Some thoughts on the Sigma DP2 Quattro (and comparison with Merrill)
I had a chance to try out the new Sigma DP2 Quattro last week, which is the first of the successors to the DP Merrill line. I also had something else much more exciting going on last week, so camera testing was pretty low on the priority list.
First, a little context for any who are new around here. I’ve shot almost exclusively with at DP2m for the past couple years, save for a brief affair with an A7r which I quickly returned (and picked up a DP3m instead) when I looked through pictures and prints and was reminded how much I love the Merrill files. I’m not going to use the “film-like” cliche, but they do have a rich, dense, organic feeling that I’ve never seen with any other camera. I’m particularly passionate about black and white work and the Merrills are outstanding monochrome cameras.
Most of my shooting is on multiday wilderness trips. I go to be out there, not to photograph, though I love taking pictures. I shoot almost exclusively handheld.
I’ve been pretty excited about the Quattro since it was announced, thinking it sounded like it’d be an even better monochrome camera with its mostly luminance-measuring top sensor layer and maybe a little better color performance above 400 ISO.
The Quattro shape has caused a bit of a stir and I feel a bit mixed on it. I do think it achieves the goal of more stable hand holding and I like the thinner profile, but the larger lens and rear-pointing grip makes it a bit large, especially if you wanted to carry more than one of the focal lengths that will be available. The stock lens hood on the Merrill is excellent; it does its job flare-wise and also does a great job of protecting the lens while carrying it around one’s neck. The Quattro hood provides basically zero protection and I think a cheap rubber hood would deal with flare better.
The Q is a bit better speed and usability wise, though I’ve honestly been fine with the M. The screen on the Quattro, however, absolutely kills the M. The lack of an EVF is one of my biggest gripes with the Merrill, but the nice bright crisp lcd on the Quattro makes seeing what you’re doing much less frustrating.
Now for the important part. If you’re looking for carefully controlled scientific comparison shots look elsewhere. I wanted to get out with both cameras and shoot the way I normally do. I’m a bit disappointed with the shots I did get; I had two quick evening hikes and was trying to get as many shots as I could before it got too dark, so no great images here. I do think I got plenty to get an idea of how the cameras compare. All of these are handheld with some rail-steadying in a couple. Quattro shots developed from X3F in SPP 6.0.3 and Merrill in 5.5.3. Sharpening seems to be different with the two cameras, Merrill shots are -.7 and Q at 0. NR completely off for both cameras. I adjusted contrast/tone curves in lightroom to try to match them somewhat but no color/sharpening/noise reduction or other settings have been adjusted. Click for full size JPEG exported from lightroom at 10 quality, no export sharpening.
Here you can see the difference in auto white balance on the two cameras. I almost always shoot in daylight WB, neutral color on my DP2M. It’s a bit less noticeable in the exported JPEG but you can see here that it seems that the Q doesn’t quite have the punch and sense of depth that the Merrill has. The developing conclusion seems to be that the Q slightly edges out the M in terms of absolute resolution but the M has more microcontrast and tends to bring out textures and fine detail better. So far I agree and in shots like this I prefer the Merrill rendering.
On distant foliage the Q appears a bit mushy compared to the M, although there are some areas of the above shot where you can see the slight resolution advantage of the Q.
The monochrome output seems more similar between the two cameras than the color output. Excellent tonality and noise is generally very pleasing and grain-like. The Q, however has a significant disadvantage at higher ISO. With the M, it’s possible to shoot at 1600 and use only the blue channel through SPP monochrome mode which significantly reduces noise and gives a much more usable image even at 3200. The channel mixer does not seem to have any effect at all on the Q when it comes to noise.
Here’s another full size ISO 1600 shot exported the same way:
I’m also not convinced that the Q gives better high ISO (in the Foveon world this means >400) performance than the M. The Q seems to have less saturation loss in the shadows, but highlight recovery is not nearly as good. With the M you can overexpose by quite a bit and pull it back in post, which helps significantly with noise. This doesn’t really work as well with the Q because highlights are lost completely much sooner than the M.
Some more comparison shots:
Below you can see how much more life and *desperately trying not to say 3d* pop the center formations have in the M pic
I feel like it’s probably wise to reserve final judgement for the release of a more final version of SPP 6 but as it stands I feel happy sticking with my Merrills and fat stack of batteries, especially at the current going rate of half the cost of the Q. The Q will probably appeal to folks wanting a more natural, smoother rendering for things like skin tones, but I kinda feel like if that’s what you’re after you’d be much better off with an x100/s, Rx1/R, A7s, etc, which will kill the Q in low light, be much more versatile, and still give gorgeous smooth color. The unique thing that the Merrill brings to the table over Bayer cameras is its pixel level sharpness, rich depth and microcontrast. Merrill files give such a sense of awe when you first open them that makes you forget what a pain it can be to get there and how limited and finicky the cameras can be; the Quattro just didn’t give me that sense in the short time I had it. Hopefully through firmware, software, or just people figuring out optimum settings there will be some kind of magic still hidden in the Q. I’ll keep watching.