Web Analytics RiceHigh's Pentax Blog: DC Watch K-5 First Review (Loads of Original Samples!)

Wednesday, November 10, 2010

DC Watch K-5 First Review (Loads of Original Samples!)

http://dc.watch.impress.co.jp/docs/review/newproduct/20101110_405733.html

(Japanese, Google English Translation Here)

As usual, we have full ISO samples, with different NR settings. And, there are also images taken for the same scene with different colour modes (Custom Image in Pentax' term). There are also some other real-life samples. The most important and meaningful thing is that they are all original in full size, directly out of the camera!

Btw, I don't see the super 14EV high range DR in action, at least the highlight is clipped as usual, as for most digital SLRs:-

http://dc.watch.impress.co.jp/img/dcw/docs/405/733/html/122.jpg.html

In contrast, I recall that my 5D still preserves better highlights than the above K-5 sample, see:-

http://ricehigh.blogspot.com/2009/03/some-new-macro-photos-to-share-dfa-100.html

My above bee on white flower photos under the bright sunlight are very similar to the situation of that K-5 sample, as it can be easily seen, the highlights and details of mine are just preserved better.

Besides, I also see the new King of Purple Fringing, just see this sample:-

http://dc.watch.impress.co.jp/img/dcw/docs/405/733/094.jpg
(Warning: Large File in 12MB!)

For your easy viewing, I provide a full-sized crop of the above sample:-



Finally, there are also two original video clip samples that can be downloaded. I do not feel good for the video quality nor the video efficiency. E.g., this first train bridge video is about 300MB for only 30 seconds. As for most of the time there is nothing in action within the frame and even though there are some, the moving portion is yet small, I am quite sure if there is some kind of video motion compression, the file could be drastically decreased.


Read Also:-

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

Ten Reasons for Thinking Twice before Decided on a K-5..

Comments (13)

Loading... Logging you in...
  • Logged in as
... any Idea wich lens was used for the King of CA???
2 replies · active 751 weeks ago
Any evidence then? At least I can't see it the last time I measurbated, for 3 different versions of the Pentax 18-55/35!
http://ricehigh.blogspot.com/2010/03/standard-zoo...
So you don't even have a clue what's the difference between PF and the regular Lateral CA? Your credibility has reached below zero levels...
I'll wait with reading this. My K-5 shipped today :)
1 reply · active 751 weeks ago
Congrats! Enjoy your new toy then!
Rice, you do realize, that dynamic range is not the same as headroom? If the highlights have clipped, the photogreapher has overexposed the images and not taken advantage of the DR.

I do suspect that Pentax autoexposure programs don't take advantage of the high DR to preserve the highlights, as that would fool several pixelpeeping review sites into thinking that the sensor is noisier than what it is (ie, it exposes roughly the same the competition does, instead of exposing less, thus preserving the highlights, and having a curve 'pushing' the midtones and shadows to compensate for the lesser exposure - this would take advantage of the lower read noise of the sensor and still offer as good midtones&shadows the competition does, instead of offering better ones, like now, when exposed identicallly).

Anyhow, I am sure the camera has an option to expand the DR of the jpgs (as my old K20D has it), and raw-shooters ought to understand to expose as required instead of blaming the camera for underperforming in unproficient hands.
2 replies · active 751 weeks ago
DR includes *both* highlight headroom and shadow deepness.
No Rice, DR is a difference between the deepest shadows and clipped highlights. You simply don't have a clue what you're talking about.
K-5 headroom is about 4EV above middle grey. That means you can use spot metering on hightlights and use +4EV compensation, to preserve the most information in the shadows. It's just your stupid idea that camera should meter exactly where middle grey is. Lightmeter is just a tool and you decide how to use it.
And a proof? Ok. here you have +4EV from a screen background (BTW, the scene contrast is 11.3EV and Blacks are at 10 if you know what that means): http://i140.photobucket.com/albums/r25/jaad75/K-5...
If I read correctly, the picture was out of camera JPG.
The high DR of K-5 sensor shall be obtained from RAW instead.
bandymelis's avatar

bandymelis · 751 weeks ago

Yes, the image of the butterfly is obviously overexposed (overexposure is the mistake Canon and Nikon cameras do and for this reason they look overly better exposed). My K10D underexposes more often then overexposes. This is an advantage, as highlights are retained.
Another thing - chromatic abberations. These are do to the 18-55 lens used for testing. Not a good lens to decide about CA's, as it produces a lot of those. Take a standard 50 mm lens and then check for CA's.
Otherwise, the results seem fine.
?????????
To me this sensor is a step in the wrong direction. All the numbers may look good, but has more plasticky feel in terms of sharpness and resolution. K-7 had more film-like feel. K-7 is more of a camera in my opinion.
USUK SendMoney's avatar

USUK SendMoney · 751 weeks ago

You need new/good lenses... and that you don't know you need new/good lenses and disparage a fantastic camera instead speaks volumes about you.

Post a new comment

Comments by