DPR is limiting availability of SD1 sample images

An interesting debate that looks at this DPR decision from a number of angles.  One of then is a question:

Is pixel level quality significant?

From reading replies in this thread, it is clear that there are people who think it is. They say they like to peruse images section by section examining the fine detail forensically close.

Clearly for those who do this, pixel level quality at the actual pixels view is paramount. In fact, to me, it sounds like it is the only thing that matters,  In this case the DPR actual pixels tool is more than a convenient tool, it is a legitimate way of enjoying images.

Foveon cameras along with monochrome cameras like the Leica MM win hands down for this kind of viewing. Bayer cameras can compete if downsized significantly but will still be a little behind.

​I'm definitely not in that camp.

I'm interested in looking at good pictures, not good pixels.  For me that value of a photograph is in the whole picture:  coherent images where subject, composition and lighting combine together to produce a satisfying and exciting picture. Call me old fashioned.

For this non-pixel peeping approach, the amount of detail captured by the sensor simply determines the maximum size at which you can reproduce the image satisfactorily. That's it really.

The DPReview tool is of little use for this as it only lets you compare actual pixels. For my uses, comparisons have to be done by comparing equal sized reproductions (in my case, prints).  I can do this myself either by printing the images to a fixed size and scrutinizing them or by resampling files to the size I want to print them at and pixel peeping the resultant files.

This tells you how much detail the files contain and how the image is affected by artefacts at the reproduction size I have chosen. The DPR tool is useless for this.

What would be useful would be a tool that compares prints rather than files at a chosen size. Imaging  Resource do a print size assessment as a key part of their reviews but they don't make the source images available. A step forward beyond what they do would be to make prints at different sizes, scan them on a high quality scanner and put the scans into a comparator tool (or sections of them).  Second best to that would be to resize the files to a given size and let the tool compare these files.  Either of these methods would be better than the current DPR tool because that tool compares only pixel quality, it cannot satisfactorily cope with files of different pixel counts.

--
Galleries and website: http://www.whisperingcat.co.uk/
Flickr: http://www.flickr.com/photos/davidmillier/