Not happy with Pentax K-3 III

wkraus · 06-23-2021, 03:20 AM

Originally posted by stevebrot

Tighter distribution infers less blur? Are we assuming each case shares the same mean?

Yes, and yes. You are aware that I am referring to a series with the same object, identical exposure parameters and identical develop settings? The Pixel Shift sample was slightly different but it was not what was at issue in the first place.

Originally posted by stevebrot

FWIW, the distribution in PS is the creation of ACR.* Processing to a TIFF using non-transformed RAW conversion might show something different.

Steve

* Pet peeve of mine.

Of course it is. But we are interested in real-world images, aren’t we? If you would like to play with the raw files I can send them to you.

stevebrot · 06-23-2021, 03:22 AM

Originally posted by Pål Jensen

If you read this thread: K-3 III bug in 2s self timer mode - Source of fuzzyness - PentaxForums.com you will see that the horizon correction is not a red herring. Several people have reported that the problem disappear when horizon correction is turned off.

I did...in detail. The issue with the horizon correction is very real and is probably as you and others have reported, but there is no reason to assign it as cause for the ISO-linked softness at lower shutter speeds. I apologize for using the term "red herring". All it means is that the observation is unrelated to the conclusion.

Steve

stevebrot · 06-23-2021, 03:41 AM

Originally posted by wkraus

Of course it is. But we are interested in real-world images, aren’t we?

Then post full-resolution crops of the same region. I am not convinced that those SD's are statistically different from each other.

Steve

Eyejason · 06-23-2021, 03:42 AM

Very frustrating for you

wkraus · 06-23-2021, 04:01 AM - **1 Like**

Originally posted by stevebrot

Then post full-resolution crops of the same region. I am not convinced that those SD's are statistically different from each other.

Steve

These are full resolution crops. Sorry, I should have mentioned it.

wkraus · 06-23-2021, 09:09 AM

Another real-world test: Sigma 70mm 1:2.8 EX DG Macro at f/6.3, 1/80, ISO 160, Manfrotto 074 tripod with X-Pro 3 head, straight exposure with cable release and no measures to reduce vibration. In the second sample I dampened the tripod by putting my hand on the head during exposure.

When looking REALLY closely there might be the tiniest amount of blur in the undampened sample but I’m not sure. Image statistics do not support that there is a difference.

You’re welcome to download and compare the samples (full resolution centre crops), which somehow always look sharper in Photoshop than on a webpage.

I have seen a little blur in this exposure range with my Manfrotto tripod but it is hard to reproduce. Proper experimentation would require all variables to be fixed except one whose effect can then be studied. However, in a system as complex as a camera plus tripod, this is all but impossible. I’ll admit that there can be some situations when what is most probably mirror shock can cause vibration-induced blur (as with any camera I have used so far) but I do not consider it a real problem, at least with my unit.

**Pål Jensen** · 06-23-2021, 10:04 AM

Originally posted by stevebrot

I did...in detail. The issue with the horizon correction is very real and is probably as you and others have reported, but there is no reason to assign it as cause for the ISO-linked softness at lower shutter speeds. I apologize for using the term "red herring". All it means is that the observation is unrelated to the conclusion.

Steve

Sure, but as I said previously I'm only aware of two instances where this has been reported. Hence, I'm not (yet) convinced that this is a real issue....

stevebrot · 06-23-2021, 11:02 AM - **1 Like**

Originally posted by Pål Jensen

Sure, but as I said previously I'm only aware of two instances where this has been reported. Hence, I'm not (yet) convinced that this is a real issue....

That is fine, though after reviewing the OPs methods and resources on their site, I found a high degree of rigor backing the observation and compelling evidence the observation is valid. Assuming, of course that they were not totally faking the results.

If I find someone locally with a K-3iii, I will attempt to reproduce using gear I have on hand. My LPL 670XL is adaptable to a copy stand and I have high-end lenses similar to those used by the OP.

Steve

(...Koch's postulates for establishing causation in disease come to mind...by analogy only...)

Class A · 06-23-2021, 08:14 PM - **1 Like**

Originally posted by wkraus

When looking REALLY closely there might be the tiniest amount of blur in the undampened sample but I’m not sure.

The second (dampened) image shows the tiniest amount of additional detail (only detectable, AFAIC, when directly overlaying the two images and switching back and forth between them).

wkraus · 06-24-2021, 12:54 AM

Originally posted by Class A

The second (dampened) image shows the tiniest amount of additional detail (only detectable, AFAIC, when directly overlaying the two images and switching back and forth between them).

Thanks for looking! So would you agree that in this specific situation (meant to be somehow similar to the DPR setup) shutter/mirror blur is a non-issue?

Class A · 06-24-2021, 03:32 AM

Originally posted by wkraus

So would you agree that in this specific situation (meant to be somehow similar to the DPR setup) shutter/mirror blur is a non-issue?

Yes, definitely.

DPReview's blur is far, far greater; I'm hoping they just happened to receive an out of spec camera.

mctaveck · 06-24-2021, 05:39 AM

Originally posted by Pål Jensen

No. Only fuzzy images while using 2s self timer and horizon correction at some medium shutter speeds. It's a bug.
Why should shutter shock be restricted to 100 and 200 ISO? Makes no sense....

I have this issue. It gets relevant as the exposure time becomes slower. Any Idea about how resolve the bug? I mean, maybe someone has news about it.
Thanks
Danilo

mtkeller · 06-24-2021, 06:43 AM

Originally posted by mctaveck

I have this issue. It gets relevant as the exposure time becomes slower. Any Idea about how resolve the bug? I mean, maybe someone has news about it.
Thanks
Danilo

Turn off horizon correction or use something other than the 2s timer if you need horizon correction.

mctaveck · 06-24-2021, 08:26 AM

Originally posted by mtkeller

Turn off horizon correction or use something other than the 2s timer if you need horizon correction.

Thanks. This is not the answer I am looking for, unfortunately. To be more explicit I was referring to Pentax fix (i.e. Firmware) or to a confirmation this is a software issue and not a technical issue. For sure i can get horizontal landscapes with 1 click in post but ....

Sorry

Danilo

stevebrot · 06-24-2021, 08:58 AM

Originally posted by mctaveck

I have this issue. It gets relevant as the exposure time becomes slower. Any Idea about how resolve the bug? I mean, maybe someone has news about it.
Thanks
Danilo

That is an interesting variation. The reports for horizon correction + 2s timer are for blur in the 1/125s range.

Steve

06-23-2021, 09:09 AM	#111
wkraus Pentaxian Join Date: Jan 2013 Location: Near Vienna, Austria Photos: Gallery Posts: 2,060	Another real-world test: Sigma 70mm 1:2.8 EX DG Macro at f/6.3, 1/80, ISO 160, Manfrotto 074 tripod with X-Pro 3 head, straight exposure with cable release and no measures to reduce vibration. In the second sample I dampened the tripod by putting my hand on the head during exposure. When looking REALLY closely there might be the tiniest amount of blur in the undampened sample but I’m not sure. Image statistics do not support that there is a difference. You’re welcome to download and compare the samples (full resolution centre crops), which somehow always look sharper in Photoshop than on a webpage. I have seen a little blur in this exposure range with my Manfrotto tripod but it is hard to reproduce. Proper experimentation would require all variables to be fixed except one whose effect can then be studied. However, in a system as complex as a camera plus tripod, this is all but impossible. I’ll admit that there can be some situations when what is most probably mirror shock can cause vibration-induced blur (as with any camera I have used so far) but I do not consider it a real problem, at least with my unit. Last edited by wkraus; 06-23-2021 at 09:14 AM.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
K-3 Mark III and GR III topping one Japanese sales chart	JPT	Photographic Industry and Professionals	27	07-31-2021 07:22 AM
Comparison K3 III versus Sony A7 III	bwgv001	Pentax DSLR Discussion	51	04-30-2021 01:20 PM
Not one, not two, not three, not four, but a wedding where half attendees are bob.	LeDave	Photographic Industry and Professionals	12	05-16-2016 03:40 AM
Happy happy joy joy	pingflood	General Talk	3	06-13-2009 07:19 PM

06-23-2021, 03:20 AM	#106
wkraus Pentaxian Join Date: Jan 2013 Location: Near Vienna, Austria Photos: Gallery Posts: 2,060	Originally posted by stevebrot Tighter distribution infers less blur? Are we assuming each case shares the same mean? Yes, and yes. You are aware that I am referring to a series with the same object, identical exposure parameters and identical develop settings? The Pixel Shift sample was slightly different but it was not what was at issue in the first place. Originally posted by stevebrot FWIW, the distribution in PS is the creation of ACR.* Processing to a TIFF using non-transformed RAW conversion might show something different. Steve * Pet peeve of mine. Of course it is. But we are interested in real-world images, aren’t we? If you would like to play with the raw files I can send them to you.

06-23-2021, 03:22 AM	#107
stevebrot Otis Memorial Pentaxian Join Date: Mar 2007 Location: Vancouver (USA) Photos: Gallery \| Albums Posts: 42,007	Originally posted by Pål Jensen If you read this thread: K-3 III bug in 2s self timer mode - Source of fuzzyness - PentaxForums.com you will see that the horizon correction is not a red herring. Several people have reported that the problem disappear when horizon correction is turned off. I did...in detail. The issue with the horizon correction is very real and is probably as you and others have reported, but there is no reason to assign it as cause for the ISO-linked softness at lower shutter speeds. I apologize for using the term "red herring". All it means is that the observation is unrelated to the conclusion. Steve

06-23-2021, 03:41 AM	#108
stevebrot Otis Memorial Pentaxian Join Date: Mar 2007 Location: Vancouver (USA) Photos: Gallery \| Albums Posts: 42,007	Originally posted by wkraus Of course it is. But we are interested in real-world images, aren’t we? Then post full-resolution crops of the same region. I am not convinced that those SD's are statistically different from each other. Steve

06-23-2021, 03:42 AM	#109
Eyejason Forum Member Join Date: Nov 2017 Posts: 61	Very frustrating for you

06-23-2021, 04:01 AM - 1 Like	#110
wkraus Pentaxian Join Date: Jan 2013 Location: Near Vienna, Austria Photos: Gallery Posts: 2,060	Originally posted by stevebrot Then post full-resolution crops of the same region. I am not convinced that those SD's are statistically different from each other. Steve These are full resolution crops. Sorry, I should have mentioned it.

06-23-2021, 10:04 AM	#112
Pål Jensen Site Supporter Join Date: Dec 2007 Location: Norway Photos: Albums Posts: 4,371	Originally posted by stevebrot I did...in detail. The issue with the horizon correction is very real and is probably as you and others have reported, but there is no reason to assign it as cause for the ISO-linked softness at lower shutter speeds. I apologize for using the term "red herring". All it means is that the observation is unrelated to the conclusion. Steve Sure, but as I said previously I'm only aware of two instances where this has been reported. Hence, I'm not (yet) convinced that this is a real issue....

06-23-2021, 11:02 AM - 1 Like	#113
stevebrot Otis Memorial Pentaxian Join Date: Mar 2007 Location: Vancouver (USA) Photos: Gallery \| Albums Posts: 42,007	Originally posted by Pål Jensen Sure, but as I said previously I'm only aware of two instances where this has been reported. Hence, I'm not (yet) convinced that this is a real issue.... That is fine, though after reviewing the OPs methods and resources on their site, I found a high degree of rigor backing the observation and compelling evidence the observation is valid. Assuming, of course that they were not totally faking the results. If I find someone locally with a K-3iii, I will attempt to reproduce using gear I have on hand. My LPL 670XL is adaptable to a copy stand and I have high-end lenses similar to those used by the OP. Steve (...Koch's postulates for establishing causation in disease come to mind...by analogy only...)

06-23-2021, 08:14 PM - 1 Like	#114
Class A Pentaxian Join Date: Aug 2008 Location: Wellington, New Zealand Posts: 11,251	Originally posted by wkraus When looking REALLY closely there might be the tiniest amount of blur in the undampened sample but I’m not sure. The second (dampened) image shows the tiniest amount of additional detail (only detectable, AFAIC, when directly overlaying the two images and switching back and forth between them).

06-24-2021, 12:54 AM	#115
wkraus Pentaxian Join Date: Jan 2013 Location: Near Vienna, Austria Photos: Gallery Posts: 2,060	Originally posted by Class A The second (dampened) image shows the tiniest amount of additional detail (only detectable, AFAIC, when directly overlaying the two images and switching back and forth between them). Thanks for looking! So would you agree that in this specific situation (meant to be somehow similar to the DPR setup) shutter/mirror blur is a non-issue?

06-24-2021, 03:32 AM	#116
Class A Pentaxian Join Date: Aug 2008 Location: Wellington, New Zealand Posts: 11,251	Originally posted by wkraus So would you agree that in this specific situation (meant to be somehow similar to the DPR setup) shutter/mirror blur is a non-issue? Yes, definitely. DPReview's blur is far, far greater; I'm hoping they just happened to receive an out of spec camera.

06-24-2021, 05:39 AM	#117
mctaveck Senior Member Join Date: Apr 2017 Posts: 165	Originally posted by Pål Jensen No. Only fuzzy images while using 2s self timer and horizon correction at some medium shutter speeds. It's a bug. Why should shutter shock be restricted to 100 and 200 ISO? Makes no sense.... I have this issue. It gets relevant as the exposure time becomes slower. Any Idea about how resolve the bug? I mean, maybe someone has news about it. Thanks Danilo

06-24-2021, 06:43 AM	#118
mtkeller Veteran Member Join Date: Feb 2019 Location: Sioux City, IA Photos: Gallery Posts: 938	Originally posted by mctaveck I have this issue. It gets relevant as the exposure time becomes slower. Any Idea about how resolve the bug? I mean, maybe someone has news about it. Thanks Danilo Turn off horizon correction or use something other than the 2s timer if you need horizon correction.

06-24-2021, 08:26 AM	#119
mctaveck Senior Member Join Date: Apr 2017 Posts: 165	Originally posted by mtkeller Turn off horizon correction or use something other than the 2s timer if you need horizon correction. Thanks. This is not the answer I am looking for, unfortunately. To be more explicit I was referring to Pentax fix (i.e. Firmware) or to a confirmation this is a software issue and not a technical issue. For sure i can get horizontal landscapes with 1 click in post but .... Sorry Danilo

06-24-2021, 08:58 AM	#120
stevebrot Otis Memorial Pentaxian Join Date: Mar 2007 Location: Vancouver (USA) Photos: Gallery \| Albums Posts: 42,007	Originally posted by mctaveck I have this issue. It gets relevant as the exposure time becomes slower. Any Idea about how resolve the bug? I mean, maybe someone has news about it. Thanks Danilo That is an interesting variation. The reports for horizon correction + 2s timer are for blur in the 1/125s range. Steve