[normhead]

As a rule I haven't found any way of doing HDR that retains detail, it's often better to select different parts of a bracketed image and paste them together yourself if possible. It's a lot of work.

[yusuf]

As things stand, my Sigma DP2M with it's 30mm lens and Merrill sensor blows my K-5 with the 31 away. This will be my first real test of the K-3 when it arrives in a few weeks. It would be great if it was even close to parity as I now always carry the Sigma since it gives me such clarity at low ISO. It is limited to ISO 200. it would be nice to use the K-3 plus FA 31 and have the full range of ISOs available and know I am not sacrificing possible resolution with the combo. I will post my results the moment I get the camera since I won't even need to wait for the battery to charge.

would love to see any update.

[audiobomber]

There's a thread right now on DPR; K-5 vs. DP3. I quote a post below from DMiller below, because he's saying something that seems pretty evident to me; Sigma images are highly cooked. Meuh's comparison thread in DPR shows the K-5 with no sharpening, and no added contrast or saturation. As expected through an AA filter and no reasonable amount of processing, the K-5 images are bland and blurry. The other significant difference in K-5 vs. Sigma is that the Foveon colours are very warm (too warm, IMO), more like Sony/Pentax CCD than current Pentax sensors. If you prefer that yellow cast, it can be achieved on a K-5 with post-processing.

K-3 WB is warmer than K-5, colours straight out of the box are more accurate than recent Pentax bodies, and more accurate than Foveon, IMO. The K-3 has higher resolution than a K-5 and a no blur filter, so should be a much better comparison point. I don't expect it to quite match Merrill resolution, but in the interests of fairness, please post at least one set where a K-3 image is punched up to the level of the DP2M. Pentax "Reversal Film" image tone with extra sharpening would be a good start.

Here's the quote from DPR:

As a long time Foveon user (Not Merrill) and a K5 user I have some comments.

- When making visual comparisons, sharpening is crucial. The eye responds more to acutance than sheer resolution and it's very easy to vary one's opinion of files simply by varying the amount of sharpening.

- I'm assuming the Sigma files were processed in SPP. That program's 0 sharpening setting is actually heavy sharpening. Opinions vary, but no sharpening is between -4 and -8 in SPP.

- The 15MP Merrill sensor at low ISO has a combination of high resolution (24 - 30 MP equivalent depending on subject matter) and very high acutance (microcontrast). Combined with high default sharpening, that's a potent mix.

- The DP3 lens is superb and you are going to have to try very hard indeed to find a lens anywhere including Zeiss and Leica to match it. But any 50mm lens is going to be pretty damn good at f8, even 30 year old ones as long long as they're not defective.

- If appropriate USM and deconvolution sharpening were applied to the K5 shot it would look a lot better - but still not as good. 24MP cameras get closer.

- The colour differences ought to be adjustable by and large.

- The 36MP D800 models easily outresolve the DP Merrils (although you'd need very large prints to see it) but the Merrills still often exhibit higher microcontrast on certain types of subject matter.

- it's worth noting that the Merrill sensor is a bit limited in dynamic range (more Canon-like than Sony)

- it's worth noting that the Merrill sensor is noisy even at base ISO. Blue skies have texture and it's easy to introduce a canvas like pattern in smooth areas. ETTR exposure is critical.

- It's also worth noting that shadows are desaturated to hide the large magenta and green blotches the sensor is prone to owing to the massive matrix colour transformations that are needed to get from Foveon's weird colour space to RGB.

Even when you have low grain across the image it is sometimes odd how easily the blotches appear (curing this problem may be the driving factor behind the redesign of the new Quattro sensor.

- all those issues aside, on a good day it's possible to get spectacular results from the DP cameras.

Re: k5 v dp3 studio test.: Pentax SLR Talk Forum: Digital Photography Review

[DDoram]

The Sigma will likely have better pixel level IQ than K3. In prints it would be difficult to pick out but the Sigam Meriils are amazing at low ISO. I have both the K3 and Sigma Dp1m.

Dale

[Na Horuk]

I remember the K-5 being slightly behind the Sigma in some tests, but the K-3 has more MP and is no AA filter, so it should be able to hold its own. Of course, even if the Sigma produces more resolution, it has too many drawbacks for most photographers. Large files, short battery life, limited Post processing software, poor high ISO performance, limited lens selecton.. I respect the Sigma's Foveon technology, but it cannot be compared directly to bayer CMOS sensors. There are many things to consider.

[carrrlangas]

but it cannot be compared directly to bayer CMOS sensors. There are many things to consider.

You don´t need to consider every features / options a system provides, just the ones you need for the task. I am waiting for a DP2 M I ordered. Why couldn´t I compare the K-5IIs with the F28/2.8 for a landcape as I usually do? I certainly will but expect the DP2M to show it´s strengths... Besides, the comparison will provide some insight in how I´ll need to expose with the sigma, dynamic range difference, lens flare and contrast, etc.

[pinholecam]

Anyway, it seems the DP1M, with it's matched lens/sensor, could fit that nitch nicely, and the new screw-on Ricoh GW-3 will allow it to serve as an extreme wide angle lens as well. Sure it won't do color well past ISO 400 or so, but for a focal length this wide, I'd expect to use it at base ISO for landscapes, up to ISO 400 on the street, and as a B&W cam for indoor shots. The only real qualm I have is focusing without a through-the-lens viewfinder, but with the DOF given on a 19mm lens, I expect that won't be too rough.
)

I seem to have missed this thread for a while, now that its resurfaced, here are my few cents.

Dp1m+GW3, been there, done that (even put up some samples on serious compacts)
Basically IQ degrades at the last 1/8 other edges.
The horizontal edges are fine though.
So in a landscape orientation, you'd get more 'height' but gain little on width.



For those thinking about the comparison, I do have the A7, DP1m, K30 (had K5), as well as the FA31ltd.
From looking at 1:1 on the images (even if they are not the same), I already know that the DP1m cannot be beaten for high res stuff.
If I'd zoom in to 1:1, I would see any text or fine detail/texture as readable on the dp1m.
With the A7 or K30 with the same lens, I'd be looking at an 'interpretation' of it.
This is even with deconvolution sharpening.



But as I have mentioned before, one is a specialized camera, the others are generalist.
Which is better depends on which side of the bed I wake up from for the day