via Information on Digital Cameras for comparing test images from the Imaging Resource by making crops and put them side-by-side.
Well, yet again, the pictures speak for themselves. How come all the crops from K-x just show that it wins over the "King of High ISO" and that "New King of APS-C DSLRs" for noise, colours and details, all in all?! :-o ;->
Related:-
I Want to Buy the K-5, But..
IQ and AF-C of the K-5 - I am Not Impressed! :-(
High K-5 DR of the DxOMark: (Hardware) Mathematical Trick of the New Sony Sensor?
K-r Vs K-x ISO Measurbations (Head to Head)
Logging you in...
Anu · 747 weeks ago
RiceHigh 110p · 747 weeks ago
Anu · 747 weeks ago
Anu · 747 weeks ago
Also, you should not look at cherry picked crops, but the images at the site. Looking at the model image's background, it is clear, Kx image has lots more noise reduction going on, yet the details seem sharper... very different processing. Regarding details - the acuity of Kx in these images seems better, but looking at the details of the fabric and other areas where sharpening doesn't come to play that much, but noise reduction does, it is clear that the K5, even in these conversions, does provide more actual detail.
The different WBs are evident if you just bother to look at the grey cards - making strong colour statements with such basic processing difference is just silly - either you are not aware of the influence of WB, making you totally incompetent to write this blog, or you are aware, which makes you biased to follow an agenda of Pentax bashing. I wonder which it is...
RiceHigh 110p · 747 weeks ago
Com'on, man. Why you are so reluctant to see the truth even hard facts and pictures are presented?
Anu · 747 weeks ago
Please answer to the following questions and tell why they don't matter and why Kx is still better and I should "see the truth" of the gospel you preach.
1. Is the WB the same in these images? And if not, why you make strong colour statements based on these images?
2. Is there lots of blurred chroma noise in Kx images especially in the background? Am I just seing things, or is it that you don't care about being objective?
3. Look at the fabric of the models clothing, all of the areas, not just some cherry picked spot - K5 provides easily better quality. Even you should be able to understand that it is a lot easier to process simple edged high contrast detail into looking good, than complicated low contrast area like fabric.
4. Look at the book's coloured area - again, K5 wins easily.
Maybe you should use your own brain instead of blindly telling how great some sites are if their images, or selected parts of them seem superficiallly to support your agenda.
DXO has by far the most scientificly accurate procedure when it comes to camera testing - not perfect, but far superior to anything that forces us to do subjective comparisons on images that have been processed in some unknown way.
I do not have the energy, the skills, the money, or the motivation to stat my own review site. Why should I? There are enough bad review sites, and I certainly can not do as proficient analysis as DXO does - note that they do this stuff to improve their raw-conversion software - informing us is only sencodary function (plus advertisement for their software).
RiceHigh 110p · 747 weeks ago
Camera tests are meant to find out the differences of different cameras, which are not to make them look equal!
Since you have highly blamed that the tests by IR and DPR are sub-standard and with faults of this and that, I doubt that if you are really the one that could do better tests. If you cannot show us anything, how should we believe you? Note that at least I do tests regularly and post it here myself..
And, post tests in the Internet is not anything difficult, as everyone are doing that now, everywhere at forums and *free* blogsphere.
Anu · 747 weeks ago
You claimed that the colours of K5 are inferior - yet you ignore the different WB of the images - either you don't like facts, or you don't understand them.
I have not claimed that IR produces sub-standard tests, please do not lie. DPR on the other hand is clueless when it comes to it's testing procedures. Pretty face of the site does not equal to understanding stuff.
I did ask you four questions of an IR image you used as evidence to advane you anti-Pentax agenda. Why not simply answer?
Anu · 747 weeks ago
The bottle crop is also a nice one - it is nicely cropped to provide partial information - is the blog you linked also made by you? If one crops a bit less, one might notice that the Kx has lots more (smudged) chroma noise, the book to the left of the bottle is more detailed and contrasty in the K5-image, especially in the coloured area, and so on.
Also, it looks like the K5 image is focused a little bit to the front of the Kx image, not much, but enough to make a little difference.
Anyhow, I am looking forward what kind of biased anti-Pentax twists will come out of your bashing mill in the future... one thing is for sure - objectivity will certainly be tossed to the trash can, right?
Barbarosa · 747 weeks ago
Erika · 747 weeks ago
K-5 looks very good at what it does. K-5 has 4 megapixel more than K-x and still manages to be about equal to the K-x (noise wise) when its pictures are scrutinized at 100% size.
That said K-5 got higher score than K-x on DxO. Most likely because it has more megapixels (as the DxO results are normalized to 8mp), and a tiny bit better noise reduction algorithms at the sensor and firmware level.
It is a question of time (and that time is close) when megapixel increase will start to degenerate IQ due to harder to manage noise and lens limitations (diffraction, sharpness etc.). The only way to get more MP and retain IQ will be to go Full Frame.
Anu · 747 weeks ago
When the lens will become the limiting factor (at mayby 200Mp or so for the better lenses, depending on the sensor size), we should be happy as then we no longer need the AA-filter.
You imply that sharpness would somhow go down with more pixels due to lens/diffraction. This is false - there will at some point be diminishing returns after which it makes little sense increasing the pixel count. Luckily we're still far from that point.
Erika · 747 weeks ago
When Tom Hogan reviewed Nikon D80 (10mp camera), he mentioned that diffraction starts to kick-in on it past F13. While for Nikon D90 (12Mp camera) the diffraction starts to kick past F11. I do not remember what lens he used though. I presume not the worst one.
I think that you are overly optimistic with 200MP. Maybe the limit is closer to 30-40MP than 200mp for APS-C size sensor.
A very good lens resolves 150lp/mm. K5 sensor size is 23.4 x 15.6 mm. 150lp = 300 lines. 23.4x15.6x300x300=33MP roughly.
Many lenses resolve just around 100lp/mm. Then: 23.4x15.6x200x200=14.6mp.
That said, Bayer sensor interpolation blurs the image so its resolution is lower than 1 line per pixel. Still though, the limit is not that far. Perhaps 20-25mp for relatively good lens (for APS-C sensor) and 35-40 for the really good one?
Syam · 747 weeks ago
Anu · 747 weeks ago
Also, diffraction is not a brick-wall. 100Mp sensor will outresoilve a 50Mp sensor even if the lens is used at f/32 - the difference will be insignificant though.
Also, you're making the mistake about resolution analysis - it does not matter what the individual pixels look like - only the resulting image is important.
You're oversimplifying the issue, I am afraid. For exanple in order to properly resolve lines, it take a Bayer-filter sensor more than just two pixels, depending on the demosaicing algorithm, and the direction of the line pairs.
Oh, also, you indicate on the top, that you think that diffraction depends on the lens used - this is not the case.
One can try to simulte the effect of smaller pixel pitch (ie. higher pixel count) by using a teleconverter. I have a decent 2x teleconverter which allows for pin sharp images with quite a few of my lenses. Effectively the lens is stressed as if I were shooting with a sensor with 4x the pixel count, in this case 14.6*4, over 50Mp. (K20D, a cheap Soviet made TC and for example East German Carl Zeiss Jena 135mm lens).
illdefined · 747 weeks ago
The same high ISO ability from a 16mp sensor as a 12mp one, but with higher resolution.
I'm sure when the DPR review comes out, you will find a way to dismiss the results if they are positive. just like you did with DxO Mark.
can you admit your negative bias?
Erika · 747 weeks ago
I think you do not read between the lines. Rice likes Pentax (else there will not be any blog about this), and criticizes it from time to time outlining its camera problems, keeping his posts controversial enough, and not painting the whole image.
This generates heated discussion and increases traffic to the blog.
So relax, and do not forget that English is not native language of Rice, so subtle irony might be lost in translation.
Cheers,
Erika
illdefined · 747 weeks ago
More importantly, all Rice has done with this purely negative site is inspire more negativity. Just look at the Recent Comments section in the right column, nearly all of the comments are of angry or upset people who don't respect him, his posts, or his attempt at "subtle irony". All Rice has done is make a site collecting negativity, seemingly for the only purpose of making himself feel important. Too bad he has become well-known, but not for anything good.