Tuesday, July 07, 2009

A Few Un-edited K-7 Samples for Noise Performance Analysis

With reference to my last report, more samples are being uploaded by new Japanese K-7 users at the Photoble:-

http://pentax.photoble.net/?L=en

Since some people are unconvinced that the K-7 is no better than the K20D in IQ and noise performance but could be even worse, there is no harm to re-visit the latest samples and see more objectively about the truth.

First look at these two pictures, the first is at ISO 200 by the K-7, the latter is at ISO 400 from the K20D. The Contrast and Sharpness settings are both at +1 and +4 respectively, NR is not applicable for both images:-

http://pentax.photoble.net/?exif=090707003 (K-7, ISO 200)

http://pentax.photoble.net/?exif=090706006 (K20D, ISO 400)

You judge yourself for which picture is more noisy!

Okay, if the ISO 200 picture of the K-7 is not good enough, let's look at the ISO 100 picture then, here is a recent example:-

http://pentax.photoble.net/?exif=090704012

The Sharpness setting is only at +1 this time. Can you see the noise? Still?

The link is always there and I recommend that anyone who are interested can always check back. In particular, I recommend to download those un-retouched samples with full EXIF and generated by the camera, instead of others that are edited and massaged and with many EXIF data destroyed. Just remember that all those uploaded samples are *Real*! And they are made by the production K-7 cameras that the new *end-users* (but not "reviewers"!) have just bought from the streets in Japan.

Nonetheless, I know that some people would still say or argue that we need to use RAW to get the best IQ out of a camera. So, still, you can look at those RAW converted images in the above end-user K-7 samples gallery, here is one shot at ISO 100:-

http://pentax.photoble.net/?exif=090704000

Now, the sharpness and details have been (much) improved, but the noise becomes so terrible. Remember, it is an ISO 100 shot! Just!

As for the FAKE green colours, look at the above shot and also this one, which is with the "Natural" custom image tone and produced in-camera:-

http://pentax.photoble.net/?exif=090704006

Again, anyone can inspect more and more of those new real user samples. I blog, you decide!

12 Comments:

Anonymous said...

Rice but you are real LOL, even can't compare corectly :D I can show you same examples from CANON

Anonymous said...

Just checking - have there been any industry/editorial reviews on K-7 using a production model camera, with the firmware level that people are taking delivery of?

Have any of the original reviews been re-published with the as-delivered performance level, and comments appropriate to that?

Per the extra sensor and added firmware, no doubt the SAFOX-VIII PLUS will outperform the K20D version of AF. What I want to see is a matchup between the new Sigma OIS lenses that mount on PK, vs. Pentax lenses in the same general focal range for zooms. Any differential in those tests will exclude the body's contribution, and it will be "lens vs lens". I have yet to see a published formal review of similar Pentax SDM vs. the earlier non-SDM lens in terms of AF speed.

I would guess that these IQ/Noise issues have to do with the "firmware level" faux-pas of what was originally tested - but like most very-late changes, it's very hard to get a noticeable betterment of what was already there through Firmware alone, unless some true "bug" existed. Panasonic delivered the GH1, and then issued a firmware upgrade.

Anonymous said...

Pentax Forums is noting some movement on the Samsung 7/7 announcements;, but I see no news coverage as yet.

At least someone in the industry is good at keeping secrets.

http://www.pentaxforums.com/forums/samsung-dslr-forum/62335-samsung-announce-july-7th.html

"I work in South Korea for a Samsung contractor company (INT Electronic), we make some semiconductor pieces.
We have been invited through a executive invitation card to special event to new products announcement on July 7th in Busan city.

Samsung will announce:
-NX3 Hybrid camera
-DSLR GX30
-WB700/HZ30W (Europe/USA market)
-Lenses and accessories"

Anonymous said...

This time you fail. I generally agree with your judgements but what I see in these pictures are sharpening artifacts from the camera and not noise. The K20D sample is pretty noisy by the way! Low ISO noise isn't the question anyways, the high ISO noise is the important factor. And since these are simply low ISO shots they are more or less useless for comparison. God knows where the noise comes from. Could be photosite response non uniformity or real "noise". Actually I don't give that much about noise anyways since all APS-C cameras appear to be on a similar level and differ mainly in noise reduction approaches. Wha would be considerably more intersting is this blue line bug and finally some comprehensive testing of the damn autofocus including the yellow light bug. It ruined several of my K20D shots last week that I shot under candle light. All have terrible front focus.

Anonymous said...

The K20D is known for perform best at ISO 400, it is likely that the K-7 is similar here. You state that the K-7 is worse than the K20D, but the K-7 does not have the high colour noise in shadows that the K20D has, and many users has reported that the K-7 does not have the banding problem that the K20D had and so on. So users that has upgraded from K20D to K-7 does see an improvement as noise are concerned.

And it is clearly an improvement over the K10D.

As for poor AF - the K-7 does not have poor AF. It is faster and more reliable in low light than the K20D, so users of K20D also sees an improvement here.

You claim that the K-7 is not much faster than the K-m, so??? Why should it be? Already the K-m is faster than the K20D! I don't see it as a disadvantage that the K-7 is slightly faster than the K-m which already is a very fast AF camera.

There has not been any professional tests made with production versions of the K-7, only pre-production samples.

Users with production K-7 do discover that the K-7 is clearly a big improvement over the K20D in many areas.

The K-7 does not have to be a machine gun camera. It is fine for most outdoor shooting and indoor too. Most Pentax users does not shoot fast action sports professionally. Pentax makes excellent cameras with excellent image quality.

Claiming that Pentax is bad because the K-7 is not a machine gun camera.... well well...
I guess that the new 645 digital is a very poor camera too then!!!

Do yourself a favour and buy a Casio p&s that does 60 fps, since this is all that seems to count for you.

Pentax is Pentax, they don't make "me too" products.

Anonymous said...

This photo EXIF shows saturation at +2. Guess green would be bright huh.... Pretty sure it's the same for the poppy as well.....
Check your facts please.
http://pentax.photoble.net/?exif=090704006

Anonymous said...

+4 sharpness will most likely cause artifacts and exaggerate noise...
you know NOTHING about digital photos do you.....

RiceHigh said...

The two +4 sharpness samples are meant to compare the noise levels between the K-7 and K20D. So, tell us which sample is better as far as noise is concerned. Mind you that the K20D sample is at ISO 400 whilst the K-7 one is only at ISO 200.

Anonymous said...

I thought you were going to stop posting about Pentax. Please do us a favor and stop ranting about equipment you haven't touched!!!

Anonymous said...

http://www.imaging-resource.com/PRODS/K7/K7THMB.HTM

Anonymous said...

I hate to say it but this comparison is totally flawed and unscientific. Compare apples to apples. I own neither camera but reading his comments are totally bogus.

Anonymous said...

dpreview says:
The final word
The Pentax K-7 manages to combine superb build quality, comprehensive customization options, excellent ergonomics and an extensive feature set with (for a camera in this class) very compact dimensions. This makes it a more than viable alternative for those (but not only for those) who like to work with semi-pro equipment that still leaves some space in the gear bag. JPEG output at high sensitivities is not quite on the same level as some of the competitors but if you revert to shooting RAW things are pretty much evened out.

Post a Comment

Related Posts

 
Creative Commons License
RiceHigh's Pentax Blog by RiceHigh is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported License.