Friday, June 26, 2009

PhotographyBLOG K-7 Review is Out (Soft JPEGs, Noisy RAW Converted Images)

The PhotographyBLOG (PB) have their latest K-7 review published. Look at their image quality review page (page 2 of the report) will quickly reveal that the JPEGs are soft with much loss of details and decreased resolution (even at ISO 100) whilst the RAW converted pics are sharper and well defined for the image but however they are all noisy to very noisy at all ISO speeds (ISO 100 has visible noise and IMO ISO 800 is already not acceptable, IMHO). See:-

It seems to me that the K-7 is particularly noisy for the Red channel, just look at Yvon Bourque's ISO comparison shots tested with his Red Jeep as the target previously:-

Furthermore, the PB comment in their Conclusion that "Exposures were generally 1/3rd EV stop under-exposed, perfect for retaining detail in the highlight areas without sacrificing the shadow areas too much, and colours were accurate using the default Bright setting." This is the same as what WDC said previously for the underexposure problem as well as my observation here as well. And, the "preserve highlights" look like a too old but classic excuse for those Pentax DSLRs to be behaving drastically for exposure and metering, which this Pentax specific persistent issue is just too annoying and would cause much degrade to the IQ indeed. (For further information on how "expose to the Left" is just not a good idea and the adverse effect on the IQ, read here.)

Besides, the SR test shows that the SR does work for the test at 82mm focal length at a shutter speed about 3-1/3 stops slower than the theoretical safely shutter speed (i.e., 1/80th sec), but however it is not perfect - see the crop.

Near the end of the report, the PB has made some remarks on their test unit is a pre-production one with production 1.0 firmware. But so what? It is just a simple disclaimer with the existence of a disclaimer in mind IMO. Should we really hope that the "production" K-7 will magically be vastly improved over the "pre-production" one with the same version of the firmware? I think it is just purely a wishful thinking with a too high unrealistic expectation (and actually last hope) by many people recently, very frankly!

Okay and lastly, you can always inspect the posted samples yourself as they are there and make your own judgement and conclusions:-


IT said...

LOL even K-7 pre-production samples are different, wait for production unit !!!

hugues said...

"The Pentax K-7 is the best-ever Pentax DSLR camera and a real challenger to the relative giants of Canon and Nikon."
"....with the K-7, Canon EOS 50D, Olympus E-3 and Nikon D300 all at the same price point. In many ways the Pentax K-7 offers the best overall package out of them all, and is a really serious contender for your cash."

What else...?

Anonymous said...

How about a comparison relavent to today's movie-oriented market for buyers without a DSLR - that would include the video-enabled DSLR's; and the GH1 with that great lens. You can add the Samsung NX to that liast shortly.

The D300 is a great 2-year-old camera that is persistently high on the Amazon 100 list, even today - to be replaced by the D400 with video.

Comparing by price is cute; but comparing by features is more appropriate. You can get $800-$900 DSLR's with simple video these days, and without the sensor shift noise forcing use of an external mic for sound recording.

shmook said...

"Soft JPEG": if you want it sharper, you can turn up sharpness in camera. They even write it in the review. I prefer Pentax' conservative sharpening policy. It's easy to sharpen up a soft JPEG in postprocessing, but it's impossible to get rid of sharpening artifacts.

"The Red channel is particularly noisy": Hum, the only color in the crops they show is red, so this conclusion is not really an informed one.

Finally, what's the point of critisizing the not-yet-released final firmware based on the pre-production firmware?

Why so negative?

Anonymous said...

Mr Rice, this is a *pre production* camera - a *prototype*. Not the final production unit. It has *pre production hardware* not production hardware. Pentax has already said, most recent in a comment made by Pentax UK, that the production units has updated hardware compared to pre-production units.
Why don't you believe them?
It is standard and normal that production units can differ from production units. They make pre-production units to test the camera and to help them adjust the hardware for production unit.

It is just stupid to test pre-production units and it is stupid to judge production units from pre-production units.

Please wait for the production units, then we can have a discussion.

Oh, and also note that the pre production units has a pre-production sensor not the final production sensor. This is also official information from Pentax. Pentax has stated that there is difference between pre production sensor and production sensor, most notably regarding noise levels at high ISO.

As for "underexposure", yes *some* of the pre-production units is adjusted so, but there are also pre-production units not adjusted this way.

So the results varies between the pre-production units and we does not know yet how the final production unit is adjusted here, because there is no production unit yet released for testing!

Anonymous said...

You are making the same misstake as others - believing that it is all in the firmware, but it is not.
The *hardware* differs between pre-production and production. So a pre-production camera with production firmware is still a pre-production camera with pre-production hardware and not the final production hardware, not the final production sensor and not the final set in stone adjustment of the metering and other *hardware* related issues.

RiceHigh said...

> "Soft JPEG": if you want it sharper, you can turn up sharpness in camera. They even write it in the review. I prefer Pentax' conservative sharpening policy. It's easy to sharpen up a soft JPEG in postprocessing, but it's impossible to get rid of sharpening artifacts.

I am so tired of such defensive sentences from some Pentax users. People are just been unable to distinguish between sharpening and resolution & image details these days!

> "The Red channel is particularly noisy": Hum, the only color in the crops they show is red, so this conclusion is not really an informed one.

Both PB's red crops and YB's red crops are so noisy. I hope not this holds true for the Greens and the Blues!

> Finally, what's the point of critisizing the not-yet-released final firmware based on the pre-production firmware?

It is with production firmware in a "pre-production" unit. But since the "pre-production" units are almost the production ones (they are not prototypes!), what huge differences and improvements would you expect?

> Why so negative?

Why so over-positive and defensive? :-)

jaad75 said...

As always... Just compare the noise of last pre-production units that were shipped for testing. Falk Lumo and Frank have uploaded a lot of samples. Even ISO6400 is useful after some post-processing:

Jarda said...

I believe 1/8 of second is full 4 stops of efficient SR. You forgot to multiply by crop factor, RH.

Regarding the noise - camera with sensor based stabilization system can never archieve as low noise level as fixed sensor - they cannot put heavy cooler on the sensor.

But I believe they pushed the K-7 to the very limits of contemporary technology.

Anonymous said...

I've never seen such confusion thus close to a launch. If, as some suggest, there's a hardware issue - that might require some dis-assembly of complete (or even packaged) units; and the sensor sits within the sensor-shift.

Even a firmware reset might involved already-packaged units.

As contrast, Olympus held a huge press junket in Berlin, with top management attending, and gave out a number of EP-1's for the press to use quite freely.

Too many excuses being made - Pentax should have clarified all of this; not the user base.

Anonymous said...

Raise the ISO, shoot at 1/125th, and the internal sensor-shift reduces markedly as an issue. At slow speeds it has time to act - at high speeds, the shot is finished.

With strobe, the flash duration can be 1/800th or less - and that takes care of "freezing" the image.

An OIS lens stabilizes before, during, and after the shot, so you know what the framing will be (and was). The AF and AE should work faster, because the image is not wandering around while viewing, and the "spots" are more consistent. The longer the lens, the more benefit. Measuring a 50mm's performance proves very little, except that it helps someone unsteady, as compare to not having IS.

Olympus abandoned mechanical S/S for video mode, resorting instead to electronic stabilization.

Anonymous said...

Hey Ricehigh,

You seem always so negative. Why don't you sell your Pentax equipment and buy Nikon?

Anonymous said...

It's not Ricehigh's job to sell Pentax - it's Pentax' job. Check Google News for "Pentax K-7" in date order and see what comes up.

If the legacy Pentax owners would stop talking about how happy they are with a 2006 K10D and BUY something, Pentax would be a lot better off.

“Sell my K10 for a K20 then sell my K20 for a K7 then sell my K7 for a K whatever then trade that camera in on another K whatever. Consumerism is rampant. I barely use the total functions of my K10 and it still amazes me at some of the pictures my K10 takes.”

NoFx200X said...

I do like critsism but RH always tends to be on the negative side. IMO.
Nevertheless, I still read this blog every day because it serves me as navigational beacon in all that pre/post/whatever K7 Hype.
I'll buy one. 12 month from now. Thats clear. BUT I would love to have that RH Blog as my primary source for neutral information. News are great and critisism too. But the always reporting the glass half empty is bothering me.

Excuse my english, my german is hopefully better.

Anonymous said...

RE: "I do like critsism but RH always tends to be on the negative side."

Nonsense - he was excited about it, and placed his own order for it.

"I've never met him, but quite a few of you should cut RiceHigh some slack. I don't always agree with what he says, but he has every right to post his thoughts. He has strong opinions, and is passionate about what he does. Last time I checked, there's no restriction on having a point of view either in print or on-line."

After the hype of the announcement, the facts get understood better. The lenses are expensive; you need a $250 external mic for sound recording; the sensor swap/upgrade called the reviews into question; no zooms tested with video that I can find; restrictions on the testers for their reviews.

Not a good beginning. We don't even know for sure the relationship of image samples to the firmware that the buyer will receive. Were there no problem, Pentax could have issued a firmware update - that's what Panasonic did.

Pentax needs this camera; so I hope that the public just reads the reviews, rather than the news.

Anonymous said...

Guys, whats to like about the current K-7 image samples? There is no way you can sum them up without looking like a bitter troll in the end or an overly hopeful fanboy.
Personally, I wouldn't be surprised if the July release date got pulled due to noise and loss of detail issues seen in this current pre-production review.

Anyway, back to my precious k20d :)

jaad75 said...

Your "precious K20D" doesn't have an image quality better than K-7 and everything else is worse, so I don't understand what's to like about it...:P

nlx said...

Desoite i'm still hoping there will be a real test here with a real final unit (if not : whats the point of all that posts from the begining ??) and camera need to be tested in hand, not in pixel pipping in the internet.

For the rest, final firmware or not it seems Pentax compete with Caniikon on high iso. Thats a bit depective. Doesn't mean the camera is crap; but just on this specific point, it seems deceptive. I don't think a 1.0.1 update could really cange that now at that point, the difference is too big to be fixed at the very last time. Can be somewhat improved but just slighty, not dramatically.

nlx said...

sorry for mistyping, can't correct posts…

(and its Pentax CAN'T compete)

Mike Princz said...

here I am again.... long live my K100D !!!
By the way gentlemen, if you really need low noise, high resolution,etc, you will have to move to nikon's d3, period. Digital photography is still too young and the electronics involved in a DSLR are an act of compromise . We have to be patient.

Anonymous said...

If the man want to see only black sides, nobody can help him to see white. Rice has real psychologic problems. Sorry sight.

There is very good proverb:
"The pig can find mud everywhere"

Paul said...

I think I get it... This blog was meant to give a new dimension to "trolling".

This is actually kind of a proof of concept in itself. A troller usually uses his "subtle art of trolling" (you can google that) within fora. The readers will come to this fora to find informational posts and interesting discussions; and some of them will fall into traps set up by trollers. This leads to hilarious inflamatory arguments.

But here we are at the next level of trolling: readers do not need to be trapped while searching for information. They actually come here directly, voluntarily, knowing what they will find and probably secretly enjoying the arguments by anticipation.

To me RH is not a Canon guy bashing his competitor. He might just well be a new species of troller. We should definitely find a name for this king of "Troll Blog", and reference this blog on wikipedia for internet historical records.

And the most important lesson of all of this: this Troll Blog just works! We come here, read, disagree, argue, flame, explode... and then we come back... ;)

Rice: no bad intention in this comment. The only goal was to make some people smile!


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.