[Pål Jensen]

Yesterday I tried to recreate Dpreviews findings that the K3 III will not give sharp images at 1/125s when the camera is mounted on a sturdy tripod. The theory is that it is due to shutter shake.
To my surprise I got the same results. It is not due to shutter shake however.

I used the camera on a tripod and engaged the 2s self-timer that locks up the mirror. I shot first at the metered shutter speed (quite slow; 1/5 - 1/2s) at 100ISO, then boosted the ISO until the speed was 1/125s. The slower shutter speed images were sharp; the 1/125s weren't. The fuzzy images were sharper in the center and gradually worse towards the edges. Worst at the far edges. At the far edges it can be seen that the subject is streaked; like like start trails at long exposure. Ie the fuzzyness is due to rotation. This can also be deduced from the fact that the two images were slightly rotated compared to each other in spite of the fact that they were shot few seconds apart while mounted on a tripod.
It is clear that the sensor is rotating during exposure. SR is supposed to be off when using this mode. I tried to turn SR off manually, but the camera (naturally) says that SR is not available in this mode.
It seems like the SR is on or malfunctioning in 2s mirror lock mode.

I tried to shoot at 1/125s in normal manner; that is without mirror lock. All images turns out super sharp. Hence, there's no shutter shock (or mirror slap for that matter) at 1/125s.

I hope this is a bug that can be fixed in software. Either SR is on during 2s self timer mode or it is malfunctioning somehow in this mode. My camera has Firmware 1.01.

---

Last edited by Pål Jensen; 06-10-2021 at 03:54 AM.

[beholder3]

Yesterday I tried to recreate Dpreviews findings that the K3 III will not give sharp images at 1/125s when the cameras is mounted on a sturdy tripod. The theory was that it was due to shutter shake.
To my surprise I got the same results. It is not due to shutter shake however.

I used the camera on a tripod and engaged the 2s self-timer that locks up the mirror. I shot first at the metered shutter speed (quite slow; 1/5 - 1/2s) at 100ISO, then boosted the ISO until the speed was 1/125s. The slower shutter speed images were sharp; the 1/125s weren't. The fuzzy images were sharper in the center and gradually worse towards the edges. Worst at the far edges. At the far edges it can be seen that the subject is streaked; like like start trails at long exposure. Ie the fuzzyness is due to rotation. This can also be deduced from the fact that the two images were slightly rotated compared to each other in spite of the fact that they were shot few seconds apart while mounted on a tripod.
It is clear that the sensor is rotating during exposure. SR is supposed to be off when using this mode. I tried to turn SR off, but the camera (naturally) says that SR is not available in this mode.
It seems like the SR is on or malfunctioning in 2s mirror lock mode.

I tried to shoot at 1/125s in normal manner; that is without mirror lock. All images turns out super sharp. Hence, there's no shutter shock (or mirror slap for that matter) at 1/125s.

I hope this is a bug that can be fixed in software. Either SR is on during 2s self timer mode or it i malfunctioning somehow in this mode. My camera has Firmware 1.01.

I would comment that I think it is perfectly possible that neither SR or anything else is specifically wrong, but that the physical actions in this scenario lead to masses being moved in a way that it creates a resonance/freuqnecy of shake that is not fully dampened on a tripod when shutter speed is 1/125 sec.
I bet the 2 sec mirror lock thing is not present in handheld modes (because of proper dampening).
I can imagine that the acceleration and deceleration of a shutter to 1/125 sec speed in the specific process done for 2 sec MU is actually causing it.

All these vibration shake things 100% depend on a) the resonance frequencies of the combined physical items (body + lens + tripod head + tripod) and b) the quality of dampening available (hands are good, ash wood tripods are good; metal or even worse carbon tripods are bad).

For the people with some scientific interests here two videos showing resonant frequencies:

You Tube

You Tube

and

You Tube

You Tube

[StigVidar]

Horizon correction feature may blur corners: Pentax SLR Talk Forum: Digital Photography Review

Conclusion by me: This is fixable in a firmware release and I am certain that Pentax will fix it.

[phoebus]

The question of SR on or off is not relevant. The Sensor is always positioned by the electromagnet assembly - the SR just adds compensating motion when it is on.
What may be happening is the strength of positioning is reduced in 2s delay, as I imagine it doesn't help battery life. Did 10s delay have the same issue?

[StigVidar]

Did 10s delay have the same issue?

I suggest you read the information on the link to dpreview in the post above yours. It will give you more information about this issue.

[c.a.m]

Horizon correction feature may blur corners: Pentax SLR Talk Forum: Digital Photography Review
Conclusion by me: This is fixable in a firmware release and I am certain that Pentax will fix it.

Thanks for posting the link. I have been following that thread at DPReview. I think that it covers the issue well.

In summary, the issue appears when Horizon Correction is ON and the 2-second, 12-second, or IR Remote 3-second timer is used. The issue appears in either viewfinder mode or Live View.

I have reproduced the effect -- camera on tripod, Horizon Correction ON, 2-second timer, DA 20-40mm Limited at 35mm, f/5.6, 1/60 s. The effect is independent of the SR mode; SR can be ON or OFF. I have not tested it while hand-holding the camera, nor at other shutter speeds.

The centre of the image is sharp, while the corners are blurry. The blurriness increases progressively from the centre towards the corners.

- Craig

[slartibartfast01]

I would comment that I think it is perfectly possible that neither SR or anything else is specifically wrong, but that the physical actions in this scenario lead to masses being moved in a way that it creates a resonance/freuqnecy of shake that is not fully dampened on a tripod when shutter speed is 1/125 sec.

I bet the 2 sec mirror lock thing is not present in handheld modes (because of proper dampening).

I can imagine that the acceleration and deceleration of a shutter to 1/125 sec speed in the specific process done for 2 sec MU is actually causing it.



All these vibration shake things 100% depend on a) the resonance frequencies of the combined physical items (body + lens + tripod head + tripod) and b) the quality of dampening available (hands are good, ash wood tripods are good; metal or even worse carbon tripods are bad).



For the people with some scientific interests here two videos showing resonant frequencies:

SDOF Resonance Vibration Test - YouTube

and

Mechanical Resonance - YouTube

If pictures taken with the 2 second timer are worse than those taken without it I think most people would think there was an issue.

[BigMackCam]

K-3 III bug in 2s self timer mode - Source of fuzzyness

Thanks for doing this test and posting the results. Even though I don't yet own a K-3 III, it's encouraging.

The question of SR on or off is not relevant. The Sensor is always positioned by the electromagnet assembly - the SR just adds compensating motion when it is on.
What may be happening is the strength of positioning is reduced in 2s delay, as I imagine it doesn't help battery life. Did 10s delay have the same issue?

It would also be interesting to see if the issue persists with SR off and shutter triggered by a remote...

[StigVidar]

It would also be interesting to see if the issue persists with SR off and shutter triggered by a remote...

The answer is in the link to dpreview above .......

[beholder3]

Thanks for posting the link. I have been following that thread at DPReview. I think that it covers the issue well.

In summary, the issue appears when Horizon Correction is ON and the 2-second, 12-second, or IR Remote 3-second timer is used.

I have reproduced the effect -- camera on tripod, Horizon Correction ON, 2-second timer, DA 20-40mm Limited at 35mm, f/5.6, 1/60 s. The effect is independent of the SR mode; SR can be ON or OFF. I have not tested it while hand-holding the camera, or at other shutter speeds.

The centre of the image is sharp, while the corners are blurry. The blurriness increases progressively from the centre towards the corners.

- Craig

Here it could be that the horizon correction somehow yanks the sensor into position at a unlucky timing or it is correcting / moving during the exposure timeframe in some way not intended.

---------- Post added 10th Jun 2021 at 15:05 ----------

If pictures taken with the 2 second timer are worse than those taken without it I think most people would think there was an issue.

I was discussing the causes of the vibration blur, not if there might be some.

I need to see comparisons made at same settings handheld versus well dampened pro ashwood tripod versus simple carbon tripod.

[bicycle]

Now I know what was going on. A week ago I was out for a walk and took a break at a picnic pavilion to escape the heat for a minute. Played with the camera and settings for a bit while I was there. I was in for a surprise when I got them on the computer at home. What a swirl effect.


The lens was the DA 60-250 with the DA 1.4 teleconverter on, which gave a focal length of 84mm. I was looking to see the effect of diffraction, but not in any real scientific way since I just did one run through. I had set the camera to 2-sec delay (which I have set to turn off SR), Horizon correction is set to on. The camera was setting on the top of the a solid, non-jiggling, picnic table. I didn't do any attempts to recreate the effect just marked it down to operator error for someone with a new camera. The swirl effect was obvious between 1/30 to 1/125 sec. The corners were pretty bad at all the speeds I used.

[c.a.m]

Now I know what was going on.

Thanks for posting. In a way, it's good to see this 'real-world' example, which correlates with the various basement-lab tests reported by several users. Someone else had suggested that a user could reasonably be expected to use Horizon Correction and the timer when the camera is not hand-held but placed on a stable platform other than a tripod -- your case precisely.

So, at least you know it was not operator error.


- Craig

[bicycle]

Here is a 100% crop of the upper left corner. Sorry couldn't figure how to add to edit this into my original post.

[Pål Jensen]

All these vibration shake things 100% depend on a) the resonance frequencies of the combined physical items (body + lens + tripod head + tripod) and b) the quality of dampening available (hands are good, ash wood tripods are good; metal or even worse carbon tripods are bad).

Strange then that it doesn't happen when used on a tripod when not using 2s delay (all else equal)...

---------- Post added 06-10-21 at 07:33 PM ----------

Horizon correction feature may blur corners: Pentax SLR Talk Forum: Digital Photography Review

Conclusion by me: This is fixable in a firmware release and I am certain that Pentax will fix it.

Yes. Horizontal correction may also be the culprit.....

[madbrain]

That looks like motion blur. Was it windy ? Are static objects also blurred? For example, the roof.