[Beau]

Currently with the K10D, and I presume the K20D and K-7, when you are using a manual lens you'll get "F --" and have to use stop-down metering in the M mode.

When the camera does the stop-down metering it could calculate the ratio between wide open and stopped down and thus measure the effective F-stop.

It should therefore be possible to put the camera in Av mode and adjust the shutter speed according to the measured F-stop, only needing a new stop-down metering when you change the aperture on the lens.

It doesn't work like this so maybe I'm missing something?

Cheers,
Beau

[r3tic]

When you have the camera in AV or TV it needs to know the aperture setting to make the necessary calculations. The camera has no way of knowing the aperture value on a manual lens unless you are in M and step the lens down.

[blende8]

When the camera does the stop-down metering it could calculate the ratio between wide open and stopped down and thus measure the effective F-stop.

What you get from this is only the difference between the two stops, not the absolute values.
The difference between f2 and f4 is the same as between f2.8 and f5.6.

One could argue then that one only needs to enter the fully open aperture once you mount the lens.

But this, too, does not work, because in K and M lenses the aperture lever does not move in linear steps but squared.
Since the Pentax DSLRs have no aperture coupler, there is no way to overcome this problem.

Once more I say:
Pentax, give us back the stop-down coupler!

[Lowell Goudge]

Currently with the K10D, and I presume the K20D and K-7, when you are using a manual lens you'll get "F --" and have to use stop-down metering in the M mode.

When the camera does the stop-down metering it could calculate the ratio between wide open and stopped down and thus measure the effective F-stop.

It should therefore be possible to put the camera in Av mode and adjust the shutter speed according to the measured F-stop, only needing a new stop-down metering when you change the aperture on the lens.

It doesn't work like this so maybe I'm missing something?

Cheers,
Beau

the idea is right in concept but wrong in execution for a couple of reasons.

first of all, in all but manual mode the K10D and (I assume) the others, will not operate the aperture lever. the lens shoots wide open all the time.

if you were to add the function you want, the idea would be to have the camera meter when you press the green button without stopping dowe, then stop the lens down meter again, and calculate the change in aperture to know how many stops, and select the shutter accordingly.

You could go one simpler, and add a metering step after stopping down, but before moving the mirror, and set the shutter by this. (but shutter lag will increase since stopping down and moving the mirror are probably done in parallel at present)

In theory it would work, it practice it can't because the K10D and K20D metering are not linear, and need to know maximum aperture to meter correctly.

What would be better, would be to input the max and min aperture and let the camera control the aperture just like with new lenses (except the relatrionship between aperture and lever position is not the same as A series lenses)

[Beau]

What you get from this is only the difference between the two stops, not the absolute values.
The difference between f2 and f4 is the same as between f2.8 and f5.6.

No. You'll get 2 values. Wide open you measure X, stopped down you get Y.
X-Y gives the difference, X/Y the ratio.

It's true you'll need to know the max aperture to calculate the F-stop from this ratio.
But you don't need to know the exact F-stop to calculate the correct shutter speed.

Suppose you measure 4 times less light stopped down. Wouldn't setting the shutter 4 times slower then the shutter speed you would get wide open in Av give the correct exposure?

the idea is right in concept but wrong in execution for a couple of reasons.

first of all, in all but manual mode the K10D and (I assume) the others, will not operate the aperture lever. the lens shoots wide open all the time.

if you were to add the function you want, the idea would be to have the camera meter when you press the green button without stopping dowe, then stop the lens down meter again, and calculate the change in aperture to know how many stops, and select the shutter accordingly.

You could go one simpler, and add a metering step after stopping down, but before moving the mirror, and set the shutter by this. (but shutter lag will increase since stopping down and moving the mirror are probably done in parallel at present)

That's more or less what I meant. Your press the green mode to do a stop-down measure. You wouldn't need to do a normal measure with the green button, this is done continuously and shutter speed would be displayed and updated like you're used to in Av mode, but adjusted according to the stop down metering.

In theory it would work, it practice it can't because the K10D and K20D metering are not linear, and need to know maximum aperture to meter correctly.

What would be better, would be to input the max and min aperture and let the camera control the aperture just like with new lenses (except the relatrionship between aperture and lever position is not the same as A series lenses)

My question comes down to this: how come you can't calculate a correct exposure from the continuous wide-open metering when you know what the effect of stopping down will be?

[stevebrot]

My question comes down to this: how come you can't calculate a correct exposure from the continuous wide-open metering when you know what the effect of stopping down will be?

Perhaps, because the camera does not know that the lens is wide open? It also does not know what the maximum aperture is.

There is also an additional consideration that Lowell touched on briefly. The movement of the aperture actuator lever on "A" type lenses is proportional to the aperture opening. This allows aperture control from the body. On K and M lenses, the assumption is that the body is always going to flip the lever all the way with the actual aperture being determined by the mechanical "stop" inside the lens.

This whole business of backwards compatibility is actually pretty complex and we are pretty lucky that it all works as well as it does. About the only thing that might have made things better is if Pentax had not "crippled" the mount to remove the mechanical aperture coupling.

Steve

[chhayanat]

...Since the Pentax DSLRs have no aperture coupler, there is no way to overcome this problem.

Once more I say:
Pentax, give us back the stop-down coupler!

Amen. Or at least a window in the prism housing to enable the aperture setting on the lens to be read: it's difficult to see the setting even looking over the prism at the aperture ring owing to the overhang in the prism or pentamirror housing.

[Lowell Goudge]

No. You'll get 2 values. Wide open you measure X, stopped down you get Y.
X-Y gives the difference, X/Y the ratio.

It's true you'll need to know the max aperture to calculate the F-stop from this ratio.
But you don't need to know the exact F-stop to calculate the correct shutter speed.

correct

Suppose you measure 4 times less light stopped down. Wouldn't setting the shutter 4 times slower then the shutter speed you would get wide open in Av give the correct exposure?

assuming the metering is accurate (like the *istD for example but more on that later)

That's more or less what I meant. Your press the green mode to do a stop-down measure. You wouldn't need to do a normal measure with the green button, this is done continuously and shutter speed would be displayed and updated like you're used to in Av mode, but adjusted according to the stop down metering.

you only need to meter stopped down, so that the exposure can be taken, unless you really want it displayed

My question comes down to this: how come you can't calculate a correct exposure from the continuous wide-open metering when you know what the effect of stopping down will be?

I wish it were that simple, and on the *istD it would have been, but there is an apparent issue with the K10D and K20D that causes exposure errors with non A lenses.

I documented this problem int he following post

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

I believe pentax uses an error curve, similar to what I measured, and knowledge of the maximum aperture of the lens, to calculate correct exposure. As a result while in theory knowing only the difference matters, in practice knowing the absolute aperture is critical/


now this is only speculation on my part, but I have confirmed part of this by taking a fast "A" lens out of A.

[Beau]

correct assuming the metering is accurate (like the *istD for example but more on that later)you only need to meter stopped down, so that the exposure can be taken, unless you really want it displayed

Ignoring the problems you mention further down your post, my setup would work like this:

I choose some aperture using the aperture ring and I do a stop down meter with the green button. This gives the camera the ratio between wide open and stopped down. This would be the only time you do an explicit metering.

As soon as the lighting conditions change, the camera would calculate a new shutter speed (using the measured ratio and the implicit continuous wide open metering), just like with an A lens in Av mode.

You wouldn't notice you're using an M lens till you change the aperture and have to do a new stop down metering to get the new ratio.

The (strange) non-linearity in your post is of course a problem but I would gladly dial down the max aperture together with the focal length when putting on a new lens.

It wouldn't be as good as a mechanical aperture coupler but it would be a big improvement over the current implementation IMHO and could all be done in firmware.

[Lowell Goudge]

Ignoring the problems you mention further down your post, my setup would work like this:

I choose some aperture using the aperture ring and I do a stop down meter with the green button. This gives the camera the ratio between wide open and stopped down. This would be the only time you do an explicit metering.

As soon as the lighting conditions change, the camera would calculate a new shutter speed (using the measured ratio and the implicit continuous wide open metering), just like with an A lens in Av mode.

You wouldn't notice you're using an M lens till you change the aperture and have to do a new stop down metering to get the new ratio.

The (strange) non-linearity in your post is of course a problem but I would gladly dial down the max aperture together with the focal length when putting on a new lens.

It wouldn't be as good as a mechanical aperture coupler but it would be a big improvement over the current implementation IMHO and could all be done in firmware.

I don't disagree, I gave pentax all sorts of options including one like this a long time ago, and the ultimate response back was that (for the K10D) they did not perceive that the metering issue was as bad as it was until production had ended.

As the K20 uses the same metering and has the same problems, I don;t know what reason thay gave K20 owners, but that was part of my reason for not upgrading. I didn't see the extra MP count as necessary for me, and unless the camera metered better I didn't see the point.

I am now re-thinking that based on high ISO performance, but also looking at the K7. Not sure what side of the fence I will fall off onto.

K7 looks interesting, and has some better performance (FPS, max shutter and image processing speed) but otherwise it almost seems like an *istDS compared to the *istD. Pentax made 7or 8 new 6 MP bodies before it stepped up to the K10D and the next "flagship product"

[ManWithCamera]

If you shoot the lens wide open, you can use Av without a problem. If you close down by a stop, use exposure compensation to tell the camera to over expose by a stop. If you close down two stop, dial in two stops of over exposure using the exposure compensation.

In my experience, it hasn't actually been this straight forward. I don't usually need to use as much exposure compensation as this, but with a little experimentation, you should be able to figure out how much is necessary.

[Lowell Goudge]

If you shoot the lens wide open, you can use Av without a problem. If you close down by a stop, use exposure compensation to tell the camera to over expose by a stop. If you close down two stop, dial in two stops of over exposure using the exposure compensation.

In my experience, it hasn't actually been this straight forward. I don't usually need to use as much exposure compensation as this, but with a little experimentation, you should be able to figure out how much is necessary.

you're right it is not that straight forward you can't stop down the lens in Av mode, the camera does not move the aperture activation lever, the only shot you get is wide open.

second, see my post above, the K10D and K20D don't meter correctly except at between F4 and F5.6. at all other apertures it has big errors, sometimes up to 2 stops.

[Marc Sabatella]

Even if the proposed idea works (and my naive notion, without thinking about it, is that it should, although I don't understand the linearity issue and maybe that really would be an obstacle), it would still be rather trouble-prone, because the camera wouldn't notice if you *change* apertures after metering. So the onus would be on you not to change apertures without remetering. Which really makes the proposed Av mdoe not that much better than M mode. OK, sure, at least it accounts for lighting changes, but it's still a matter of needing to Green button every time you change apertures.

[code4code5]

Try this... With your M lens set to minimum aperture, mount the lens to the body as you are looking through the front element. As you twist the lens, you will see the blades opening slightly. Back off until the blades are closed. With the lens mounted in this fashion, you can change aperture in Av mode and the camera will instantly calculate the correct shutter speed, but some EV compensation may be needed, to be determined with experience.

Sure, your lens isn't securely mounted and you might drop it, but you have an instant Av and M meter that doesn't require a green button press when changing aperture. It's a functional "work-around" with the equipment we have.

[Steve Beswick]

LOL - I call this technique "K1000 D.O.F. Preview"

Try this... With your M lens set to minimum aperture, mount the lens to the body as you are looking through the front element. As you twist the lens, you will see the blades opening slightly. Back off until the blades are closed. With the lens mounted in this fashion, you can change aperture in Av mode and the camera will instantly calculate the correct shutter speed, but some EV compensation may be needed, to be determined with experience.

Sure, your lens isn't securely mounted and you might drop it, but you have an instant Av and M meter that doesn't require a green button press when changing aperture. It's a functional "work-around" with the equipment we have.