Nikon D4s: Bayer processing breakthru?

Started 3 months ago | Discussions thread
richard stone
Senior MemberPosts: 1,580
Like?
Re: Print analysis: matching Foveon to CFA
In reply to Tom Schum, 3 months ago

Tom Schum wrote:

Just for fun I printed the test image on 19"x13" paper (Inkpress Fine Art Matte, 12mil, 220GSM) on my Epson R2000 printer.

My wife immediately picked the image on the right as the better one because "the pink is pinker". This is the X-E1 image.

I noticed that the DP3M image does have slightly less saturation, but still excels in the depiction of surface textures in all the plastic (pots, crates, etc). Also, textures of the foliage and the red-painted metal rack were also more visible.

I did most of my acutance processing based on the printing visible in the second pot from the left on the top row, and the two of them match up quite well in the print. At the far right, on the top row, the tag printed "Ashley's annuals" has a slight sharpening halo in the X-E1 image and a more natural blurring in the DP3M image.

In the DP3M image, the edges of the plastic plant markers for the white pots in the top row on the left seem a little bit too sharp to me, almost stark, compared with these in the X-E1 image.

In the bottom right of the DP3M image, the plastic plant marker seems to have some edge jaggies compared to the X-E1 image. The barcode in the bottom right corner is more detailed in the DP3M image and almost jumps out at you,while in the X-E1 image it is somewhat muted.

The foliage in the DP3M image in the bottom left is much better rendered than in the X-E1 image. The leaf textures make it almost palpable. But the fern leaves in the top left seem to be pretty much the same in both images.

I think because of the differences in textural detail between the two images, that I prefer the X-E1 image. To me, it's just more pleasant to look at. The lack of textural detail makes it look smoother and nicer to me. The DP3M image has a grittier feel to it, more realism, you see all the warts.

-- hide signature --

Tom Schum

But your argument is exactly why I prefer the Foveon images.

My view is that the Foveon sensor gives you more detail, more texture, more realism, and when it does that it somewhat changes the nature (a pun?) of the image. A "sport" image from a Foveon is going to show more grit and dirt (fine detail) and show it differently than a Bayer image. I have always considered Bayer images to look more plastic and cartoon-like, less reality-based, than a Foveon image.

But is all that texture and reality what we (or most viewers) want in a sports image? Or in a wedding image, as usually imagined (another pun)? I think not. The Foveon sensor may in that sense change the nature of the image. For some people as well, landscape shots may not in fact benefit from excessive (meaning more...) detail. For them we have Thomas Kinkade.

We can take as an advantage the ability to print large as the issue, or be dismayed at how the Foveon sensor does portraits, but the Bayer sensors "work" by reducing detail and simplifying it. The most used Bayer sensors have two sets of filters, color and blurring, and then as a part of the Bayer concept they have to  "interpolate" (reconstruct) the results to get an image, so what can we expect? But in the end the images are pleasing to some, and effective for many uses, precisely because they have that "plastic" look.

I would propose that one reason people do not use Sigma's for sports images is that the detail produced is far beyond what is necessary, and the plastic look of sports images (and weddings, for that matter) are simply more accepted and enjoyed in that context, as compared to the more detailed, grittier, Foveon images.

Richard

-- hide signature --
Reply   Reply with quote   Complain
Post (hide subjects)Posted by
(unknown member)
(unknown member)
(unknown member)
(unknown member)
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark post MMy threads
Color scheme? Blue / Yellow