[jbinpg]

Jim - Re. the performance of the K7 SR, you may want to check out the last post in this thread in this forum: New "silent " K7 firmware

Jack

[ScooterMaxi Jim]

Thanks Jack for that reference. Certainly good news for K-7 owners - SR will be very slightly sharper at speeds near the flash sync. This on the level of 1-2 pixels difference - very small but still important. Makes you wonder if the K-X claimed softness at specific shutter speeds (around 1/125th) is related.

However, the testing at DPreview indicated that SR on the K20D was a solid two stops, nearly twice as good as the results from the K-7. The shutter speeds tested were generally slower than the problem the firmware addresses. The DPreview definition of sharpness and blur involves a large number pixels - probably in the dozens. Of course, testing isn't very precise - it would take a robot to properly replicate conditions.

The only way to know if there really is any difference would be the impressions from those who have used both bodies extensively. I'm not picking on the K-7; I certainly would rather have that body rather than the K20D - but at the time I bought earlier this year I could get twice the factory warranty at half the K-7 price.

[Lowell Goudge]

Thanks Jack for that reference. Certainly good news for K-7 owners - SR will be very slightly sharper at speeds near the flash sync. This on the level of 1-2 pixels difference - very small but still important. Makes you wonder if the K-X claimed softness at specific shutter speeds (around 1/125th) is related.

However, the testing at DPreview indicated that SR on the K20D was a solid two stops, nearly twice as good as the results from the K-7. The shutter speeds tested were generally slower than the problem the firmware addresses. The DPreview definition of sharpness and blur involves a large number pixels - probably in the dozens. Of course, testing isn't very precise - it would take a robot to properly replicate conditions.

The only way to know if there really is any difference would be the impressions from those who have used both bodies extensively. I'm not picking on the K-7; I certainly would rather have that body rather than the K20D - but at the time I bought earlier this year I could get twice the factory warranty at half the K-7 price.

just look at the heron i posted earlier from the K7

1/40th hand held with a 510mm lens I can't do that with the K10D

[blende8]

I have directly compared the K20 with the K7 with respect to metering with K and M lenses. While we all know that the K20 is unacceptable, the K7 is much better in this respect.
Below is an example:

Metering wise it would seem that the K-7 has it but I've got a couple of lenses that don't work so good on my K-7 but do work on my K20D.

[Lowell Goudge]

I have directly compared the K20 with the K7 with respect to metering with K and M lenses. While we all know that the K20 is unacceptable, the K7 is much better in this respect.
Below is an example:

as opposed to metering on a complex scene with a broad range of lighting, a single uniform surface is more appropriate.

[blende8]

Feel free to repeat the test with a grey card.
Anyway, with the K10/K20 you get overexposure of about +2 EV at higher apertures, with the K7 this is much less, but still not perfect.

It would be nice if one could do a calibration curve with a certain lens on the body. The calibration is then stored in the camera.

[Lowell Goudge]

Feel free to repeat the test with a grey card.
Anyway, with the K10/K20 you get overexposure of about +2 EV at higher apertures, with the K7 this is much less, but still not perfect.

It would be nice if one could do a calibration curve with a certain lens on the body. The calibration is then stored in the camera.

I agree, I think the camera (k7) still has a problem at fast apertures but I agree fully that at middle to small apertures the K7 meters much better, a fact displayed by the metering with my 2x TC on my 70-200F2.8 being spot on as opposed to +1.5 stops over exposed.

The problem with a calibration curve is that the camera does not know maximum aperture or minimum aperture, BUT if you could enter maximum aperture I am sure the metering behaves in a consistent manner so they could calculate the preset aperture by the change in exposure, it just would not be as linear as you would expect.

You should propose this to pentax, they haver stopped listening to me on the subject of metering with K mount lenses

[mrpackerguy]

While I love my K200D, I have several M lenses as well. I find green-button metering consistently gives me one stop underexposure most of the time. Bracketed exposure shooting helps. Didn't mean for this to be off topic, but I, too am interested in the K7.

[ScooterMaxi Jim]

The example with Lowell Goudge's graphic is very helpful - not so much relating to the relatively minimal differences between the K-7 and K20D - but as an apparent confirmation that the particular lens has a slightly sticky aperture. Study closely, and you will see some telltale inconsistencies on both bodies. As an aperture heads toward failure, it doesn't perform consistently - and that's what we see here.

Some of us have spent so much longer shooting film (I've shot only about 200,000 digital photos - but many of them with M lenses), these relatively minor differences are interesting but not insurmountable. It is rare that I don't have an opportunity to do test exposures at the particular shutter speed and aperture for my shot; this is a luxury that pretty much renders the exposure discussion into the theoretical - not a practical concern.

If you use an M or K lens, expose using center weighted metering, do a quick test exposure, and get a CLA on your old lenses when they start exhibiting exposure inconsistencies such as the histogram shown - you'll do just fine with the K20D or any other Pentax DSLR. Sorry if I sound like a Grampa saying, "In my day we didn't have it so easy..." but this thread is begging for it.

[GeneV]

I added a split screen to my K10D for just this use. I find it works well but don't have a K-7 to compare metering. I don't see much difference with the K20D in terms of metering although it is a bit better at high ISO's.

But I'm posting because the K10D is OK at 800 and can pull off high ISO shooting if you slightly over expose the shot. Dial in +.3 or +.7 Ev and you'll find the high ISO shots are much better.

If you dial in that exposure compensation, then aren't you really shooting at ISO 600 or so?

[Lowell Goudge]

The example with Lowell Goudge's graphic is very helpful - not so much relating to the relatively minimal differences between the K-7 and K20D - but as an apparent confirmation that the particular lens has a slightly sticky aperture. Study closely, and you will see some telltale inconsistencies on both bodies. As an aperture heads toward failure, it doesn't perform consistently - and that's what we see here.

Some of us have spent so much longer shooting film (I've shot only about 200,000 digital photos - but many of them with M lenses), these relatively minor differences are interesting but not insurmountable. It is rare that I don't have an opportunity to do test exposures at the particular shutter speed and aperture for my shot; this is a luxury that pretty much renders the exposure discussion into the theoretical - not a practical concern.

If you use an M or K lens, expose using center weighted metering, do a quick test exposure, and get a CLA on your old lenses when they start exhibiting exposure inconsistencies such as the histogram shown - you'll do just fine with the K20D or any other Pentax DSLR. Sorry if I sound like a Grampa saying, "In my day we didn't have it so easy..." but this thread is begging for it.

not m graphic, but any way

[blende8]

The example with Lowell Goudge's graphic is very helpful - not so much relating to the relatively minimal differences between the K-7 and K20D - but as an apparent confirmation that the particular lens has a slightly sticky aperture. Study closely, and you will see some telltale inconsistencies on both bodies. As an aperture heads toward failure, it doesn't perform consistently - and that's what we see here.

No, Jim, this is not correct.
You probably do not know this, but we discuss this malbehavior of the K10/K20 for the last 3 years. It happens with all K and M lenses and is definitely not a sticky aperture.
Our idea is that it depends on a non linear transmittance of the focussing screen. The more the aperture is closed a disproportionately small amount of light gets to the metering sensor and results in overexposure.
And these are not "relatively minimal differences" but strong overexposure of about +2 EV. The above histograms show this.

[Lowell Goudge]

No, Jim, this is not correct.
You probably do not know this, but we discuss this malbehavior of the K10/K20 for the last 3 years. It happens with all K and M lenses and is definitely not a sticky aperture.
Our idea is that it depends on a non linear transmittance of the focussing screen. The more the aperture is closed a disproportionately small amount of light gets to the metering sensor and results in overexposure.
And these are not "relatively minimal differences" but strong overexposure of about +2 EV. The above histograms show this.

just to remind every one about the K10D performance with manual lenses

https://www.pentaxforums.com/forums/241716-post69.html

I really need to test my K7 with the same lens. I've been bad had the K7 for 6 months and not benchmarked my lenses yet

[jogiba]

[IMG]ttp://i8.photobucket.com/albums/a21/jogiba/Pentax%20K-7/IMGP2324.jpg[/IMG]