D-7000 vs. K-5?

[DuncanM1]

D7000: ISO6400



K5: ISO6400



D7000: ISO25600



K5: ISO25600



You can download the images and exif data from here:
http://picasaweb.google.com/Dunmunro1/K5TestImages#

darn, I don't know why that didn't load properly.

 

[ET2]

And yes, D7000 will also score around 1100 for lowlight, same as K-5. Overall, K-5 might end up scoring higher due to DR at ISO 80, but lowlight scores will be similar (around 1100 --- and A580 will also score around 1100 for lowlight).

Matters very little, The K-5 produces results better than D700 at ISO 25600+ and similar results below. which the D7000 does not.

We will have answer to that once IR images for K-5 are posted. Until then, it's just unproven claims.

 

[Oleg_V]

The only odd part about K-5 score is DR score at ISO 80.

I personally think that those fascinating DR at base ISO are more or less the tricks. The most suspicious thing about it is the slight kick up of the curve which first appeared in D5000. If you read carefully the dxomark description of the DR rating http://dxomark.com/...-more/Understanding-DxOMark-Database/Measurements/Noise you can observe that DR may be easily tweaked just by adding some constant offset to the signal. If the absolute dark signal exceeds the noise dispersion the DR obtained by dxomark method just goes to infinity. I guess sony just tweaks offset to be close enough to the noise dispersion ..

 

[Oleg_V]

The only odd part about K-5 score is DR score at ISO 80.

I guess sony just tweaks offset to be close enough to the noise dispersion ..

The following just illustrates the point



Loosing contrast with increasing exposure correction is obvious artificial trick.

 

[AngryCorgi]

Well, the DXO results are in...The D7000 is practically identical to the K-5. Any "magic" that was presumed to be in the K-5's variant of the sensor hasn't shown up. So all the "Pentax egineers have worked a miracle" can mostly cease. The K-5 goes down to ISO80 (actually measured as closer to ISO64 than ISO80), giving it that huge extra 0.2EV of DR at base ISO. Kudos, I guess, to Pentax for dropping the base sensitivity a hair to grab those extra bits. LOL. I guess we can pass the "mircale-working engineer" claims to the proper location...Sony. They made the sensor, and as it performs almost identically between the two cameras, it would be foolish to assume the praise goes anywhere else.

Above ISO1600, the NR kicks in and assists in helping the camera get a slightly higher rating above ISO1600. Of course, as you can see by the samples above, there is some signal loss in the NR (edges of letters degrade some) and it looks like any other software-based-NR, so its nothing that can't be duplicated in post.

Both are great cameras. Sorry the pen1s-measuring didn't end as hoped for the either side of crazy-claim posters.
--
-AC

 

[spbStan]

Considering now effective some post processing NR programs are in allowing selective NR and fine tuning based on image characteristics, Pentax's heavy permanent NR is a trick rather than a positive feature. Obviously they do not care about IQ, just test results in single criteria tests.

Now that it is settled, will the obnoxious Pentax fanboys go home. You have been shown to have baseless wild claims of superiority. You are still not addressing the poor software development that has plagued the brand for years, or poor AF and metering.

Just go away, we know too much to be swayed by the Tea-baggers of of the camera world, know nothings just yelling and making things up.
--
Stan
St Petersburg Russia

 

[teodorian2]

Well, the DXO results are in...The D7000 is practically identical to the K-5. Any "magic" that was presumed to be in the K-5's variant of the sensor hasn't shown up. So all the "Pentax egineers have worked a miracle" can mostly cease. The K-5 goes down to ISO80 (actually measured as closer to ISO64 than ISO80), giving it that huge extra 0.2EV of DR at base ISO. Kudos, I guess, to Pentax for dropping the base sensitivity a hair to grab those extra bits. LOL. I guess we can pass the "mircale-working engineer" claims to the proper location...Sony. They made the sensor, and as it performs almost identically between the two cameras, it would be foolish to assume the praise goes anywhere else.

Above ISO1600, the NR kicks in and assists in helping the camera get a slightly higher rating above ISO1600. Of course, as you can see by the samples above, there is some signal loss in the NR (edges of letters degrade some) and it looks like any other software-based-NR, so its nothing that can't be duplicated in post.

Both are great cameras. Sorry the pen1s-measuring didn't end as hoped for the either side of crazy-claim posters.
--
-AC

And as I repeat to death:

We must compare the end results where lenses, exposure, raw conversion, NR and sharpening as well as printer and paper determine the perceived image quality.

Will it be possible to see a significant difference between images from the same scene shot by the D7000, D700, K-5 or A580? I highly doubt it.

 

[awaldram]

People with investment in Pentax glass will buy Pentax.
People with investment in Nikon glass will buy Nikon.

People with little photographic skill will rant incessantly trying to have pxssing contest over figures they don't understand.

When it comes down to it the quality of the glass will have a greater affect on the finished image than the difference in any modern body.

Considering now effective some post processing NR programs are in allowing selective NR and fine tuning based on image characteristics, Pentax's heavy permanent NR is a trick rather than a positive feature. Obviously they do not care about IQ, just test results in single criteria tests.

Now that it is settled, will the obnoxious Pentax fanboys go home. You have been shown to have baseless wild claims of superiority. You are still not addressing the poor software development that has plagued the brand for years, or poor AF and metering.

Just go away, we know too much to be swayed by the Tea-baggers of of the camera world, know nothings just yelling and making things up.
--
Stan
St Petersburg Russia

--
My PPG

http://www.pentaxphotogallery.com/home#section=ARTIST&subSection=1471087&subSubSection=0&language=EN
My Photo Stream
http://www.flickr.com/photos/awaldram/

 

[rwl408]

If anyone is "really" tempted by k-5, wait until this time next year because it was well-overpriced. Just look at the price of K-7, Pentax's top of the line last year, which was priced at US$ 1300 at introduction but you can get one at below US$ 800 now. The truth is this has always been the price story of Pentax's top of the line forever.
--
Rick

 

[steelski]

Considering now effective some post processing NR programs are in allowing selective NR and fine tuning based on image characteristics, Pentax's heavy permanent NR is a trick rather than a positive feature. Obviously they do not care about IQ, just test results in single criteria tests.

If you examin the D7000 results you will see that there is evidence of NR at ISO 6400+

What I have seen is that it leaves a heavy coarse grain which is a real pain to get rid of.

PP NR is almost certainly not at the point of hardware NR. But you would not know that. CMOS sensors have noise canceling circuitry that allows such sensitivity. Is that a trick? Canon white paper on CMOS anyone. Sony has had programable NR on sensor for RAW since the D300 came out, which certainly has NR. as does the D90 and D5000. Its at a lot lower level than what Pentax does. but it is very destructive to colors, but leaves a lot of detail. The A700 was gutted by DPreview for too much NR. in contrast the Pentax K-x using the same sensor (or there abouts) managed to get passed their BS detector and come to the conclusion that it provides the highest IQ of any APS-c camera to date.
Personally I think the K-x was heavy on NR.

The K-5 is a lot lighter, and gets real results, its usable at 25600 where the D7000 struggles with some banding and color casts, and having processed 51200 files which acording to DXOmark are very poor, to get a usable result just blows the mind. You can pull whatever you want, call us whatever, but you will see that the K-5, with its NR and tricks pulls better results at 25600 and 51200 than the D700 and D7000.

Here is my disclaimer, I believe that the D7000 has been getting a rough deal against the D700 because of the Imaging resources samples being a 1/3 stop underexposed. I predicted long ago from looking at the A55 samples from IR that the sensors are within 1/2 a stop at the highest ISO and slightly more below that.

People put them at around a stop apart, I put them closer, possibly 1/2 a stop. I also see with the Pentax NR that the K-5 gets a further 1/2 stop more on the D7000 at higher ISO. And the last but not least, the Pentax files have a different grain, and it is said that the NR Pentax performs has atributed to the demosaic algorythmes which give the possibility of a cleaner looking rich RAW capture. Which does not show in RAW data.

Whatever you believe, I will wager that when IR puts the K-5 results up, we will see what all the fuss is about, and wether the DXOmark numbers matter anything. I would say that the DXO numbers are almost always very close to the final truth. but pure numbers are not what matter, I am a K20D owner, I see that acording to DXO my camera is aprox 2 stops behind the K-5. But in reality its actually closer to 3. why? because the K-5 51200 file is about as maluble at the 6400 file from the K20D. I have samples that come close to the 2 stops, but they are flat and have no latitude. If I underexpose by 1/2 stop the file is unusable! The K-5 files I have worked with are just plain better even 3 stops further.
Its not simply about numbers and "facts"

There are other areas which make the cameras different apart from the sensor.

Now that it is settled, will the obnoxious Pentax fanboys go home. You have been shown to have baseless wild claims of superiority.

They are not baseless. DXO is half the story, Real samples are the other. If DXO manage to get the exposure right for the K-5 then we could be in for a treat.

 

[Howzit]

The comment about the D7000 is exactly how I felt on picking it up the first time. So true. It's a dumbed-down D300 in functionality, a portly D90 in stature. I was surprised by how graceless it felt in the hand. I'm usually totally gung-ho for Nikon bodies of all shades, too. (The D50, D5000, and now the D7000 are the only Nikons in recent memory I've disliked for hold.)

The K-5 design is pretty close to perfection in comparison, though a fraction cramped around the grip for my hands personally.

I could not get comfortable with the D90 or D7000 but the K-5 needs to sell for $1200 if Pentax ergo's are to win over. Until then the D7000 is a superb camera at an awesome price.

 

[steelski]

The comment about the D7000 is exactly how I felt on picking it up the first time. So true. It's a dumbed-down D300 in functionality, a portly D90 in stature. I was surprised by how graceless it felt in the hand. I'm usually totally gung-ho for Nikon bodies of all shades, too. (The D50, D5000, and now the D7000 are the only Nikons in recent memory I've disliked for hold.)

The K-5 design is pretty close to perfection in comparison, though a fraction cramped around the grip for my hands personally.

I could not get comfortable with the D90 or D7000 but the K-5 needs to sell for $1200 if Pentax ergo's are to win over. Until then the D7000 is a superb camera at an awesome price.

Absolutley true.

What is very strange is that everywhere else the two cameras sell for around a $50 difference. which is very strange. The US price will drop as soon as demand drops. for the moment it is selling out like the d7000, so they see no need to drop. but they are missing on people like you. which does not help in the long run.

 

[kayone]

Does the K-5 have one of the most impressive high ISO APS-C sized sensors on the market? Maybe they do, and you know what, that's fine, Pentax owners and fans can have bragging rights because frankly, it's irrelevent to me since all my Nikon accessories and lenses are useless with a Pentax body. I simply don't care, it doesn't make the D7000 image sensor any less impressive.

 

[Lifer]

Know what, the high DXO rating given the K5 and D7000 won't help most of these people in this useless contest of "my camera is better than your camera" take better pictures.

--
Richard R. Price

 

[Voikkari]

Hi!

You are absolytely right. I have Pentax lenses and I'm going to use them as you use your gear Nikkors!

Take care.

--
K20 / K10D/*istD/MZ5n/Z70/SF7/MESuper/MX/Spotmatic F
Pentax since Spotmatic F

 

[Richard Smals]

If anyone is "really" tempted by k-5, wait until this time next year because it was well-overpriced. Just look at the price of K-7, Pentax's top of the line last year, which was priced at US$ 1300 at introduction but you can get one at below US$ 800 now. The truth is this has always been the price story of Pentax's top of the line forever.
--
Rick

Strange reasoning..the Nikon D90 lost even more here in NL: 580 Euro now. Started with over 1000 Euro is approx. 420 euro difference.

Nikon D300S on start € 1800 and now € 1130, resulting in 670 euro less.

The K-7 is € 1100 start and now € 790 = 310 euro difference

Don't fool yourself: a camera is not a good investment, the prices will always fall. And with Nikon just as hard.
--
Bye4now



http://www.indots.nl

I have the deepest respect for all those people who like me.

 

[neil holmes]

, it's irrelevent to me since all my Nikon accessories and lenses are useless with a Pentax body. I simply don't care, it doesn't make the D7000 image sensor any less impressive.

Thats not quite true.....

Nikon lenses work very well on Pentax bodies......I ONLY use my old Nikon 85 1.8 on my Pentax K-x stabilized....never on my Nikon D50.

I would not even try to use G lenses (though did for a stabilization test) and anyone dumb enough to do this does it at their own risk but for a bit of fun or for lenses either not available and/or cheap enough to try, I will use them from time to time (and that 85 1.8 whenever I need/want to).

Nikon lenses mount the reverse way on Pentax and most do not "lock" but hey they work, are secure enough for me.

The K-5 and D7000 both look like very nice cameras.

neil
http://www.flickr.com/photos/26884588@N00/