DxO comparison 40D, 7D, Pentax K-5

The K5 looks like it may be very impressive at higher ISOs:
https://photography-on-the.net …?p=11184141&pos​tcount=196
https://photography-on-the.net …?p=11207635&pos​tcount=222

DxO mark presents sensor data via general ratings that are of very little use to photographers. Their data is kind of a raw sensor performance figure that is pixel specific. Their data does not account for sensor size, pixel count or in-camera RAW processing.

So really, their ratings are meaningless. If you peruse their overall scores you will find numerous examples of ratings that seem to make no sense. Is Nikon's D90 really the best camera ever made for IQ? Not really, and thus you know everything you need to about how useful these ratings are.

i'd take it if it was FF

stsva wrote in post #11236935
The K5 looks like it may be very impressive at higher ISOs:

JeffreyG wrote in post #11236940
DxO mark presents sensor data via general ratings that are of very little use to photographers. Their data is kind of a raw sensor performance figure that is pixel specific. Their data does not account for sensor size, pixel count or in-camera RAW processing.

So really, their ratings are meaningless. If you peruse their overall scores you will find numerous examples of ratings that seem to make no sense... .

Not sure if I would call them meaningless, but you're right, they should not be the only measure. However, if you take a look at the image in the thread, that is one impressive ISO 51200 shot. Maybe there has finally been a sensor breakthrough and we are about to see a new batch of high ISO low noise bodies.

gjl711 wrote in post #11237076
Not sure if I would call them meaningless, but you're right, they should not be the only measure. However, if you take a look at the image in the thread, that is one impressive ISO 51200 shot. Maybe there has finally been a sensor breakthrough and we are about to see a new batch of high ISO low noise bodies.

My post should not be taken to mean that the K-5 isn't good. I honestly have no idea.

But I have used a D90 (one of DxO's highest rated sensors) and it sucks hind tit against many lessor rated cameras that I have also used (like the 5D Mark II). So I do know exactly how useful DxO's sensor ratings are.

Which is to say, their ratings are very close to useless.

on paper the k-5 looks really good. i wouldn't mind giving one a test run.

Lemmie try and make a chart to explain the uselessness of DxO

Camera........Overall Score.........'Bits'..​.....'DR'.........'ISO​'
1D Mark IV........74..........​.....22.8.......12....​.....1320
5D Mark II.........79.........​......23.7......11.9..​......1815
5D...............71...​............22.9......​11.1........1368
D90...............73..​.............23.......​..12.5.........977

Hmmmm.....So according to DxO the original 5D and the 1D Mark IV have similar ISO performance? I own both and uh....no farking way.

And the 5D Mark II is not better at 'ISO' than the 1D Mark IV in my direct experience.

And just what goes into the 'overall' ratings? The D90 seems to suck for 'ISO' (I agree, it does compared to the other three choices in the real world too) and seems to be similar for 'bits'. So how do these overall ratings get created?


If you have used several cameras seriously (as I have) the more you paw through the DxO data the less sense it makes. I find endless examples of data from them that does not translate into real world experience. The above handful are just a couple examples of many.

Very good points, JeffreyG.

JeffreyG wrote in post #11237249
Lemmie try and make a chart to explain the uselessness of DxO

Camera........Overall Score.........'Bits'..​.....'DR'.........'ISO​'
1D Mark IV........74..........​.....22.8.......12....​.....1320
5D Mark II.........79.........​......23.7......11.9..​......1815
5D...............71...​............22.9......​11.1........1368
D90...............73..​.............23.......​..12.5.........977

Hmmmm.....So according to DxO the original 5D and the 1D Mark IV have similar ISO performance? I own both and uh....no farking way.

And the 5D Mark II is not better at 'ISO' than the 1D Mark IV in my direct experience.

And just what goes into the 'overall' ratings? The D90 seems to suck for 'ISO' (I agree, it does compared to the other three choices in the real world too) and seems to be similar for 'bits'. So how do these overall ratings get created?


If you have used several cameras seriously (as I have) the more you paw through the DxO data the less sense it makes. I find endless examples of data from them that does not translate into real world experience. The above handful are just a couple examples of many.

Exactly. I have the 1D4 and the 5D2 and the 1D4 smokes the 5D2 not only in ISO performance but in overall IQ. The only advantage the 5D2 has is the FF DOF, if it wasn't for that I would be using the 1D4 all the time. It's going to be interesting to see what they do with the 1Ds4, if it has the ISO performance of the 1D4 I may need to buy one.

K6AZ wrote in post #11237402
Exactly. I have the 1D4 and the 5D2 and the 1D4 smokes the 5D2 not only in ISO performance but in overall IQ. The only advantage the 5D2 has is the FF DOF, if it wasn't for that I would be using the 1D4 all the time. It's going to be interesting to see what they do with the 1Ds4, if it has the ISO performance of the 1D4 I may need to buy one.

Would you mind taking a couple of comparison shots? I'm really curious to see the differences.

How come on every picture posted for the K5 there's that funny strip off to the right side. Are these pictures non edited versions?

SunTsu wrote in post #11239049
Would you mind taking a couple of comparison shots? I'm really curious to see the differences.

For ISO or IQ?

The Pentax K-5 definately has some fantastic image quality and an amazingly clean sensor, with great dynamic range in the shadows. Definately is a cleaner sensor than the 7d. Great for Pentax users to get such a nice sensor! However, dxo-mark is absolute b.s. and has no bearing on reality

woos wrote in post #11240936
The Pentax K-5 definately has some fantastic image quality and an amazingly clean sensor, with great dynamic range in the shadows. Definately is a cleaner sensor than the 7d. Great for Pentax users to get such a nice sensor! However, dxo-mark is absolute b.s. and has no bearing on reality

What are you basing the K-5 to 7D comparison on?

Even if the K-5 is the second coming there is a huge issue with lens selection and pricing for Pentax lenses. Have a look at their primes.

K6AZ wrote in post #11240981
What are you basing the K-5 to 7D comparison on?

Even if the K-5 is the second coming there is a huge issue with lens selection and pricing for Pentax lenses. Have a look at their primes.

Well, I'm basing it on some posted 51.2k ISO examples, and some examples of some files from the K-5 being pushed 4+ stops, on low ISO files, and still looking 'okay'. More than my 7d can do. I love the 7d's resolution though, and I have no complaints as to it's dynamic range (I could use more highlight DR, which unfortunately the K-5 doesn't seem to help with at all--shadows aren't really a problem for me)--I mostly do landscape stuff, and the 7d is absolutely perfect for me--the only way I'll upgrade is for a significant megapixel boost (or maybe to a significantly higher MP full frame body). The iso 200-3200 files from the 7d look gorgeous...

It's great that pentax users have a competitive body available to them though, more competition is good for all of us. But yes, glass flat out matters more than the body does--and that will always be true--and that is why so many of us love Canon. But it's okay to think Nikon, Pentax, etc, stuff is good too. I actually like the Pentax ergonomics generally, not as well as Canon (can't stand the way the Nikon bodies operate, but I have coworkers who feel the opposite way from me and can't stand Canon, hey, to each their own). Hell, I have a coworker who is a great landscape photographer...his choice of gear? An ep-1 with a leica tri-elmar lens.