accurately at all apertures using the green button. My Pentax-M 200/4 meters accurately wide open, but overexposes at f/16-f/32. My Pentax-M 50/1.7 underexposes about 1 stop wide open, while my Jupiter-9/2 underexposes 2-3 stops at f/2. Both the 50 and the J-9 are accurate at apertures f/5.6 and narrower. The Tamron 28/2.5 and Rikenon 50/1.7 behave similar to the M 50/1.7. The Zenitar 16/2.8 overexposes 2-3 stops at all apertures.

My intuition is that this is not a simple issue with a single cause. I suspect that numerous compromises were introduced to support the matrix metering and that good support for center-weighted/spot meter in stop-down mode was not deemed cost-effective and was left out of the development effort. There is no profit in supporting sales in the used lens market!

So, when I explain this issue to noobies I just say the reasons are poorly understood, cuz the fact is that I don't know, Pentax is not saying, and to the best of knowledge there is no readily available hardware/software fix.

Steve
10-14-2008, 04:33 PM   #26
Site Supporter
Site Supporter
Lowell Goudge's Avatar

Join Date: Jan 2007
Location: Toronto
Photos: Gallery | Albums
Posts: 17,869
QuoteOriginally posted by stevebrot Quote
Glad to hear that you had good success using that method. I don't have access to a *istD screen and can't comment. I have read reports of people having good luck with the Katz Eye screen, though other people say that changing to that screen had no affect.



This has not been my experience. The type and degree of over/under exposure is related to aperture, but seems to be lens-specific. Case in point being my Tamron 70-150/3.5 with PK-M mount. That lens meters fairly accurately at all apertures using the green button. My Pentax-M 200/4 meters accurately wide open, but overexposes at f/16-f/32. My Pentax-M 50/1.7 underexposes about 1 stop wide open, while my Jupiter-9/2 underexposes 2-3 stops at f/2. Both the 50 and the J-9 are accurate at apertures f/5.6 and narrower. The Tamron 28/2.5 and Rikenon 50/1.7 behave similar to the M 50/1.7. The Zenitar 16/2.8 overexposes 2-3 stops at all apertures.

My intuition is that this is not a simple issue with a single cause. I suspect that numerous compromises were introduced to support the matrix metering and that good support for center-weighted/spot meter in stop-down mode was not deemed cost-effective and was left out of the development effort. There is no profit in supporting sales in the used lens market!

So, when I explain this issue to noobies I just say the reasons are poorly understood, cuz the fact is that I don't know, Pentax is not saying, and to the best of knowledge there is no readily available hardware/software fix.

Steve
Steve

I would be willing to take a set of shots, from each of your lenses made over the entire range of apature clicks, and plot the results.

The shots should be as follows.

Wide open, adjust ISO to get between 1/2000 and 1/3000, then start shooting, apature click by apature click until minimum apature is reached. metering with green button for all shots. Record each apature click (actual F stop) and use 1/2 stop in camera steps as it matches closest to the lens.

I take histogram from central ~10% of screen.

What I ahve noted is the following.

fast lenses under expose wide open by between 1 and 1.5 stops, but expose correctly by F4, then over expose by up to 2 stops by F8 settling back to 1.5 stops by F22 or F32.

there may be some variation between different lenses, and if so, I suspect that it is a function of how close the lens rear element is to the sensor/mirror, and how accute the angle of light is at the focusing screen, as I believe it is the off axis scatter
10-14-2008, 10:35 PM   #27
Otis Memorial Pentaxian
stevebrot's Avatar

Join Date: Mar 2007
Location: Vancouver (USA)
Photos: Gallery | Albums
Posts: 42,007
QuoteOriginally posted by Lowell Goudge Quote
Steve

I would be willing to take a set of shots, from each of your lenses made over the entire range of apature clicks, and plot the results.

The shots should be as follows.

Wide open, adjust ISO to get between 1/2000 and 1/3000, then start shooting, apature click by apature click until minimum apature is reached. metering with green button for all shots. Record each apature click (actual F stop) and use 1/2 stop in camera steps as it matches closest to the lens.

I take histogram from central ~10% of screen.

What I ahve noted is the following.

fast lenses under expose wide open by between 1 and 1.5 stops, but expose correctly by F4, then over expose by up to 2 stops by F8 settling back to 1.5 stops by F22 or F32.

there may be some variation between different lenses, and if so, I suspect that it is a function of how close the lens rear element is to the sensor/mirror, and how accute the angle of light is at the focusing screen, as I believe it is the off axis scatter
Thanks Lowell. That type of systematic approach is much appreciated. I did do exposure series, but did not actually record and plot the values. I just eye-balled the series much as I would a proof strip in the darkroom. Your comments regarding rear element position, light angle and off-axis scatter are also good considerations.

The thing that puzzles me is why A-mount lenses in the A-setting don't have problems while the same lens using the aperture ring does. Why, for an f/2 lens, would an f/2 aperture setting meter differently when set by the body than at f/2 using the aperture ring (delta = 1-1.5 stops)? Both are metered with the lens wide open so the optical environments should be the same. The only difference is that the body "knows" it is metering with the lens at f/2 when on the A-setting. The only conclusion I can come to is that the meter reading is "interpreted" and adjusted to fit the meter/detector characteristics.

On a tangentially related note...If you really want to muddy the waters. Do an exposure series on a subject (say a still-life with controlled lighting) using an A-mount lens on a 35mm film SLR at ISO 200 (AV mode, your XR2s will do nicely) and repeat using the same lens on a K10D on the same ISO setting, AV mode, A-setting, using matrix, center-weighted, and spot. You don't actually have to do the exposures, just record the settings. Then repeat with a gray card (fill the frame) under the same light.

How about the "sunny 16" rule on the K10D?

Clearly, Toto, we are not in Kansas any more...

Steve

(Hint regarding the last three paragraphs...apparently ISO 200 means something different to the K10D than it does to film...except, perhaps, in regards to rendering 18% gray...)

(Another hint...there is a Wikipedia article on ASA/ISO where "digital" ISO is discussed.)

Last edited by stevebrot; 10-14-2008 at 10:45 PM.
10-15-2008, 04:06 AM   #28
Site Supporter
Site Supporter
Lowell Goudge's Avatar

Join Date: Jan 2007
Location: Toronto
Photos: Gallery | Albums
Posts: 17,869
Thanks steve

QuoteOriginally posted by stevebrot Quote
That type of systematic approach is much appreciated. I did do exposure series, but did not actually record and plot the values. I just eye-balled the series much as I would a proof strip in the darkroom. Your comments regarding rear element position, light angle and off-axis scatter are also good considerations.
I consider the systematic approach part of knowing my equipment. when you eyeball things you cant measure as accurately as the meter, which makes a problem
QuoteQuote:
The thing that puzzles me is why A-mount lenses in the A-setting don't have problems while the same lens using the aperture ring does. Why, for an f/2 lens, would an f/2 aperture setting meter differently when set by the body than at f/2 using the aperture ring (delta = 1-1.5 stops)? Both are metered with the lens wide open so the optical environments should be the same. The only difference is that the body "knows" it is metering with the lens at f/2 when on the A-setting. The only conclusion I can come to is that the meter reading is "interpreted" and adjusted to fit the meter/detector characteristics.
I came to the same conslusion. specifically pentax must have compensated in software for the error as all metering is done at maximum apature, and performance is relitively consistent (by this I mean the error is well defined) lens to lens at maximum.
QuoteQuote:

On a tangentially related note...If you really want to muddy the waters. Do an exposure series on a subject (say a still-life with controlled lighting) using an A-mount lens on a 35mm film SLR at ISO 200 (AV mode, your XR2s will do nicely) and repeat using the same lens on a K10D on the same ISO setting, AV mode, A-setting, using matrix, center-weighted, and spot. You don't actually have to do the exposures, just record the settings. Then repeat with a gray card (fill the frame) under the same light.
the only issue is the shutter speed, the XR2s is infinately vatiable in Apature priority mode. How do you accurately read it? It would have been better if the meter was in +/- ev for this test. Perhaps a newer body like a MX, which as I recall gave +/- ev but that is why I compared to the *istD. It meters very well indeed with manual apature lenses and the results are digital and can be evaluated the same way.
QuoteQuote:
How about the "sunny 16" rule on the K10D?

Clearly, Toto, we are not in Kansas any more...

Steve
actually steve, I think the expression should be "beam me up scotty"
QuoteQuote:
(Hint regarding the last three paragraphs...apparently ISO 200 means something different to the K10D than it does to film...except, perhaps, in regards to rendering 18% gray...)

(Another hint...there is a Wikipedia article on ASA/ISO where "digital" ISO is discussed.)
there has been lots of discussion about this and I am not sure where it is headed so I won't add to it
10-15-2008, 08:51 AM   #29
Otis Memorial Pentaxian
stevebrot's Avatar

Join Date: Mar 2007
Location: Vancouver (USA)
Photos: Gallery | Albums
Posts: 42,007
QuoteOriginally posted by Lowell Goudge Quote

there has been lots of discussion about this and I am not sure where it is headed so I won't add to it
I'm done too...next case...
10-15-2008, 09:37 AM   #30
Loyal Site Supporter
Loyal Site Supporter
Canada_Rockies's Avatar

Join Date: Nov 2007
Location: Sparwood, BC, Canada
Photos: Gallery | Albums
Posts: 12,385
Sunny 16 works.

QuoteOriginally posted by stevebrot Quote
On a tangentially related note...If you really want to muddy the waters. Do an exposure series on a subject (say a still-life with controlled lighting) using an A-mount lens on a 35mm film SLR at ISO 200 (AV mode, your XR2s will do nicely) and repeat using the same lens on a K10D on the same ISO setting, AV mode, A-setting, using matrix, center-weighted, and spot. You don't actually have to do the exposures, just record the settings. Then repeat with a gray card (fill the frame) under the same light.

How about the "sunny 16" rule on the K10D?
I did a series using sunny 16 with my M 400. It works. My tests indicated that the problem is with the metering, not with the mechanicals, which is what I was checking. If the problem was with the mchanicals, I would have been terribly disappointed. The metering I can work around if I have to. You can see the test shots on my Flickr site, if you want to have a peek.
Reply

Bookmarks
  • Submit Thread to Facebook Facebook
  • Submit Thread to Twitter Twitter
  • Submit Thread to Digg Digg
Tags - Make this thread easier to find by adding keywords to it!
aperture, camera, dslr, lens, lenses, meter, photography, step
Thread Tools Search this Thread
Search this Thread:

Advanced Search


Similar Threads
Thread Thread Starter Forum Replies Last Post
Manual Metering on Pentax DSLRs with Manual and Automatic Lenses Adam Pentax DSLR and Camera Articles 20 01-06-2016 03:38 AM
accuracy of K-x metering manual / M42 lenses? kuuan Pentax DSLR Discussion 7 01-22-2010 01:01 PM
K-7 LV metering with manual lenses Jadran Pentax DSLR Discussion 6 01-12-2010 12:57 PM
Metering with Manual lenses.. arbib Pentax SLR Lens Discussion 21 01-03-2010 10:22 PM
K10D, manual lenses, and metering vinzer Pentax DSLR Discussion 40 06-30-2007 07:54 AM



All times are GMT -7. The time now is 09:01 AM. | See also: NikonForums.com, CanonForums.com part of our network of photo forums!
  • Red (Default)
  • Green
  • Gray
  • Dark
  • Dark Yellow
  • Dark Blue
  • Old Red
  • Old Green
  • Old Gray
  • Dial-Up Style
Hello! It's great to see you back on the forum! Have you considered joining the community?
register
Creating a FREE ACCOUNT takes under a minute, removes ads, and lets you post! [Dismiss]
Top