I feel bad for saying this... but who else is waiting for the k-7's DXO mark?

IsaacEastgate · 05-29-2009, 04:58 AM

Now im not one of those pixel-peeping types, but one thing really annoyed me when i got my first DSLR half a year ago, the k20d. Coming from both film and digital compact, i knew the vast difference in dynamic range they offer, and to see a semi-pro DSLR only placing itself in the middle was a surprise to me. After looking around the internet it is common opinion and a repeatable result that the k20d has 0.5-1.5 stops LESS (effective) dymamic range than the k200d, k-m, or k10d. Thats the difference between a light sky and a completely blown out one.
It appears that in the rush to create pentax's 14.6 mega-sensor they just didnt get some of the fundamentals right. Or perhaps they sacrificed that dynamic range for some excellent detail rendering (which im ok with really).
So for now, the only thing that could make me want to buy a k-7 is if pentax really have rebuilt that sensor from the ground up, and if the noise and D-range (which have so far been kept under the covers) really measure up to the competition.
I'm not expecting full frame results, of course, these cameras come into a league of their own, but if the k-7 can get close to the d90's frankly spectacular sensor (sorry guys), that'll be all i need

P.S.

you can find DXOmark here: DxOMark Sensor
I know about the huge inaccuracies and irrelevance in some cases, dont worry. I realise that the GX-20 got an all round higher mark than the k20d for some reason, and have taken this into account.
Please dont tell me about HDR or D-range, i dont carry a tripod with me and shoot raw most of the time
I'm not k-7 "bashing". I think its an awesome camera and i dont know how they made it so small either. I just think there is no information on the new sensor and it ought to be good to justify the price tag

xxx

Ishpuini · 05-29-2009, 05:53 AM

I have always been at a loss concerning these claims. My K20D actually has the largest DR of all my DSLRs to date. Better than the *istDS and *much* better than the K10D... Strange that these tests show different...

But... (I know you told us not to mention this, but I am going to anyway)... I do have D-Range activated... And I exclusively shoot in RAW as well. I have the impression D-Range also does something in RAW, of course you don't see as many blown highlights in the review but that could still be related to the processing of the embedded JPG, but I noticed I could bring back more detail then before when I didn't have D-Range activated.

I always refer to this example:
First a shot I accidentally overexposed dramatically but for some reason never deleted (which I normally do). I had been shooting a pano in M mode and had adjusted aperture prior to this exposure without realising that Shutter speed wasn't being adjusted... No adjustments when developing the RAW.

I pulled back 3 stops in Lightroom with this result:

Whatever tests may show, the K20D's sensor's DR is OK. I could never do that with either of my 2 previous Pentax DSLRs using Sony CCD sensors... Hence I'm OK with the K-7 using a sensor that evolved from the one in the K20D!!

Wim

**Buddha Jones** · 05-29-2009, 06:02 AM

The only thing I am waiting on is for my bank account to replennish itself.

jfsavage · 05-29-2009, 06:20 AM

I'm not a pixel peeper, but I'd love a detailed review of the K-7.

kevinschoenmakers · 05-29-2009, 06:48 AM

Ishpuini, I would say the overexposed shot has a charm of it's own

Ishpuini · 05-29-2009, 07:03 AM

Originally Posted by kevinschoenmakers

Ishpuini, I would say the overexposed shot has a charm of it's own

I know I pulled back the exposure too much, but I did so to show to what degree details can be brought back.

Wim

smdeep · 05-29-2009, 07:36 AM

Originally Posted by kevinschoenmakers

Ishpuini, I would say the overexposed shot has a charm of it's own

Hi

I was just about to say the same thing! The overexposed shot looks wonderful to me! Would love to have a print of that one!

Sudeep

jeffkrol · 05-29-2009, 08:16 AM

Originally Posted by IsaacEastgate

you can start here...
Analysis: K-7 DNG raw Beta 0.2...: Pentax SLR Talk Forum: Digital Photography Review

*Wheatfield* · 05-29-2009, 09:17 AM

DXO and it's ilk are just another meaningless way for people to waste time. Any DSLR made in the last 5 or so years is capable of rendering a perfectly acceptable picture, so why be concerned about useless measurements?

Maffer · 05-29-2009, 11:41 AM

Originally Posted by Ishpuini

I have always been at a loss concerning these claims. My K20D actually has the largest DR of all my DSLRs to date. Better than the *istDS and *much* better than the K10D... Strange that these tests show different...

But... (I know you told us not to mention this, but I am going to anyway)... I do have D-Range activated... And I exclusively shoot in RAW as well. I have the impression D-Range also does something in RAW, of course you don't see as many blown highlights in the review but that could still be related to the processing of the embedded JPG, but I noticed I could bring back more detail then before when I didn't have D-Range activated.

I always refer to this example:
First a shot I accidentally overexposed dramatically but for some reason never deleted (which I normally do). I had been shooting a pano in M mode and had adjusted aperture prior to this exposure without realising that Shutter speed wasn't being adjusted... No adjustments when developing the RAW.

I pulled back 3 stops in Lightroom with this result:

Whatever tests may show, the K20D's sensor's DR is OK. I could never do that with either of my 2 previous Pentax DSLRs using Sony CCD sensors... Hence I'm OK with the K-7 using a sensor that evolved from the one in the K20D!!

Wim

Whoa...Lightroom seems to be lightyears ahead pentax photolab. I just cannot get that kind of results when I try to pull back overexposed details of an image taken with K20D. I guess I should really try out the lightroom next. Thanks for the example.

and · 05-29-2009, 11:47 AM

I find dxomark an interesting concept but in practice it doesnt seem to be very usefull at all. And I expect the k7 to perform very similarly to the k20d, I believe the main differences in the sensor is the higher readout / better live view / CAF / Video and not the still quality. although I dont expect it to be any worse, possibly a bit better but no huge change.

ryan s · 05-29-2009, 11:49 AM

Wim, great job! I would have considered the recovery of the sky hopeless

But...there it is. Man, I love how Pentaxes render the blues and all colors for that matter.

Originally Posted by Wheatfield

DXO and it's ilk are just another meaningless way for people to waste time. Any DSLR made in the last 5 or so years is capable of rendering a perfectly acceptable picture, so why be concerned about useless measurements?

Thank you.

I was going to ask "What's the point of DXO?" even though I know the "reason" it was created

johnmflores · 05-29-2009, 12:10 PM

Nope. It's bound to be better than the K20d and that's already awesome. I'd like but don't need even more awesome.

IsaacEastgate · 05-29-2009, 01:00 PM

Originally Posted by Ishpuini

First a shot I accidentally overexposed dramatically but for some reason never deleted (which I normally do). I had been shooting a pano in M mode and had adjusted aperture prior to this exposure without realising that Shutter speed wasn't being adjusted... No adjustments when developing the RAW.

I pulled back 3 stops in Lightroom with this result:

Thats pretty amazing. I will try D-range alongside raw from now on, i have never got results like those.
I think lightroom probably has some clever algorithms for retrieving highlights too.

this link from jeff: Analysis: K-7 DNG raw Beta 0.2...: Pentax SLR Talk Forum: Digital Photography Review
Shows some good and bad points. disappointing that reds will clip more (especially for band photography where i already have trouble) and that the overall noise levels are similar, but it sounds like the quality and appearance of noise has improved a lot and the quality of the sensor is just higher.

Hmmm.

I guess we just wait and see./

mk07138 · 05-29-2009, 01:28 PM

Originally Posted by Ishpuini

I have always been at a loss concerning these claims. My K20D actually has the largest DR of all my DSLRs to date. Better than the *istDS and *much* better than the K10D... Strange that these tests show different...

But... (I know you told us not to mention this, but I am going to anyway)... I do have D-Range activated... And I exclusively shoot in RAW as well. I have the impression D-Range also does something in RAW, of course you don't see as many blown highlights in the review but that could still be related to the processing of the embedded JPG, but I noticed I could bring back more detail then before when I didn't have D-Range activated.

I always refer to this example:
First a shot I accidentally overexposed dramatically but for some reason never deleted (which I normally do). I had been shooting a pano in M mode and had adjusted aperture prior to this exposure without realising that Shutter speed wasn't being adjusted... No adjustments when developing the RAW.

I pulled back 3 stops in Lightroom with this result:

Whatever tests may show, the K20D's sensor's DR is OK. I could never do that with either of my 2 previous Pentax DSLRs using Sony CCD sensors... Hence I'm OK with the K-7 using a sensor that evolved from the one in the K20D!!

Wim

Wow. . . . . .

That's all I can say. With my K20D and Lightroom 2 I have never succeeded in pulling something back like that. . . . . . I apparently need more practice in using Lightroom or I definitely need to use the D-Range setting.