Web Analytics RiceHigh's Pentax Blog: DPR K-3 Studio Test Shot Comparison has been Out

Sunday, November 24, 2013

DPR K-3 Studio Test Shot Comparison has been Out

For instance, this is K-3 Vs D7100 Vs K-500 Vs A7 at ISO 1600 in RAW for the centre of the test shot image:

http://www.dpreview.com/reviews/image-comparison?attr18=daylight&attr13_0=pentax_k3&attr13_1=nikon_d7100&attr13_2=pentax_k500&attr13_3=sony_a7&attr15_0=raw&attr15_1=raw&attr15_2=raw&attr15_3=raw&attr16_0=1600&attr16_1=1600&attr16_2=1600&attr16_3=1600&normalization=full&widget=1&x=0&y=0

So, measurebate endlessly yourself to go figure all the differences for all other combinations of different camera models and at different ISO and etc. I blog, you decide!

Comments (13)

Loading... Logging you in...
  • Logged in as
Anonimity is Great's avatar

Anonimity is Great · 589 weeks ago

Against the D7100, the K-3's IQ is generally same, though there are some areas where the K-3 wins a bit. This is to be expected.

Against the OM-D EM-1, clearly the K-3 wins. Anywhere you look the K-3's is more detailed and less noisy. Again, to be expected.

Against the D610, clearly the D610 wins. About a stop less noisy and more detailed. Once again, to be expected.

What is not expected is against the Sony A7. There are no clear advantage from the A7! It is a bit cleaner than the K-3, however there is also less detail! This is surprising. It seems the A7 apply too much noise reduction even in RAW.

So, in conclusion if you're looking for FF image quality, don't buy the A7! Go for the D610 at least! However, if you're looking for the best new camera for under 1500 dollars, no doubt the K-3 it is.

Note: comparison done on ISO3200 and 6400
4 replies · active 589 weeks ago
How about ISO 1600 and below? What do you think?
Anonimity is Great's avatar

Anonimity is Great · 589 weeks ago

At 1600 my first evaluation still apply, although there's much less difference between each camera, and the detail on the A7 is now on the same level of the others.

Rankings For Noise:
D610>A7>K-3=D7100>EM-1.

Rankings For Detail:
D610>A7=K-3=D7100>EM-1

Honestly, if you are only using up to ISO1600, even in the dark, I see no reason to choose the more expensive cameras, unless there are specific features that you really, really need that the K-3 doesn't have.
SpentalotonPentax's avatar

SpentalotonPentax · 589 weeks ago

Rice. You may have changed my mind about the A7r. Has to be said the K-3 is doing extremely well there. This is before firmware updates for either camera mind.
I do not know about you, but the ISO1600 image of the k3 (that is displayed inside the comparometer you linked to) seems to be a little out of focus and / or blurred ( might be either because of motion blur {e.g. un-steady tripod ; shot without SR-off and mirror lock-up ; mirror-slap} / miss-focused / de-centered lens / wrong lens / AA-blur=on istead of 'off')

Center looks fine but the 7100 image with a sensor of basically comparable spatial resolution should not work out to give that much more detail / sharp differentiation than the K-3's sensor does. My point of focus: center of bottom-left rectangle with the feather and green plant target. Whatever it is (might be even field curvature in the lens' optical formula) either i cannot see the light for the life of me (I just cleaned my monitor to make sure I can) or there appears to be something amiss in this set of test-shots (I have the strong suspicion the actual test-shot displayed for the K-3 does not represent what it can do).

On the plus-side: the test-scene that dpreview use here exhibits so much fine detail - and all the cameras I went through in the display-boxes handle it so damn-well when put into "global perspective" -- I think the minor variance in these sample-shots taken will tell every sane person: Them DSLRs and mirrorless cameras are all good-enough, already!

Perspective - something to consider. We all are shooters, after all - Photographers should honor perspective as much as maximum-detail (pun intended)!
3 replies · active 589 weeks ago
to back up my claim: check that spot for the native ISO, as well: These feathers joining in on the green stuff don't appear to sharpen up one bit. Either the lighting (which seems consistent to my eyes) is off for the K-3's shot or my first impression and thought-process hits the nail on its head!

Bottom line:
The image is not softer because of the sensor - but it sure is softer, albeit - I suspect - for a different reason. Since the ISO-100 shot does not reveal any finer detail than the ISO-1600 shot (1/16th the amount of light captured). Might even be the RAW-processing at fault here - who knows? I did not take the shot - so I cannot tell definitively.
All based on the belief that Sony would supply a new sensor that does worse for per-pixel-fidelity than its previously used ones.
Anonimity is Great's avatar

Anonimity is Great · 589 weeks ago

I still think it's the AA blur on. I can see on the fine lines the K-3 has less moire than the even the FF D610 and A7. Now this begs another question. I thought the K-3 is just a bit less noisy than the D7100 in many parts of the pictures (not all), could this be due to the AA blur? Is it possible the AA blur also gives us noise reduction? :D
good question. However: Rice's original criticism was about noise-patterns and noise-related image-rendering in general. In that respect: sharpness and/or brilliance of the sample-shot won't matter as much as for a 'real' photograph. Maybe I let my words slip ahead of myself a little too much, thereby leaving the original point of interest.

However: I say the next noteworthy bump in image-rendering (read: not just noise-performance) will eventually set in when applied sensor-technology is leaving the "Bayer"-Color Matrix. The Sigma Merill-Chip (ex-"Foveon") as of right now is the nearest thing money can buy to a full-color-pixel design - yet still seems to exhibit flaws in the two fields that are either a popular differentiator or need at the moment: high-ISO shooting and acurate color-reproduction.
For photographs that don't require high-ISO and/or exact, life-like color-reproduction but benefit from precisely what a certain group of people characterizes as "technical flaws", that chip already provides a viable alternative - unfortunately exclusively in a Sigma camera.

But there are other full-color-pixel designs in the works right now. Thing is: Because Bayer-Chips have matured to the highly-developed state they are in, now, they are good-enough for the majority of shooting-scenarios the world throws at them.
As I understand it: in a global economy where economy of scale and mass-market availability is king: "good enough" is just about enough reason for other attempts at sensor-design to have a hard and slow road ahead (as in funding and timing of Research&Development), once a competing technology has made its way into the mainstream.

But eventually I hope that the diffraction limit will tempt the die-hard enthusiasts to yearn for even better "image quality" from the same size of photo-sensitive area - and camera makers will try and meet their customers needs and produce cameras with another, not-flat color-matrix sensor technology.

It is my personal understanding that on precisely that day, where these chips will be adopted globally, all of our better lenses will receive the next considerable leap forward in image quality - basically "for free". Hopefully Pentax/Pentax-Ricoh will still be alive and kicking, making affordable K-mount and 645/6x7-mount cameras, then!

Cheers
Heiko
Just change the lightning conditions to Tungsten - it clearly shows pretty destructive noise reduction on the K-3 file. While it's slightly better than this from the K-500, it's still not much worse than fine grain with detail from the D7100. http://www.dpreview.com/reviews/image-comparison/...
1 reply · active 589 weeks ago
I meant "it's still worse than a grain from the D7100".
P.S. I knew that something critical about the K-3 will definitely go through your censorship, Rice...:D
I am surprised how much are the images from A7 and K-3 close to each other. On some settings i would prefer one, on other another, but in the end it's not about image quality, which is the about the same. I thing it's about features you prefer (DSLR/mirrorles, shallower depth of field/broader depth of field). It would be a hard time for me to decide between those two cams, what a luck they are out of budget for me at the moment!:-D

Post a new comment

Comments by