DPR's tests need the same lens brand

[Martin Frost]

"Switching to a standard RAW converter (in this case our benchmark;
Adobe Camera RAW) means that the image processing pipeline is
equalized between the cameras."

Yes, but what about lenses ?? It's not only the processing pipeline
that needs to be equalized, the lenses too.

The result would be the sharpness of the sensor and AA-filter.

I think this is what dpreview measures, but what is only one of 4 parts in real world.

Real world sharpness =°=
Sensor and AA-Filter sharpness

• Typical-System-Lens-Sharpness-at-F-5 ( not F-10 )
• Image-Processing
• perhaps Anti-Shake-Performance but that last thing may be too difficult to measure

In systematic view, dpreview should either test all 4 components and then compute the result, or test it together as real-world-example

Even the color will not be rendered the same, each lens has a slight
yet existing color cast and what about the peripheral sharpness ??

Being an Oly user, I just want to throw in ( offtopic ) that the Oly lens color is equalised because 4/3-lenses know their own color cast, tell it to the camera, and the camera corrects it. Funny. But that was just a remark.

. One lens for the
4/3 format, one for the 1.6/1.5 one for the 1.3 and one for the FF

in the above systematic meaning, IF dpreview creates a formula to add the 4 single sharpness results to one result, and the sharpness-of.the-sensor-plus-AA-filter should be measured alone,

then this is correct, but it should be the best lens then, which can change over time:

. One lens for the
4/3 format, one for the 1.6/1.5 one for the 1.3 and one for the FF

or dpreview should state their results as "sharpness on system prime lenses" test

cheers
Martin F.

E330andE500

Typing errors are intended to provide a basis for global amusement.

 

[Fabio Amodeo]

Not all lenses work the same with different brands of camera. Not all lenses work the same with different cameras with different sensor from the same manufacturer. Not all RAW converters work the same with different RAW files (it depends on the level of information the software firm had from the camera manufacturer). There are simply too many variables for any kind of scientific data.

Dpreview gives us an idea at large of what result we can expect for a camera. I think that's enough for me. For any further information, I would plan to put my hands on the camera, and test it in my environment. No other way seems possible to me.

Fabio

 

[Ludovic Monchat]

I would be 100% OK with your statement below

Dpreview gives us an idea at large of what result we can expect for a
camera. I think that's enough for me.

My only problem is that a lot of people actually takes DPR's comments on sharpness for granted, hence the wierd battle between the Nikon trolls and the Canon's regarding the D700 definition...

Using a single lens would solve a lot of speculations

--
Ludo from Paris
Tankers of tools, thimbles of talent
BestOf http://ludo.smugmug.com/gallery/1158249

 

[Erik Magnuson]

Using a single lens would solve a lot of speculations

No, it would just change them a bit. Those who want to believe that one camera "blows away" another when pixel peeping is involved will still find many reasons why the test is/isn't valid depending on whether it does/doesn't match their preconceptions.

--
Erik

 

[Chris59]

The resolution, contrast, dynamic range etc. of a camera is dependent on the sensor, lens and processing. In the real world, people don't use Canon lenses on Nikons or vice versa and the processing done is usually by the RAW converter supplied with the camera.

I do take your point that it is a little hypocritical for DP Review to on the one hand extol the virtues of using the same RAW converter to create a level playing field (more about this anon) and then use different lenses when rating a camera's ability to record a quality image. However, using the same lens and RAW converter is in my opinion the wrong way to go!

I believe it is more important to keep the proprieary lenses and use the sharpest available at the sharpest aperture, but to ALSO use the RAW converter that comes packaged with the camera not only because in the real world that is how most photographers will convert their images but also because it is most likely going to be the best available for that particular camera.

There is absolutely NO guarantee that Abobe, or whoever, will be able to do an equally good job of converting the different maufacturers' RAW files into TIFFs or JPEGs. The reason is that they are always back engineering their converters and are not privy to all the subtleties that exist in the RAW files.

In short, keep the lenses that are being used now and ditch ACR for RAW conversion in favour of the RAW converter that comes with the camera. I think that with this strategy you will find that there will in fact be greater differences between JPEG (which is the camera manufacturer's JPEG engine) and RAW than is now the case.

I mean who can say definitively that Olympus has an excellent JPEG engine which is not improved upon by shooting RAW because of the quality of their JPEG engine or the LACK of quality in the ACR conversion. In my opinion testing cameras means testing camera, lens and RAW converter NOT in some utopian level playing field to determine who has the best sensor, but in real world conditions using the lenses and RAW converter 95% of users will be using when they purchase the camera.

At the very least, to take some of the heat out of this debate, DP Review should test BOTH proprieary and "best possible" lens as well as ACR and proprietary RAW converter and let us make up our own mind which path to take. I realise that this creates more work, but once they have chosen the lens to be used, it shouldn't take up much more time - and if it did, so what, I'm sure most people here would prefer to wait one or two more days for a review which included results with "best possible" lens and converter.

 

[Nathan Yan]

I believe it is more important to keep the proprieary lenses and use
the sharpest available at the sharpest aperture, but to ALSO use the
RAW converter that comes packaged with the camera not only because in
the real world that is how most photographers will convert their
images but also because it is most likely going to be the best
available for that particular camera.

There is absolutely NO guarantee that Abobe, or whoever, will be able
to do an equally good job of converting the different maufacturers'
RAW files into TIFFs or JPEGs. The reason is that they are always
back engineering their converters and are not privy to all the
subtleties that exist in the RAW files.

This is very true. I notice a pretty significant difference between ACR and DPP, especially on high-ISO images.

--
Camera blog: http://shuttersounds.thedailynathan.com/
Portfolio: http://portfolio.thedailynathan.com/
Class: http://www.photodecal.org