[Stefan Jr]

Someone might how to correct this issue. Cause I can't find how to solve it.

As the title says, my K-1 overexposure in Live View, but exposure correctly when using the OVF.
Live View need a -0,7 (almost 1,0) EV correction.

Examples shown are in shot in AV mode. With a Carl Zeiss Planar 50/1,4 ZK (set to A).

Photo 1 - through the OVF.
Photo 2 - with Live View (no exposure compensation)
Photo 3 - with Live View (-0,7 EV compensation)

Thanks in advance for any insight.

((Oddly enough, in M mode, exposure is fine with Live View. But the white balance is weirdly way off)).

Attached Images

     

[biz-engineer]

I have 0.3 ev more in LV compared to an exposure with OVF closed. I have more than 0.3 ev in LV when the OVF is not obstructed (light goes through the OVF pushing the camera AE to underexpose), leading to believe that LV overexpose, while it's OVF that actually underexpose when exposed to light from the outside.

[pschlute]

I think biz-engineer has your answer.

LV exposure and OVF exposure use two different methods, so it is not right to expect them always to be exactly the same.

You could as an experiment cover the viewfinder with the VF cover and see if it still exposes as you think it should.

[Stefan Jr]

I have 0.3 ev more in LV compared to an exposure with OVF closed. I have more than 0.3 ev in LV when the OVF is not obstructed (light goes through the OVF pushing the camera AE to underexpose), leading to believe that LV overexpose, while it's OVF that actually underexpose when exposed to light from the outside.

I actually shielded the OVF (fully - with the plastic cover). And although the general appearance (and a few reflections) improved - the exposure difference remained.

I would say it's the Live View that is overexposed, the whites get blown out. While OVF-photos are better balanced, both highs and lows. The photos attached are all
original RAW (saved for the web as JPEG), so there is no JPEG artifacts involved.

---------- Post added 04-09-20 at 11:21 AM ----------

I think biz-engineer has your answer.

LV exposure and OVF exposure use two different methods, so it is not right to expect them always to be exactly the same.

You could as an experiment cover the viewfinder with the VF cover and see if it still exposes as you think it should.

Sure - different methods. Although - the difference is on the hefty side.

I thought I check with you guys, with some luck there might have been a fine adjustment deep in the menus that I've missed.

---

Last edited by Stefan Jr; 04-09-2020 at 11:09 AM.

[MJKoski]

Setting "wrong" WB can itself blow off some channels for 8bit JPG image whereas in RAW highlights are not clipped.

I just expose as far as the sensor can handle the highlights and then drop the EV adjustment in post.

Zeiss 50/1.4 has high vignetting wide open which may affect liveview. Pentax LV keeps the aperture as open as possible all times and my Zeiss 50mm makro-planar ZK also overexposes in LV due to vignette wide open.

[Stefan Jr]

Setting "wrong" WB can itself blow off some channels for 8bit JPG image whereas in RAW highlights are not clipped.

I just expose as far as the sensor can handle the highlights and then drop the EV adjustment in post.

Zeiss 50/1.4 has high vignetting wide open which may affect liveview. Pentax LV keeps the aperture as open as possible all times and my Zeiss 50mm makro-planar ZK also overexposes in LV due to vignette wide open.

I basically use the same exposure technique.
The problem is that I base it on what I see on the LV.

The photos are at f5,6, and even at f8 the the same issue remained. And auto WB - tried the multi WB too.

I guess I need to mindset the exposure difference for different scenarios. I just wish there was a way to decline the difference between the LV and OVF a bit.

[biz-engineer]

Exposure in LV M mode is measured with lens aperture stopped down. Exposure with OVF is measured with lens aperture wide open.
So , any deviation of actual lens aperture relative to theoretical lens aperture assumed by the camera will offset the exposure.
Exposure settings can toggle by +-0.3ev because it is the minimum step for the camera. So 0.333ev of systematic error from the lens from wide open to stopped down + 0.333ev of setting step will produce up to 0.666ev of exposure error.

[Stefan Jr]

Exposure in LV M mode is measured with lens aperture stopped down. Exposure with OVF is measured with lens aperture wide open.
So , any deviation of actual lens aperture relative to theoretical lens aperture assumed by the camera will offset the exposure.
Exposure settings can toggle by +-0.3ev because it is the minimum step for the camera. So 0.333ev of systematic error from the lens from wide open to stopped down + 0.333ev of setting step will produce up to 0.666ev of exposure error.

OK - thank you for the theoretics!That make sense to my situation/scenario.

I got my EV set to 1/3 steps and end up in about -0,7 EV to level out between OVF and LV.
There's a small underexposure with OVF, but larger overexposure with LV.

I was hoping the was a setting that magically could be found with you guys help.
But I'm guess I'm simply stuck with the small regular tricks I usually use.

Thanks again!

[MJKoski]

No the exposure in LV is not measured stopped down if it gets dark enough. When in dim conditions the lens opens up to max possible aperture and camera simulates the rest of the exposure including the histogram. This completely throws off exposure calculation when the attached lens has high enough vignetting.

This is also why most MILC bodies have baked-in lens corrections (vignetting and/or distortion) enabled at all times. To ensure even metering of the scene.

[acoufap]

LiveView depends on the chosen Customer Color Profile. What about if you try to use another one - by example Flat. Does that change anything?

[Stefan Jr]

LiveView depends on the chosen Customer Color Profile. What about if you try to use another one - by example Flat. Does that change anything?

I will check and see!
And if I eventually find anything useful, I'll post an update.

I always use "natural" setting, with no tweaking. As neutral as ever possible.

---------- Post added 04-09-20 at 02:11 PM ----------

No the exposure in LV is not measured stopped down if it gets dark enough. When in dim conditions the lens opens up to max possible aperture and camera simulates the rest of the exposure including the histogram. This completely throws off exposure calculation when the attached lens has high enough vignetting.

This is also why most MILC bodies have baked-in lens corrections (vignetting and/or distortion) enabled at all times. To ensure even metering of the scene.

Thanks - that make sense.
But my issue didn't improve as I was stopping down to 5,6 and 8. As it normally does. The difference was still significant.

[MJKoski]

Stopping down in liveview to whatever aperture from wide open may not help here if the camera body determines light is too dim. It tries to calculate the correct exposure from wide open liveview feed taking into account the selected aperture. Then this happens: If wide open aperture is f/2 and the lens vignettes heavily wide open and lens aperture ring is set to A, selecting f/4 does not stop the lens down. Instead, f/4 exposure is calculated from vignette influenced dark(ish) picture feed. Result is overexposure. Stopping down does not help when the base exposure calculation fails due to vignette. Camera will continue overexposing at any selected aperture.

Pentax liveview is currently a catastrophe. First, one needs at least A-type lens or more modern one for it to work at least somewhat properly in manual mode. It cannot simulate exposure if fully manual lens is used (Zeiss .ZK counts as one when aperture is adjusted from the aperture ring itself). And then, only green button metering gives some kind of result. And even if green mode metering works, histogram does not upgrade correctly.

You could try green mode metering. For it will stop the lens down and then meter the exposure. Try setting the aperture from the lens before you do that.

I suffer from these issues all the time with my .ZK Zeiss lens and Samyang PK lenses. I mostly use K-1 in liveview and in fully manual mode. The LV implementation should have a total makeover from scratch and include proper exposure simulation for starters.

[Bob 256]

You didn't mention if this happens with all your lenses or if you're just seeing this with one in particular. Have you tried other lenses? Also, what metering is set for the OVF? You have several choices (press and hold the AF button on the left of the lens, and the back-side adjustment wheel will dial you through those available). It could be possible that the OVF is metering off a different location than the LV. Shoot an out-of-focus gray card (or a gray surface) which fills the entire field and see if you get different exposures with the OVF and LV??

[biz-engineer]

AE metering modes: matrix, center weighted, spot. Why use matrix metering when the overexposed part is the white subject.

Anyway, no camera in the world even exposes perfectly, this is the reason why most ilc cameras have an exposure compensation feature and modes.

Typically to check exposure metering you dont use a scene with pure white parts and high contrast. A good start would be to use a grey card.

Typically, when shooting portraits you use center weighted (subject in center) or spot ae. Exposure basics. No need to jump on horses to say how camera metering is faulty.

[Stefan Jr]

I want to thank everyone for your feedback.
I will take all into account and continue my testing.

I've checked further and now ensured, that in controlled/directed light and at full aperture, exposure was good in LV.
So I know for sure the lens is OK.

And - I know my K-1 are OK too. Especially with the Pentax 100mm macro WR - exposure is silly accurate. And with,
for example, the CZ distagon 35/2. And other fast fifties.

But for some reason this camera and lens marriage ride into a little bump, which I'll simply just need to work around.
Thanks again for sharing your minds.

Best // Stefan