Pixel shift K-1 sample on IR

maceoQ wrote:

Kendall Helmstetter Gelner wrote:

I also find it REALLY funny that people would like the K1 results which plainly do not have nearly as much texture going on. The back wall, the cardboard in the crayon box, all over you can see that Pixel Shift is not enough to bring out fine texture the way the Quattro can - which in theory was not as good as the Merrill! How many people complaining about the loss of microcontrast from Merrill to Quattro are now saying they really like the K1...

In my opinion, there is such a big difference, that i'm not even willing to add crops to make my point.
The "super-high" and sharpened Quattro version is just a noisy, over sharpened, aliased mess that hurts my eyes,

You mean, a mess like this? (standard size DP2 Quattro out of SPP, resized to match K1 size and processed in LR)

100% crop

This really doesn't hurt my eyes. But I don't use SPP Standard settings, which can mess things up a bit.

while the K1 is super clean.
If you now take this super clean DNG and add tonal contrast and sharpening to your taste, then you get enough texture, enough resolution and still a far cleaner result.

I had a go and got this:

100% crop, no resizing

Both images have patch 22 adjusted to neutral grey and near 50%.

I won't compare noise because a multi-shot composite should have much lower noise even than a single-shot image from the same camera. Wait for the new HDR multishot Quattro files and compare noise then.

The interesting bit for me is detail. The Pentax should easily out-detail the Quattro because after pixel shifting it has 36M green values whereas the Quattro has only 19.5m luminance values. Yet I think the Quattro has done very well and is only slightly behind -- especially considering that the Quattro image has been through a resize process. Looking back, I may have over sharpened the Pentax a bit (label of bottle on right looks halo-enhanced to me), because I was looking to see its detail abilities.

Since the release of Kalpanika we all know how much contrast-processing SPP "adds" to the X3F-files, so it's fair enough to push those K1 images also a bit.

I tried, above.

But anyway: As much as i like the pixel shifted result, since it needs 4 shots it is completely uninteresting for me.