[stevebrot]

I feel we should be able to turn it off

I agree, though I would caution that in truly dim conditions live view is essentially unusable due to noise. I just did a little look-see in the laundry room and found that noise, not brightness is the limiting factor for dim light manual focus in live view.


Steve

[Volker76]

ok, I reproduced it with my DA 50mm f1.8. It is a very minor dimming, easy to miss.

BUT: it also screws around with autofocus!

Started at f1.8 focused my computer monitor.
Scrolled up to f11 and observed the dimming effect.
Scrolled down to f1.8 and with the last step to f1.8 it defocused! The AF kicked in for a moment and everything was nice and blurry. What the hell?

[vonBaloney]

The workaround is to move the aperture ring off the "A" setting. The camera only does this monkey business when the "A" contact is detected. For lenses lacking the aperture ring, another workaround would be to do focus testing, calibration, and such in relatively dim light or with an ND filter attached. FWIW, I seem to remember comments that this type of live view behavior was present in non-Pentax cameras as well.

If in manual focus mode, using the magnify function will keep the aperture wide-open while it is digitally zoomed in...

[Volker76]

Thanks to all responders, but to you both particularly! I was wondering if my camera was maybe misconfigured. While I agree this might possibly be considered a "feature" when shooting natural light, I think we can agree it's a bug when shooting manual flash, and under all circumstances I feel we should be able to turn it off. When I'm shooting in dim light to begin with, sometimes the display goes so dim that I can barely see a thing, and it becomes more or less impossible to focus (by moving the camera and using magnified liveview to achieve focus). And sometimes even if there's enough light, if I'm looking at the camera from an odd angle (remember, I shoot macro), or the sun is shining into the LCD, then the dimming is also very aggravating.

As I wrote before, using the aperture ring is a viable, if inconvenient, workaround. But I guess I won't be buying that D-FA 100mm WR macro I've been thinking about to get a WR macro body & lens combo...

actually with the 100mm I haven't seen the behaviour. If the effect is there with that lens it is really, really tiny. The DFA100 is just awesome.

[Stone G.]

This would be potentially useful when doing intentional over/underexposure in M mode. Yes...a feature, not a bug.

I agree - and I appreciate that feature.

Hmmmmmm...this is not good since the DOF displayed is not always the actual DOF for the exposure-time aperture.

I may misunderstand something here - but to actually see the resulting DOF for the set aperture, shouldn't one then always use DOF-preview ? And DOF-preview does work in Liveview - just tested it with my D-FA 100/2.8 WR and my FA 50/1.4 on a K-3.

[stevebrot]

but to actually see the resulting DOF for the set aperture, shouldn't one then always use DOF-preview ?

Yes, that would be the appropriate, but not intuitively obvious best practice. The common mantra spoken by myself and others is that live view is the gold standard for focus-related stuff, literally WYSIWYG. Apparently that should be qualified.


Steve

[stevebrot]

Scrolled down to f1.8 and with the last step to f1.8 it defocused! The AF kicked in for a moment and everything was nice and blurry. What the hell?

That is strange. AF should keep its peace unless active by button push. Are you sure it was not the aperture actuator?


Steve

[stevebrot]

If in manual focus mode, using the magnify function will keep the aperture wide-open while it is digitally zoomed in...

Cool! I just confirmed. I will amend my earlier comment.


Steve

[Doundounba]

I agree, though I would caution that in truly dim conditions live view is essentially unusable due to noise. [...]

Yes, noise does become an issue. But I've made shots where I basically half-guessed the framing because the LCD is just so dim, and then a few pixels kinda light-up from the focus peaking and - wham! - nice in-focus shot. Of course, high magnification macro is a numbers game anyway: one accepts that the DoF is so thin, the creature can and does move, you're never perfectly still, and so on, so it's a given that you just get very few keepers. It's just annoying when the camera makes it even harder than it needs to be. It took me a while to figure out this was happening - I just kept thinking, "is it just me or is the K-3 LCD dimmer than the K-01?" until I figured out that this was the cause.

I agree - and I appreciate that feature.

But don't you already have the eV bar that tells you if the shot will be under- or over-exposed?

In any case, for macro and in dim conditions, you do not want to be shooting with a pre-closed aperture. That is precisely the advantage in using, say, a 100mm macro plus a strong diopter rather than a reversed lens on tubes (at equivalent magnification). You're then focusing with a wide-open aperture, and can just see a whole lot better. If I'm shooting in tons of light, then I'm more likely to shoot with a reversed lens, where the aperture is preset. But when shooting under a thick canopy, in a dark forest, I can achieve focus with a regular lens plus diopter under conditions where using a reversed lens would just be hopeless (without some kind of continuous focus-assist light).

And at any rate, when you're using manual flash, there's no way the camera can preview the exposure. The LCD dimming in that case is really just a bug, IMHO...

[stevebrot]

It's just annoying when the camera makes it even harder than it needs to be. It took me a while to figure out this was happening - I just kept thinking, "is it just me or is the K-3 LCD dimmer than the K-01?" until I figured out that this was the cause.

Yes, I can see that for macro at say, f/22 for DOF and reasonably fast shutter speed and manual flash, live view is going to be rather dim. I thought there was a chance that Av mode with -5 stops exposure compensation might be an option, but the "dimmer" is on for exposure comp too.

But don't you already have the eV bar that tells you if the shot will be under- or over-exposed?

The EV bar does not tell you if you have shadow detail.


Steve

[vonBaloney]

I may misunderstand something here - but to actually see the resulting DOF for the set aperture, shouldn't one then always use DOF-preview ? And DOF-preview does work in Liveview - just tested it with my D-FA 100/2.8 WR and my FA 50/1.4 on a K-3.

The point is that the camera stops down while you are focusing (or just viewing) and so it is achieving focus or confirming focus (if manual) with possibly a greater depth of field than you will actually shoot at. So if you take your 1.4 AF lens, set it to 1.4, and go use CDAF autofocus *in bright sunlight*, the camera will be autofocusing at f/5.6, then you take the shot at f/1.4 and whatever you focused on is very very likely BADLY out of focus (because in order to achieve critical focus at f/1.4, the aperture actually needs to be wide-open while you do so, and it is not). I think you will find it is nearly impossible to get in-focus shots (of what you want) with a fast lens using a fast aperture when using CDAF liveview AF (or even MF if you aren't using the zoom function). And I find it strange that I seem to be the only one to have ever noticed this...

[stevebrot]

the camera will be autofocusing at f/5.6, then you take the shot at f/1.4 and whatever you focused on is very very likely BADLY out of focus (because in order to achieve critical focus at f/1.4, the aperture actually needs to be wide-open while you do so, and it is not).

This may not be an accurate assumption. Remember that while DOF is real, for any target, there is only one true point of focus. This is true regardless of aperture and is readily demonstrable by simply viewing an image at high magnification. The question is whether the CDAF system has adequate focus sensitivity (the ability to detect OOF condition) when contrast is artificially high due to the smaller aperture.

It may be useful to do a series of twenty or so shots using a ramped calibration target in order to determine whether the stop-down is a problem.


Steve

[stevebrot]

And I find it strange that I seem to be the only one to have ever noticed this...

You aren't. I noticed it within the first month of taking delivery of my K-3 as did several other users at about the same time. It may not have continued as a hot topic because of how people use live view or it may be that it did not make much difference in people's results. To be honest, I think that more people are concerned about focus shift on stop-down (a minor problem in practice) than this matter.


Steve

[vonBaloney]

This may not be an accurate assumption. Remember that while DOF is real, for any target, there is only one true point of focus. This is true regardless of aperture and is readily demonstrable by simply viewing an image at high magnification. The question is whether the CDAF system has adequate focus sensitivity (the ability to detect OOF condition) when contrast is artificially high due to the smaller aperture.

It may be useful to do a series of twenty or so shots using a ramped calibration target in order to determine whether the stop-down is a problem.

You aren't. I noticed it within the first month of taking delivery of my K-3 as did several other users at about the same time. It may not have continued as a hot topic because of how people use live view or that it did not make much difference in people's results. To be honest, I think that more people are concerned about focus shift on stop-down (a minor problem in practice) than this matter.

But it has been doing that on all the previous models as well going back years. And I can tell you it is a real problem because I've tried to use fast lenses at fast apertures and can NEVER get the subject in focus unless I switch to manual and use the magnify function. (Even on an isolated subject where there is nothing else to focus on.) It is essentially causing itself reverse focus shift by design...

[stevebrot]

I've tried to use fast lenses at fast apertures and can NEVER get the subject in focus unless I switch to manual and use the magnify function.

I am not a live view user except for when I am doing focus calibration and even then always use magnified manual focus. Without doing some hands-on research, I cannot comment on the subject beyond the observation that CDAF complaints on this site are relatively rare and the anecdotal accounts from the camera reviewers on this site that the CDAF implementation has superior accuracy and precision compared to the PDAF system. It may be that bright lighting at wide apertures has not been a use case for those people.


Stee