RiceHigh's Pentax Blog: I Want to Buy the K-5, But..

Wednesday, November 17, 2010

I Want to Buy the K-5, But..

Still, I needed to do my research before decided. Even though I knew about the 10 major shortcomings of the K-5 and Pentax, and even though the K-5 has got the highest ever DxOMark amongst all APS-C DSLRs, I am still looking at the IQ of the K-5 in *real-life*.

Yesterday, I browsed through 10 pages of the Japanese Pentax Photoble samples database for K-5, download them to inspect (all are unedited originals and in full size), what I would say is that I am totally not impressed! :-(

Most of the pictures are lacking contrast and sharpness, images are soft, colours are odd, sometimes focuses are out and so on.. What a big disappointment for those real-life photos from those real K-5 users indeed! >:-((

And then, I've simply backed to the pixel peeping mode for some more scientific and systematic comparisons, I've re-visited DPR's studio photos with its new comparison tool, which I just reported a few days before, here:-

http://www.dpreview.com/reviews/canoneos60d/page17.asp

I replaced the 50D with the K-x and surprisingly found that the K-x just had better IQ than the new K-5 flagship! :-o Don't believe? Won't believe? Unbelievable? Below is a screen capture of the comparison!

As you can see from the above: colours, sharpness, contrast, noise, details are *all* better for the K-x image! :-o (Oh, Well..)

In fact, last time when I saw the DC Watch first review on the K-5, I had already disappointed with all those posted samples. Yes, I'd downloaded those original samples to inspect, already. Btw, I still recall that DC Watch usually took some best sample images amongst all review sites on this planet!

So, my purchase plan is now withheld. Unless I could see more positive samples from the K-5 in the future, I won't re-consider the K-5! :-( Just don't tell me that K-5 has low high ISO noise. Yes, I knew from the very beginning! I just wish to ask: What else is more important than low ISO Image Quality for a DSLR??

The most unfortunate thing is that last time K-r have already failed my tests, which I indeed was also interested to buy. Now, what new Pentax DSLR body in the Pentaxland can I consider and acquire?! ]:-((

More:-

High K-5 DR of the DxOMark: (Hardware) Mathematical Trick of the New Sony Sensor?

K-7 Wins Over K-5 in IQ (Specifically)!

Pentax DSLRs Require Fast Cards for A Bit Longer Video Recording!

Scratched Sensors of Brand New K-5 Units

K-r Vs K-x ISO Measurbations (Head to Head)

Comments (17)

Logging you in...

Dashboard | Edit profile | Logout

Logged in as

+3 Vote up Vote down

skina · 747 weeks ago

Let me tell you, high dynamic range can make the color contrast dull because the sensor captures more information in the shadow and highlight.
It's the task of post processing to map the information. Shifting histogram/curve left-right up-down.

5 replies · active 747 weeks ago

-3 Vote up Vote down

RiceHigh 110p · 747 weeks ago

I know, but if the high DR is achieved by the trick of lowering the contrast. Then I have to Say NO! :-(

We buy cameras for real-life shooting, not for obtaining high benchmarks!

0 Vote up Vote down

keithj · 747 weeks ago

Actually, looking at the test results, I agree with rhermans regarding the fact that the K-5 image is over-exposed relative to the K-x. I think it's something thats often overlooked when doing comparisons: over-exposure (even slight) will result in apparent lack of contrast and saturation.

But regarding the DR, it must be true to say that, the more DR you try to portray in an image, the lower the contrast for any given object. Now this may well work against you if you're wanting a "punchy" result, but all it means is extra work in PP.

I have a Fuji P&S camera, which allows DR expansion by taking 2 concurrent (yes, concurrent!) shots, each using different shutter speeds. Expansion can be up to 800%, or 3 stops, and I will say that, if you use this setting, you should expect to see a lower-contrast image. However, all the information is there. You just need to modify the tone curve in PP if you want to bring out subject detail, or sky detail or maybe shadow detail - the latter two of which would be impossible with standard DR.

Another thing which I think should be taken into account is the fact that all the DPR shots were presumably taken using different lenses, and of course this will also affect sharpness/colour/contrast (even if the lenses are nominally the "same").

+1 Vote up Vote down

skina · 747 weeks ago

I saw most K-5 users are happy with their purchase/upgrade.

0 Vote up Vote down

jaad75 · 747 weeks ago

Once again you just don't understand that "trick". Real life shooting is a proof of high DR, but you're just too blind too see, or trying to justify your cheapness.

0 Vote up Vote down

Anu · 747 weeks ago

No offense meant, but what you say above makes absolutely no sense - and this is not the first time. Please do try to learn about the camera and sensor technology a bit more. I do not mean to offend you, but your knowledge of this technology is very limited, far too low for you to write a blog which is supposed to have "Technical Articles on DSLR Gear and Technologies" in-between Pentax-news and criticism (wether warranted or not). You are not proficient enough in this field to write technical articles.

I am aware that you're an experienced photographer and used to be an avid Pentax user, but this does not give you understanding of today's technology, so it would be great, if you had the honesty of not pretending that you did. Your value for the community would go up significantly, and your criticism would have far more weight in it if you actually knew what you were wriring about.

0 Vote up Vote down

rhermans · 747 weeks ago

Ok, on the picture you showed here there is a difference and the k-x looks like it has more detail. So I downloaded the two jpg's and messing around a bit in photoshop I noticed a difference in exposure of 0.12 so I dropped the exposure of the K-5 to that and ... the difference is already getting a lot less. Then took the originals found the brightest white spot in those images and adjusted the curves with the same spot as white reference. The result is that they look almost the same, only the higher resolution of the K-5 gives it more detail at 100%. Both resized to 6MP I can't find the difference. (did resize both so both had the same interpolation and other *** to loose detail.)

So imho the difference has a lot to do with the settings of the camera and the jpg's they produced and not with the quality of the sensor.

Rice or anyone who wants to answer this or point me in the righ direction.
I'm still figuring out how to interpret the DR. DxO gives the camera a DR of 14EV at lowest iso. (landscape mode) Now I would like to know how this DR changes when raising the iso. I know it gets smaller, but is there a point when you lose so much DR when raising the ISO that it would be better to Keep the ISO setting and in post production just push the image.

+1 Vote up Vote down

ArcticExposure 45p · 747 weeks ago

This weekend, most likely, I'll be taking some pictures with both the K-x and the K-5 with the same lenses attached to compare. I'll post a link in here so you can see the results.

1 reply · active 747 weeks ago

-1 Vote up Vote down

RiceHigh 110p · 747 weeks ago

Thanks in advance.

0 Vote up Vote down

garedagmad · 747 weeks ago

Yes, I agree with the others: K-X image is less exposed than K-5, so its black looks more black resulting in a more contrasted image, and also a bit more saturated. Also it has lower resolution, so noise is not being shown (although I neither can see noise on K-5), and the image seems more neat, but again, it is due to a higher sharpenned image: take a look at the mountains shape: blue sky is soft until the end for K-5, even for D7000, but a white line halo appears in K-x and 60D images.

-1 Vote up Vote down

Dimitri · 747 weeks ago

I would rather wait till FW-update befor buying K5. I have a K5 for a short test and the write times of the JPGs (! NOT RAWs!) on the class 10 SDHC were just awful. I do not know if it was jast a bad K5 unit or a system error. Can somebody post own experience with SD write times of the K5 please, TIA!

1 reply · active 747 weeks ago

+1 Vote up Vote down

RiceHigh 110p · 747 weeks ago

Have you tried another card? If problem persists, contact Pentax.

0 Vote up Vote down

Anu · 747 weeks ago

So, scientific tests of camera's capability is not as important for you when you consider buying a camera, than some folks real life test pictures with more unknown variables than known ones... Fact is that K5 (and D7000) do perform extraortdinarily well and are superior image capturing devices (vis-a-vis image quality) than any other APS-C sensor camera. The fact that 99% of photographers could not take a representative shot even if their life depended on it, doesn't change the fact.

Still, I'll probably skip K5 as my next camera will probably be one without a mirror, at least if a decent size sensor is combined with an image stabilizer.

0 Vote up Vote down

jaad75 · 747 weeks ago

Rice, is it really that hard to see, that all Dpreview's K-5 samples are all front focused? Plus JPG comparison is meaningless - nobody who's really into photography pays no attention to JPEG output.

1 reply · active less than 1 minute ago

0 Vote up Vote down