[cooldude14es]

Hello all.

Just 2 days ago the Pentax Spanish Forum made an interesting review of the new K-7 and I wanted to share the results with you. Hopefully, you dont need to understand spanish because images speak by itself.

The idea was taking the same photo using a Pentax K-7, K20D, K10D, Nikon D300 and Canon 50D. As you will see there are different elements on the set so one can check dynamic range, color, IQ and noise at different ISO values.

All pictures were shot in RAW mode, auto WB, and noise reduction off. There is no post processing on any of them, just setting WB to Tungsten using Lightroom software, so NO noise reduction, NO sharpening, just export to jpeg.

The same lens was used on all cameras (different mount, of course) and it was a Sigma 105mm f2.8 Macro EX DG.

Here is the link:

Analisis K-7 vs Nikon D300 vs Canon 50d vs K20D vs K10D


* First test images are shots at ISO 100, no cropping.

* Then you have some 100% crops from 2 different parts of the photo at:

- 1/10, f:11, ISO 100 (except for D300 and 50D that due to a mistake on the test crops from these cameras are at 1/30, f 6.3)
- 1/20, f:11, ISO 200
- 1/40, f:11, ISO 400
- 1/80, f:11, ISO 800
- 1/160, f:11, ISO 1600
- 1/320, f:11, ISO 3200
- 1/640, f:11, ISO 6400

* Next post, there is a direct comparison K-7 vs K10D cropped and re-scaling the K-7 image to match the 10Mpx resolution of the K10D:

Analisis K-7 vs Nikon D300 vs Canon 50d vs K20D vs K10D



* Next post, same picture but without WB adjustment, just direct from cameras.
Also, you can download the RAW files:

Analisis K-7 vs Nikon D300 vs Canon 50d vs K20D vs K10D


Regards.

[PolishMike]

These are great, thanks for posting them here!

ISO 3200/6400... ouch, Pentax. No amount of "Pentax preserves texture, you can clean it up in post" talk can really save Vader's nose and eyes in the K7 ISO6400 shot.

Also, does the Pentax images being darker for a given shutter/aperture/ISO combo mean that Pentax are labelling their sensitivity ratings even further from "true ISO" than the competition?

[snogglethorpe]

ISO 3200/6400... ouch, Pentax. No amount of "Pentax preserves texture, you can clean it up in post" talk can really save Vader's nose and eyes in the K7 ISO6400 shot.

Indeed -- but the nikon and canon results for those settings seem just as bad!

[pingflood]

These are great, thanks for posting them here!

ISO 3200/6400... ouch, Pentax. No amount of "Pentax preserves texture, you can clean it up in post" talk can really save Vader's nose and eyes in the K7 ISO6400 shot.

Also, does the Pentax images being darker for a given shutter/aperture/ISO combo mean that Pentax are labelling their sensitivity ratings even further from "true ISO" than the competition?

In the case of the 50D, ISO 100 is actually more like ISO 125. IS0 200-1600 or so is more or less dead on. So the Pentax is probably just about right, Canon for some reason sandbags a bit.

[8540tomg]

I liked the looked of the the results for all the cameras. I was pretty pleased how well the K10 stood up in comparison. We always want the latest and greatest but based on these results, for the type of shooting I do, I think I can wait another generation before I upgrade.

Tom G

[pingflood]

Yeah, there aren't any quantum leaps these days, just incremental improvements. Goes to show Ben's point from another thread that most all cameras today are "good enough" for most people.

[konraDarnok]

Any test I've seen like this, the K10D always holds it's own and more at ISO 100.

[PolishMike]

In the case of the 50D, ISO 100 is actually more like ISO 125. IS0 200-1600 or so is more or less dead on. So the Pentax is probably just about right, Canon for some reason sandbags a bit.

Oh, I always assumed they over-estimated ISO ratings to make sensors look better at "higher" ISO. Looking at the Nikon D90 values, which over-estimate the ISO at higher sensitivities, I assumed it would be the same across brands. Shows what I know.


And, snagglethorpe, the Nicanon shots do look noticeably better, with less overall noise - in the Pentax ISO6400 shot the chroma noise has obliterated a lot of detail, which I can still make out in both the others. You could argue that the Pentax shot is underexposed, so noise is more visible, but since the shots are taken with the same exact settings, this means that for a given low-light situation the K7 is going to do worse (either higher noise or longer shutter speed).

The (kind-of) good news for me is that the difference is only becoming visible at ISO1600, and I'm fine staying within those bounds. Would be nice to be able to take quality pictures at high ISO, but I think a larger sensor is the only real way to really achieve that.

[mgoblue]

The k7 looks astonishingly sharp and 3D up to ISO 1600, especially compared to the 50D which looks badly smudged. Up to 800 I'd surely take the K7 (the D300 does preserve more color detail with its lower saturation though); beyond 1600 the K7 looks worse with lots of chunky chroma noise and less detail than the Canikons. I'm not sure, however, that I'd be willing to use the Canikon outputs either!

Can anyone explain why the 50D gets sharper and less smudgy at higher ISO than it is at low ISO? That really surprises me.

[pingflood]

The k7 looks astonishingly sharp and 3D up to ISO 1600, especially compared to the 50D which looks badly smudged. Up to 800 I'd surely take the K7 (the D300 does preserve more color detail with its lower saturation though); beyond 1600 the K7 looks worse with lots of chunky chroma noise and less detail than the Canikons. I'm not sure, however, that I'd be willing to use the Canikon outputs either!

Can anyone explain why the 50D gets sharper and less smudgy at higher ISO than it is at low ISO? That really surprises me.

Judging by the shutter speeds it could be vibration induced. My 50D certainly doesn't get sharper at higher ISO.

And yikes, at 6400 the K-7 just obliterates the detail in blotchy noise. Surprised at the difference there.

[snogglethorpe]

And, snagglethorpe, the Nicanon shots do look noticeably better, with less overall noise

The detail is hard to judge, because the k7 shot is underexposed in this area, but the amount of noise looks very similar to me.

It'd be nice if the test site had taken more care to make the exposure similar!

[PolishMike]

The detail is hard to judge, because the k7 shot is underexposed in this area, but the amount of noise looks very similar to me.

It'd be nice if the test site had taken more care to make the exposure similar!

Well, they *did* - same settings used for every shot. It's not their fault that camera manufacturers lie about ISO values. Of course if you adjusted the test so that more light was hitting the Pentax sensor, the signal-to-noise ratio would look much better - this would have nothing to do with the sensor itself.

If the Pentax shot at the *same settings* produces a darker and therefore noisier image then it is very much a flaw of the Pentax - it means that a photographer trying to take a shot in low light and being limited in terms of shutter speed would get a poorer shot.

[mgoblue]

On the other hand, the 50D is chronically overexposed -- by that logic making the sensor look better for noise -- but the photos look washed out and the highlights are burned badly. Less noise at high ISO, yes, but still not useable to me b/c instead of black, there is brown...and the images are bleached and burnt at all ISO (the screw on the paper and the shading on the clock face in the crop almost vanish). But, I tend to prefer highlight details to shadow details and others may have a different preference.

To me the K7 is exposing the best...The Nikon splits the dif...and notably it splits it for noise too. As was discussed in another thread, exposure seems to be making the big difference...not the sensor.

As an aside, the Nikon is also exhibiting nasty flare off the right edge of the clock...interesting since it is the same lens on all three.

[Edvinas]

What an excellent comparison! Finally someone got it right! Honestly, I hate when so called review sites make test shots of semi-pro cameras with all kinds of different lenses (even cheapest kit lenses) and then make final verdict about camera quality. Respect to Spanish guys who managed to use the same lens model with all cameras.

Very revealing samples, after all K-7 is not that bad And I wouldn't use any of those cameras above ISO1600. K10D limit is ISO800.

[Erik]

For any sane purpose, these will all look more or less identical as prints. There's no significant difference between any of these cameras, honestly.