DPR Article on Pixel Shift

wanderer2 · 11-14-2017, 05:38 PM - **1 Like**

I'm not sure if this has already been posted (it didn't turn up in a search) or if this is the right forum to post it, but Digital Photography Review released an article this week on the pixel shift resolution in the new Sony a7R III. DPR absolutely raved about how wonderful it is. What a stark contrast to their review of the K-1 last year in which they listed the pixel shift as a "con" only useful with stationary objects because the Adobe programs couldn't process the K-1's pixel shift with motion control files. A link to the article:

https://www.dpreview.com/news/5645755619/sony-a7r-iii-pixel-shift-lifts-a-ve..._dpr_nl_285_18

A lot of us felt the DPR K-1 review was biased in several ways, pixel shift being one of them. This latest from DPR reinforces that perception, at least for me.

stevebrot · 11-14-2017, 05:54 PM - **1 Like**

Originally posted by wanderer2

A lot of us felt the DPR K-1 review was biased in several ways, pixel shift being one of them. This latest from DPR reinforces that perception, at least for me.

Hmmm...Are we to be discussing the A7R III, Sony's pixel shift implementation, or DPReview? The DPR review of the K-1 has been thoroughly discussed on other threads.

Steve

**Rondec** · 11-14-2017, 05:55 PM

They do mention that the K-1 produces similar results, but they sure did pound the K-1's pixel shift feature as being useless for real world situations when it came out.

pinholecam · 11-14-2017, 07:02 PM - **1 Like**

Another site that I often ask folks to abandon (its forums) and just let them loose the Pentax userbase there totally. (not that they really value it anyways)

I do recall the few articles they had about the K1 (and pixel shift) and the underlining tone was always that it was left wanting due to limited use situations and no Adobe support.
This A7RIII is different in what they are saying now :

"Are we impressed?

How could we not be? Landscape, cityscape and architecture photographers will absolutely love this new feature paired with the already excellent sensor in the a7R III - as long as they steer clear of (or clone out) moving objects in the scene. The increase in resolution and decrease in aliasing Pixel Shift brings is obvious in both our studio scene and real world result. It's frankly dramatic in the latter.
Are we impressed? How could we not be? Landscape, cityscape and architecture photographers will love this

There can be obvious artifacts in anything moving though, so that's a potentially significant (albeit expected) caveat for landscapes with motion (water, fast clouds), telephoto shots prone to movement from wind and vibrations, etc. You'll want to use a sturdy tripod with a remote release or self-timer. Furthermore, for now, using Sony's 'Imaging Edge' software is clunky, but once Adobe incorporates support, we can't wait to start shooting landscapes and perhaps tougher subjects in this mode to see how well it copes."

pakinjapan · 11-14-2017, 07:15 PM - **3 Likes**

Originally posted by pinholecam

I do recall the few articles they had about the K1 (and pixel shift) and the underlining tone was always that it was left wanting due to limited use situations and no Adobe support.
This A7RIII is different in what they are saying now :

"Are we impressed?

How could we not be? ..."

They impressed at Advertising money Sony throw that them... I guess,

**Rondec** · 11-15-2017, 04:24 AM

Originally posted by pinholecam

Another site that I often ask folks to abandon (its forums) and just let them loose the Pentax userbase there totally. (not that they really value it anyways)

I do recall the few articles they had about the K1 (and pixel shift) and the underlining tone was always that it was left wanting due to limited use situations and no Adobe support.
This A7RIII is different in what they are saying now :

"Are we impressed?

How could we not be? Landscape, cityscape and architecture photographers will absolutely love this new feature paired with the already excellent sensor in the a7R III - as long as they steer clear of (or clone out) moving objects in the scene. The increase in resolution and decrease in aliasing Pixel Shift brings is obvious in both our studio scene and real world result. It's frankly dramatic in the latter.
Are we impressed? How could we not be? Landscape, cityscape and architecture photographers will love this

There can be obvious artifacts in anything moving though, so that's a potentially significant (albeit expected) caveat for landscapes with motion (water, fast clouds), telephoto shots prone to movement from wind and vibrations, etc. You'll want to use a sturdy tripod with a remote release or self-timer. Furthermore, for now, using Sony's 'Imaging Edge' software is clunky, but once Adobe incorporates support, we can't wait to start shooting landscapes and perhaps tougher subjects in this mode to see how well it copes."

So much of it is in the way things are phrased. I think the point that needs to be made with Sony's implementation is that this is a clunky version of exactly what Pentax has been doing for several years. At this point, Pentax's pixel shift is more refined (can be done in camera and the images are taken more rapidly) motion correction can be applied in camera as well producing a jpeg if desired.

I'm glad that Sony is doing this. In the long run it should produce better support for pixel shift in third party software. It's just that you can't have a glowing review of Sony's version if a significantly more refined Pentax pixel shift was disparaged just a year and a half ago.

wanderer2 · 11-15-2017, 05:30 AM

<<It's just that you can't have a glowing review of Sony's version if a significantly more refined Pentax pixel shift was disparaged just a year and a half ago.>>

Those are also my thoughts. Thanks to everyone for the comments.

serothis · 11-15-2017, 09:50 AM

It's not even just the phrasing. If I remember correctly, when they tested the K-1's pixel shift, they did it on a waterfall and then tried to process it in ACR which had (and I think still has) zero support for pixel shift. and then they were SHOCKED at the artifacts.

With the Sony they did their best to shoot a completely static scene. Maybe this is just a learning curve of knowing pixel shift's strengths.

EDIT: although I should say that I shoot flowing water occasionally with PS. RawTherapee does a better job than silkypix at motion correction.

surfar · 11-15-2017, 04:17 PM - **1 Like**

DPR...Dependable Phony REVIEW$$$$$$$$$$$$$$

Those guys are accountants...the more money they count the higher the review score.

ruggiex · 11-15-2017, 05:17 PM

They were so impressed that they forgot to mention about .5 second delay between shots.

**monochrome** · 11-15-2017, 05:53 PM

DPR can claim they aren’t biased all they want, but here they’re clearly demonstrating Confirmation Bias.

EDIT: I really don't like that Forum at all - the comments are a hot mess.

beholder3 · 11-16-2017, 12:45 AM

Originally posted by wanderer2

This latest from DPR reinforces that perception, at least for me.

Amazon/dpr is an advertisement job, similar to Dxo and others. Don't mistake it for anything close to journalism or neutral/professional.

I once wrote a personal product review there and a North Korean "moderator" censored/deleted it quickly since my view "did not fit to their experiences". It's a little propaganda outfit.

**Rondec** · 11-16-2017, 03:50 AM

The funny thing is that Rishi is in the comments telling folks that they always really liked Pentax's implementation of Pixel Shift, same as Sony's. They just didn't like Olympus's version for some reason. I remember things a little differently, but then again, I shoot with Pentax.

JohnBee · 11-16-2017, 04:52 AM - **1 Like**

Originally posted by wanderer2

A lot of us felt the DPR K-1 review was biased in several ways, pixel shift being one of them. This latest from DPR reinforces that perception, at least for me.

The funny about developing technologies such as this is that it comes with what I'd call an adjustment period. That is to say that while DPReview may seem biased at this point in time, I'd add that we can trace back the development of their stance as they came to understand the technology and it's limitations. Which I believe could be likened to a learning process. In short, when DPReview tested this on the K-3 and eventually the K-1, it becomes apparent that the reviewer was not happy with the results. To which I'd add, they tried taking various images in various conditions with limited success. Not based on the camera of course, but more specifically, on part of the users and their expectations. And so the response came across negatively as the reviewer was left wanting more(or less) ie, more benefits and less artifacts

Fast forward up until today and the reviewer in question now has established an experience to work with. Granted, the a7r III is nowhere near as sophisticated as the K-1, in where the reviewer didn't know what to look for. And guess what... we can clearly see that experience put to use in his latest review. ie, the reviewer set-up his test shots to maximize on the limitations of the technology(stills) as he then go-on to compare the results against his own past experiences.

Which leads me to contemplate the following;
Is it fair? Hardly
Is it common? Most definitely

Having said all that, my take on this is that Pixel Shift technology is here to stay and that Pentax has the upper hand. And while it may feel as though the K-1 got a bum-wrap as the control. It remains that the next iteration to come-out of Pentax's doors will go unnoticed - this only works once

beholder3 · 11-16-2017, 08:35 AM

Originally posted by Rondec

The funny thing is that Rishi is in the comments telling folks that they always really liked Pentax's implementation of Pixel Shift, same as Sony's. They just didn't like Olympus's version for some reason. I remember things a little differently, but then again, I shoot with Pentax.

Hm... if you like a girl that is the language you would use:

Quote:

Conclusion
The amount of detail in the Pixel Shift files is markedly better than ..., but for landscape work ... artifacts ... can still cause issues. I'm glad that things aren't quite as bad as we feared..., but even... shows some movement artifacts. And, ... not the most stable or easy to use software out there. 9 out of 10 times we tried to use it, we got an unusable image preview even at 100% - making editing cumbersome, if not impossible.... don't currently support ...at all.
The reality, then, is ... will have to resort to (the buggy and cumbersome) SilkyPix. That's a real limitation, but not an insurmountable one - you can perform rudimentary ...
In absolutely perfect shooting conditions ... impressive when movement isn't a factor.
It is worth noting, however, that if there is a chance of any movement occurring in the frame I would definitely recommend backing your files up with non-Pixel Shift

I am not even mentioning that only a complete fail would choose a constantly moving waterfall to test a feature which is declared as working best with stationary objects.
And "surprisingly" then someone turns to 100% static test charts and photos of concrete city buildings, when covering the same question for the brand he is a fan of.

11-14-2017, 05:38 PM - 1 Like	#1
wanderer2 Veteran Member Join Date: Jul 2016 Posts: 301	DPR Article on Pixel Shift I'm not sure if this has already been posted (it didn't turn up in a search) or if this is the right forum to post it, but Digital Photography Review released an article this week on the pixel shift resolution in the new Sony a7R III. DPR absolutely raved about how wonderful it is. What a stark contrast to their review of the K-1 last year in which they listed the pixel shift as a "con" only useful with stationary objects because the Adobe programs couldn't process the K-1's pixel shift with motion control files. A link to the article: https://www.dpreview.com/news/5645755619/sony-a7r-iii-pixel-shift-lifts-a-ve..._dpr_nl_285_18 A lot of us felt the DPR K-1 review was biased in several ways, pixel shift being one of them. This latest from DPR reinforces that perception, at least for me.

11-15-2017, 05:53 PM	#11
monochrome Loyal Site Supporter Join Date: Mar 2008 Location: Working From Home Photos: Gallery \| Albums Posts: 26,276	DPR can claim they aren’t biased all they want, but here they’re clearly demonstrating Confirmation Bias. EDIT: I really don't like that Forum at all - the comments are a hot mess. Last edited by monochrome; 11-15-2017 at 10:54 PM.

11-16-2017, 04:52 AM - 1 Like	#14
JohnBee Veteran Member Join Date: Sep 2008 Location: Newrfoundland Photos: Albums Posts: 4,667	Originally posted by wanderer2 A lot of us felt the DPR K-1 review was biased in several ways, pixel shift being one of them. This latest from DPR reinforces that perception, at least for me. The funny about developing technologies such as this is that it comes with what I'd call an adjustment period. That is to say that while DPReview may seem biased at this point in time, I'd add that we can trace back the development of their stance as they came to understand the technology and it's limitations. Which I believe could be likened to a learning process. In short, when DPReview tested this on the K-3 and eventually the K-1, it becomes apparent that the reviewer was not happy with the results. To which I'd add, they tried taking various images in various conditions with limited success. Not based on the camera of course, but more specifically, on part of the users and their expectations. And so the response came across negatively as the reviewer was left wanting more(or less) ie, more benefits and less artifacts Fast forward up until today and the reviewer in question now has established an experience to work with. Granted, the a7r III is nowhere near as sophisticated as the K-1, in where the reviewer didn't know what to look for. And guess what... we can clearly see that experience put to use in his latest review. ie, the reviewer set-up his test shots to maximize on the limitations of the technology(stills) as he then go-on to compare the results against his own past experiences. Which leads me to contemplate the following; Is it fair? Hardly Is it common? Most definitely Having said all that, my take on this is that Pixel Shift technology is here to stay and that Pentax has the upper hand. And while it may feel as though the K-1 got a bum-wrap as the control. It remains that the next iteration to come-out of Pentax's doors will go unnoticed - this only works once Last edited by JohnBee; 11-16-2017 at 05:58 AM.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
K1/Samyang 24mm Tilt/Shift and Pixel Shift	DDoram	Pentax K-1 & K-1 II	6	12-14-2021 06:47 PM
DPR Pixel Shift Comparison	quh86	Pentax K-1 & K-1 II	9	05-07-2016 03:26 PM
Pixel Shift Corruption?	amoringello	Digital Processing, Software, and Printing	19	01-28-2016 03:59 PM
Imaging Resource Article on Ricoh Pixel-Shifting	mgvh	Pentax DSLR Discussion	26	03-30-2015 05:03 AM

11-14-2017, 05:54 PM - 1 Like	#2
stevebrot Otis Memorial Pentaxian Join Date: Mar 2007 Location: Vancouver (USA) Photos: Gallery \| Albums Posts: 42,007	Originally posted by wanderer2 A lot of us felt the DPR K-1 review was biased in several ways, pixel shift being one of them. This latest from DPR reinforces that perception, at least for me. Hmmm...Are we to be discussing the A7R III, Sony's pixel shift implementation, or DPReview? The DPR review of the K-1 has been thoroughly discussed on other threads. Steve

11-14-2017, 05:55 PM	#3
Rondec Loyal Site Supporter Join Date: Mar 2009 Location: Gladys, Virginia Photos: Gallery Posts: 27,653	They do mention that the K-1 produces similar results, but they sure did pound the K-1's pixel shift feature as being useless for real world situations when it came out.

11-14-2017, 07:02 PM - 1 Like	#4
pinholecam Pentaxian Join Date: Oct 2009 Location: North Photos: Gallery Posts: 4,706	Another site that I often ask folks to abandon (its forums) and just let them loose the Pentax userbase there totally. (not that they really value it anyways) I do recall the few articles they had about the K1 (and pixel shift) and the underlining tone was always that it was left wanting due to limited use situations and no Adobe support. This A7RIII is different in what they are saying now : "Are we impressed? How could we not be? Landscape, cityscape and architecture photographers will absolutely love this new feature paired with the already excellent sensor in the a7R III - as long as they steer clear of (or clone out) moving objects in the scene. The increase in resolution and decrease in aliasing Pixel Shift brings is obvious in both our studio scene and real world result. It's frankly dramatic in the latter. Are we impressed? How could we not be? Landscape, cityscape and architecture photographers will love this There can be obvious artifacts in anything moving though, so that's a potentially significant (albeit expected) caveat for landscapes with motion (water, fast clouds), telephoto shots prone to movement from wind and vibrations, etc. You'll want to use a sturdy tripod with a remote release or self-timer. Furthermore, for now, using Sony's 'Imaging Edge' software is clunky, but once Adobe incorporates support, we can't wait to start shooting landscapes and perhaps tougher subjects in this mode to see how well it copes."

11-14-2017, 07:15 PM - 3 Likes	#5
pakinjapan Veteran Member Join Date: Jun 2012 Location: Tokyo Photos: Gallery Posts: 1,041	Originally posted by pinholecam I do recall the few articles they had about the K1 (and pixel shift) and the underlining tone was always that it was left wanting due to limited use situations and no Adobe support. This A7RIII is different in what they are saying now : "Are we impressed? How could we not be? ..." They impressed at Advertising money Sony throw that them... I guess,

11-15-2017, 04:24 AM	#6
Rondec Loyal Site Supporter Join Date: Mar 2009 Location: Gladys, Virginia Photos: Gallery Posts: 27,653	Originally posted by pinholecam Another site that I often ask folks to abandon (its forums) and just let them loose the Pentax userbase there totally. (not that they really value it anyways) I do recall the few articles they had about the K1 (and pixel shift) and the underlining tone was always that it was left wanting due to limited use situations and no Adobe support. This A7RIII is different in what they are saying now : "Are we impressed? How could we not be? Landscape, cityscape and architecture photographers will absolutely love this new feature paired with the already excellent sensor in the a7R III - as long as they steer clear of (or clone out) moving objects in the scene. The increase in resolution and decrease in aliasing Pixel Shift brings is obvious in both our studio scene and real world result. It's frankly dramatic in the latter. Are we impressed? How could we not be? Landscape, cityscape and architecture photographers will love this There can be obvious artifacts in anything moving though, so that's a potentially significant (albeit expected) caveat for landscapes with motion (water, fast clouds), telephoto shots prone to movement from wind and vibrations, etc. You'll want to use a sturdy tripod with a remote release or self-timer. Furthermore, for now, using Sony's 'Imaging Edge' software is clunky, but once Adobe incorporates support, we can't wait to start shooting landscapes and perhaps tougher subjects in this mode to see how well it copes." So much of it is in the way things are phrased. I think the point that needs to be made with Sony's implementation is that this is a clunky version of exactly what Pentax has been doing for several years. At this point, Pentax's pixel shift is more refined (can be done in camera and the images are taken more rapidly) motion correction can be applied in camera as well producing a jpeg if desired. I'm glad that Sony is doing this. In the long run it should produce better support for pixel shift in third party software. It's just that you can't have a glowing review of Sony's version if a significantly more refined Pentax pixel shift was disparaged just a year and a half ago.

11-15-2017, 05:30 AM	#7
wanderer2 Veteran Member Join Date: Jul 2016 Posts: 301 Original Poster	<<It's just that you can't have a glowing review of Sony's version if a significantly more refined Pentax pixel shift was disparaged just a year and a half ago.>> Those are also my thoughts. Thanks to everyone for the comments.

11-15-2017, 09:50 AM	#8
serothis Veteran Member Join Date: May 2016 Photos: Gallery \| Albums Posts: 538	It's not even just the phrasing. If I remember correctly, when they tested the K-1's pixel shift, they did it on a waterfall and then tried to process it in ACR which had (and I think still has) zero support for pixel shift. and then they were SHOCKED at the artifacts. With the Sony they did their best to shoot a completely static scene. Maybe this is just a learning curve of knowing pixel shift's strengths. EDIT: although I should say that I shoot flowing water occasionally with PS. RawTherapee does a better job than silkypix at motion correction.

11-15-2017, 04:17 PM - 1 Like	#9
surfar Pentaxian Join Date: Jan 2015 Location: mid nth coast,nsw Photos: Gallery Posts: 6,141	DPR...Dependable Phony REVIEW$$$$$$$$$$$$$$ Those guys are accountants...the more money they count the higher the review score.

11-15-2017, 05:17 PM	#10
ruggiex Veteran Member Join Date: Feb 2010 Location: East Bay Area Photos: Gallery Posts: 811	They were so impressed that they forgot to mention about .5 second delay between shots.

11-16-2017, 12:45 AM	#12
beholder3 Pentaxian Join Date: Dec 2011 Posts: 3,112	Originally posted by wanderer2 This latest from DPR reinforces that perception, at least for me. Amazon/dpr is an advertisement job, similar to Dxo and others. Don't mistake it for anything close to journalism or neutral/professional. I once wrote a personal product review there and a North Korean "moderator" censored/deleted it quickly since my view "did not fit to their experiences". It's a little propaganda outfit.

11-16-2017, 03:50 AM	#13
Rondec Loyal Site Supporter Join Date: Mar 2009 Location: Gladys, Virginia Photos: Gallery Posts: 27,653	The funny thing is that Rishi is in the comments telling folks that they always really liked Pentax's implementation of Pixel Shift, same as Sony's. They just didn't like Olympus's version for some reason. I remember things a little differently, but then again, I shoot with Pentax.

11-16-2017, 08:35 AM	#15
beholder3 Pentaxian Join Date: Dec 2011 Posts: 3,112	Originally posted by Rondec The funny thing is that Rishi is in the comments telling folks that they always really liked Pentax's implementation of Pixel Shift, same as Sony's. They just didn't like Olympus's version for some reason. I remember things a little differently, but then again, I shoot with Pentax. Hm... if you like a girl that is the language you would use: Quote: Conclusion The amount of detail in the Pixel Shift files is markedly better than ..., but for landscape work ... artifacts ... can still cause issues. I'm glad that things aren't quite as bad as we feared..., but even... shows some movement artifacts. And, ... not the most stable or easy to use software out there. 9 out of 10 times we tried to use it, we got an unusable image preview even at 100% - making editing cumbersome, if not impossible.... don't currently support ...at all. The reality, then, is ... will have to resort to (the buggy and cumbersome) SilkyPix. That's a real limitation, but not an insurmountable one - you can perform rudimentary ... In absolutely perfect shooting conditions ... impressive when movement isn't a factor. It is worth noting, however, that if there is a chance of any movement occurring in the frame I would definitely recommend backing your files up with non-Pixel Shift I am not even mentioning that only a complete fail would choose a constantly moving waterfall to test a feature which is declared as working best with stationary objects. And "surprisingly" then someone turns to 100% static test charts and photos of concrete city buildings, when covering the same question for the brand he is a fan of.