Thursday, June 20, 2013


I have been using Nikon equipment all my life. In fact, I took my first photograph, decades ago, with a glorious Nikon F and a 50mm, and have been using Nikon since. Nevertheless, I consider myself brand-agnostic: I do not use any brand exclusively, nor do I refuse to use a particular brand on purpose; during my photographic journey, besides my faithful Nikons I have been shooting with Leica, Hasselblad, PhaseOne, Aptus, Canon, Fuji, Sony, Holga, you name it... I am always looking for the best tools for the job at hand, and I enjoy experimenting with different equipment, whenever something new and interesting comes out on the market. At the moment, to create images for my VIERI BOTTAZZINI FINE ART PHOTOGRAPHY business I am using mainly a D3X and a D800E, both wonderful cameras, and I am quite happy with Nikon's offer of cameras & lenses at this point in time.

But - there is always a but, isn't it?

But last fall I got interested in the Sigma DP Merrill series of compact cameras, and decided to give the Foveon technology a try. I was hooked. The results coming out of these small, portable cameras with fixed focal lenses were so good that I started thinking about getting a Sigma SD1 Merrill to enjoy the Foveon sensor in a DSLR body, with the choice of lenses and accessories that this allows. Finally, a couple of months ago I took the plunge - got the camera and a few lenses and started experimenting with it (see my first impressions about the SD1 Merrill on the article ANSWERING FOVEON'S CALL: FIRST LOOK AT THE SIGMA SD1 MERRILL if you are interested).

Of course, since I am always purpose-driven when I buy and review equipment for my personal use, my sole concern is this: will the Sigma SD1 Merrill and Sigma lenses be up to the task of creating artworks for my business? How will they compare with my Nikon D800E and its lenses? Will the difference in sensor size (Full Frame vs APS-C) make a big difference when it comes to image quality? Will the prints be good enough to make me and my customers happy?

The Sigma SD1 & the Nikon D800E with some lenses

To answer this questions and decide what to use on my outings from now on, I decided to do a thorough comparison between the Sigma SD1 Merrill and my Nikon D800E with as many lenses as possible. Here we go!

No contest here: the D800E is without a doubt the feature-richer camera of the two, offering among others (in no particular order) video, electronic level, live view, possibility to use an external GPS, etc. that the SD1 Merrill lacks. As well, I wish the SD1 Merrill offered double card slots like the D800E does: being able to instantly back up your pictures in the field on two different cards provides you with an extra layer of protection that might save your bacon if anything goes wrong with your main card. More, I also love the D800E's shutter on the viewfinder's eyepiece, which is very useful to prevent stray light to get in from the VF during long exposures, something missing on the SD1 Merrill. On the other hand, the SD1 Merrill offers better timer / mirror up control, a more logical organisation for its custom feature sets (selectable on the mode dial) and the very useful QS menus for quickly setting up the most frequently used parameters. The D800E has a traditional status LCD on top of the camera, which is very useful to quickly check settings without looking into the viewfinder; on the other hand, the SD1 Merrill got rid of the status LCD, and your status information will appear on the main screen pressing the FUNC button. Personally, I prefer Nikon's solution, but your mileage may vary.

The D800E's features are broadly customisable via the menus, even too much so for someone, while the SD1 Merrill is a monument to the Spartan Camera, following a sort of "the less customisation the better" philosophy that might appeal to some and not to others; however, as a big plus it provides you with many different ways to reach the main controls (direct dedicated buttons, the QS menu and the regular menus). If you are a professional, in short, the D800E is a more complete camera, one that you can use for different jobs and customise to suit various shooting styles; while it excels as a studio, landscape and fine art camera, you can use it in a pinch for sports and PJ as well (though a Nikon D4 would of course be better). The SD1 Merrill on the other hand is a more limited tool; it excels at Landscape / Fine Art and in the studio (with people as well, if you shoot at a slow pace). However, having a choice I'd never take the SD1 Merrill on a PJ job, and it wouldn't certainly be my first choice to shoot sports either.

So, while feature-wise the clear winner here is the D800E, according to what you shoot the SD1 Merrill can prove to be the best camera for you if its features are enough for your work and if you are one of those guys that don't like to fiddle with overly complicated menus and like to set the camera once and forget about it. For my kind of work, both cameras provide me with all the features I need; of course, I'd appreciate it if the SD1 Merrill offered at least live view, electronic level and GPS, and a way to close the viewfinder's eyepiece during long exposure as well.

Torc, Ireland - Nikon D800E, 50mm f/1.4 AF-S, 1 sec. @ f11 (multiple shots, stitched)

As I said, I have been shooting Nikon cameras all my life. Since the advent of digital, in particular, I used the D70, D2X, D2XS, D300, D700, D3, D3X and the D800E: I can easily say that I can operate the D800E with my eyes closed and one hand behind my back. I always felt that ergonomically Nikon cameras are among the best out there, and I am really comfortable using pretty much any of them. That said, contrary to my expectations picking up the Sigma SD1 Merrill was surprisingly easy: the camera felt really good in my hands, all controls are placed very logically (with the exception of the ISO button, which requires a bit of contortion to be reached with your RH index finger!) and those that aren't located where I expected, I managed to get acquainted with in a very short time.

Both cameras feel really solid in the hand, very well built and easy to hold and use. No winner here for me personally, I'd be equally happy to pick either camera and start shooting with it.

Mantua, VII - Sigma SD1, 35mm f/1.4 HSM "A", 30 sec. @ f8

Auto-focus. The D800E is the clear winner here: the 51-points CAM-3500FX is faster, locks more decisively, hunts less and (especially important for me) works between -2 EV and 19 EV, while the SD1 Merrill's AF system works only between 0 and 18 EV. Personally, for my Landscape / Fine Art work I only use the central AF point; I never need AF tracking or 51 AF points (nor 11, for that matter). I only prefer the D800E's AF over the SD1 Merrill's for its low light capability and for that bit of extra decisiveness with which it locks focus. That said, the SD1 Merrill's AF system almost never misses, is about as fast as the D800E's and if you work like I do (using the central AF point to focus & recompose) you'll not notice much difference between the two.

Continuous shooting. The SD1 Merrill is a faster machine-gun shooter than the D800E (5-6 frames per second vs 4 fps at full resolution); however, the D800E has a deeper buffer, is faster in writing to the card and keeps shooting after the fourth frame, whereas the SD1 Merrill will stop after the sixth (14th if you shoot at Medium Speed). Since I wouldn't use either cameras anyway for any situation where machine-gun shooting is required, I'd call this a toss up.

Writing files speed. The D800E works just fine, even if it's not the fastest camera out there; you never wait for the camera, you can playback your images almost instantly after shooting them and you can access all camera's functions while it's writing files to the card - considering the amount of data that the camera is moving around, it is a pretty impressive result. The SD1 Merrill on the contrary feels like the camera of yesteryear: it stops working almost completely when busy writing files to the card; taking a new picture is pretty much the only thing you can do during such times. In this respect, the SD1 Merrill feels like a Medium Format digital back. My experience tells me that the SD1 Merrill's buffer is just fine: you can take up to 14 pictures in continuous shooting (at Medium speed) before filling the buffer, so I think that the problem is in processing the data and / or writing them to the card. The SD1 Merrill uses Dual TRUE II processors, and while Sigma touts that "the Dual TRUE II processes vast volumes of data at blazing speed", with all due respect to Sigma I'd have a different opinion about what "blazing speed" means. This is the only really REALLY annoying performance issue I have with the SD1 Merrill, and I hope that Sigma will either fix in it FW (if at all possible) or in the next iteration of the camera. In short, when it comes down to file writing speed, sadly there is no contest: these two cameras aren't just in different leagues, they are playing two different sports...

Shutter & Mirror. The Nikon D800E features a classic, heavy-duty Nikon shutter & mirror assembly: the electronically controlled vertical-travel focal-plane shutter is solid, has a long lifespan, features a quick-return mirror and shutter speed ranging between 30 sec. and 1/8000 sec. plus Bulb. However, it's certainly neither silent nor vibration-free. The Sigma SD1 Merrill also features an electronically controlled focal plane shutter with quick return mirror; however, Sigma opted for separate motors for shutter and mirror, and created a very silent, incredibly well damped shutter / mirror assembly; shutter speed ranges also between 30 sec. and 1/8000 sec. plus Bulb (though this is limited to 2 minutes in Extended Mode, 30 sec in normal mode). As a result, the SD1 Merrill is much more silent and creates much less vibrations than the D800E. Flash sync speed is 1/250 sec. (1/320 sec. in extended mode) for the D800E, 1/180 sec. for the SD1 Merrill. Therefore, unless you really need the D800E's higher flash sync speed, the SD1 Merrill's shutter wins hands down here.

Battery life. The Nikon D800E uses the EN-EL15 battery, with a CIPA life of 900 shots. The Sigma SD1 Merrill uses the BP-21 battery, with a life of about 300-400 shots (my measurements, Sigma doesn't provide a CIPA standard measurement). While with the D800E one battery is enough for a day shooting in the field, with the SD1 Merrill you'll need to carry at least one spare battery to be safe.

Devil's Corn Field and Mesquite Dunes
Devil's Cornfield & Mesquite Dunes - Nikon D800E, 70-200mm f/4 @ 110mm, 1/250 sec. @ f8

Both camera meter slightly conservatively, underexposing by 1/3 to 1/2 stop to protect the highlights; when using a f/1.4 lens, both cameras tend to underexpose around 2/3 to 1 stop shooting wide-open. Stopping down improves things, and around f/5.6 both cameras meter much closer to a correct exposure. While I understand the physics of shooting a lens wide-open, I am surprised that manufacturers didn't implement an automatic meter compensation of sorts for this.

Both cameras offer evaluative, spot and center weighted metering modes; the D800E of course offers more customisation options than the SD1 Merrill. In the end however they behave in the same way, underexposing more or less equally. No clear winner here.

Swans in the park - Sigma SD1, 85mm f/1.4 HSM (127mm FOV), 1/4000 sec. @ f1.4

The Nikon D800E's viewfinder cover 100% of the image, is very bright and a pleasure to use; besides focus points, shooting info etc., it offers an electronic level which I find very useful. The focussing screen is replaceable, a great feature for people using manual focus lenses. The Sigma SD1 Merrill's viewfinder is indeed particularly good for a APS-C camera; is very bright, covers 98% of the frame and offers shooting info and focus points, but no electronic level. The focussing screen is fixed, full matte and while it works pretty well when used in manual focus you don't have the possibility to replace it. I'd rate both viewfinders equally good, with the D800E having the advantage of featuring a replaceable screen and an eyepiece shutter, particularly useful to avoid stray light coming in both during long exposures and when using Live View.

The Nikon D800E's LCD is 3.2" and has a resolution of 920.000 dots. The Sigma SD1 Merrill's, on the other hand, is 3" with a resolution of 460.000 dots. The D800E's LCD wins here: brighter, more detailed, clearer and better in all kind of light.

Le Corti - Nikon D800E, 85mm f/1.4 AF-S, 1/250 sec. @ f4

When you buy a camera, you buy into a system: according to what you shoot, it is wise to examine the system in details and decide accordingly. For instance: if your work requires you to use Tilt-Shift lenses, then the Sigma is a no-go, and either Nikon or Canon will necessarily be your system of choice. More, if you need an extremely wide angle range T-S lens, then Canon is the only choice short of a technical camera sporting a digital back. If you need weather-proof lenses, then Sigma's offer pales compared to Nikon's or Canon's. One more thing to consider is that while Nikon, Canon, Sony and Pentax cameras can use third party lenses & accessories besides their own, if you get into the Sigma system you'll only have access to Sigma lenses and accessories. On the other hand, if you need lenses such as a 120-300mm f/2.8, a 150mm f/2.8 Macro, a 180mm f/2.8 Macro, a 50-500mm super zoom, a 18-35mm f/1.8, an esoteric fish-eyes and so on, or if you just are into the Foveon's look, then Sigma is your brand - of course, you can also use Sigma lenses on your favourite brand of camera (without the Foveon sensor and its particular look, though).

In short, buying into Nikon with the D800E you buy into one of the most developed camera systems out there; the lack of a 17mm Tilt-Shift lens aside, between OEM and third-party lenses & accessories there is nothing you will end up wanting. Buying into Sigma with the SD1 Merrill, on the other hand, as things are today you buy into a system that - while quite developed, and more so than some of the competition - isn't as developed as Nikon's. As always, only you can say if this is a serious limitation for your kind of shooting or not; however, it is something you'll seriously have to consider before jumping in one system vs the other. For Landscape / Fine Art work, with the exception of Tilt-Shift lenses, Sigma offers a pretty complete system, one I felt comfortable with when I decided to move on with my purchase.

Mantua, III - Sigma SD1, 50mm f/1.4 HSM (75mm FOV), 3.2 sec. @ f8

Comparing two completely different imaging systems is not an easy task, and even less so in a case such as this. We are comparing cameras with different sensor sizes, so lenses' fields of view will not exactly match, nor will depth of field at equal apertures; there isn't any software able to develop RAW files from both cameras; sensor technology is completely different; etc. All this considered, my methodology for this comparison has been as follows:

- Shoot 3 sets of images for each set, re-setting AF every time (using Live View on the D800E), choosing the sharpest set of the three;
- Shoot RAW with both cameras in A mode without any exposure compensation, to be able to compare the two cameras' metering behaviours; 
- Set the WB using corresponding preset WB in each camera (i.e., shade) and correcting it in software via a spot WB on the exact same point of the images;
- Develop the files in Sigma Photo Pro and Nikon Capture NX 2 (since no available software is able to develop the RAWs from both cameras, I went for the softwares I'd use with each camera for my own work); in particular, I sharpened both cameras' files as I normally do for my work (setting Sharpening to -1.0 for the SD1 Merrill and using a combination of High Pass + Unsharp Mask for the D800E); I turned off all Noise Reduction and Distortion Correction options; I left on CA & Fringe correction for both cameras;
- Compare lenses providing as similar as possible an angle of view on both cameras (i.e., 35mm on the SD1 Merrill vs 50mm on the D800E);
- To help you evaluate sharpness, I decided to upload files at native resolutions, as well as files resized to 30 MP (6720 px wide): the (arbitrary) choice of 30 MP has been done according to Sigma's claim that "the luminance resolution of this sensor is, in fact, equivalent to that of a 30MP CFA sensor as measured on the standard B&W resolution chart used in conventional digital camera resolution testing". Of course, such a choice should favour the D800E by some margin, since it downsizes the D800E's file while upsizing the SD1 Merrill's. However, I found the results interesting enough to be worth a look.

One last consideration. While there are many different ways to do a comparison such as this, I decided to proceed in a more user-oriented and result-oriented way rather than embarking in a classic studio comparison with resolution targets and so on. My goal when using either camera is to shoot Landscape and Fine Art, create large high-quality Fine Art prints of my images and sell my artwork. I don't shoot and sell images of resolution charts; my shooting distance and light conditions are very different than those required to shoot charts; etc. Hence, I decided to compare lenses in such a way as to get results as relevant as possible to my kind of shooting.

In this first instalment, I decided to start with a classic, the "fast normal", comparing the Sigma 35mm f/1.4 DG HSM on the Sigma SD1 Merrill vs the Nikon 50mm f/1.4 AF-S on the Nikon D800E.

Here is the usual, uninspiring test scene:

Uninspiring test scene, Sigma SD1, 35mm f/1.4, 1/640 sec. @ f1.4

Now, let's see some centre crops (600x600 px, 100%), at the original resolutions of 36 MP (D800E) vs 15 MP (SD1 Merrill):

As you can see, in the centre the Nikon 50mm f/1.4 is just OK wide open, gets better at f/2 and is very very sharp between f/2.8 and f/8 - no difference in centre sharpness between these apertures. At f/11, diffraction kicks in, but the image is still very much usable.
The Sigma, on the other hand, is amazingly sharp wide open, and stays like that until f/8, where diffraction starts showing its effects; f/11 is softer still, but quite usable if you need the depth of field.

What is clearly evident is how the SD1 Merrill's colour rendition changed between shots: these photographs have been taken at a few seconds interval and processed exactly in the same way; I can exclude any possible external cause for the colour changes, such as light changes or the like. Out of curiosity, I tried and managed to bring the picture to look exactly the same in SPP in a few seconds, but I chose not to do so in order to let you appreciate the camera's behaviour when left on its own.

Now, let's have a look at some corner crops (600x600 px, 100%), again at the original resolutions of 36 MP (D800E) vs 15 MP (SD1 Merrill):

In this comparison, again the Nikon is softer than the Sigma wide open; they both become very sharp at around f/5.6, with the Nikon taking the lead. At f/8 they are equally sharp, and at f/11 diffraction starts to make things softer again. For both lenses, f/8 is the sweet spot if you want maximum sharpness into the corners.

From this crop, you can also see how the Nikon 50mm vignettes much more than the Sigma 35mm, which is to be expected since the Nikon is a full frame lens on a full frame camera, while the Sigma is a full frame lens on an APS-C camera (thus using the sweet spot of the lens only).

While the D800E has evidently more resolution, you can see how the SD1 Merrill holds its own very well; its files show actually more micro-contrast and fine detail (see inside the tiles) than the D800E's files. Indeed the rendering of homogenous coloured areas looks more detailed, more tridimensional in the SD1 Merrill.

And now, let examine some centre crops resized to 30 MP (600x600 px, 100%):

Now this is an interesting comparison, since Sigma claims that the 45 MP Foveon sensor used in the SD1 Merrill has a luminance resolution equivalent to that of a 30MP Bayer sensor. Here, of course, I am presenting a colour test scene, not a black & white one designed to test luminance resolution. However, the upsized file from the SD1 Merrill holds details very well against the downsized D800E's file, more so if you consider that downsizing a file improves noise and detail, while upsizing it exacerbates noise and reduces detail. Again, if you check the surface of the tiles, you still can see more micro-contrast and fine details in the Sigma files compared to the Nikon's.

Now let's look at some corner crops, also resized to 30 MP (600x600 px, 100%):

Here slightly downsizing the D800E's files improves their perceived sharpness, while upsizing the SD1 Merrill's ones exaggerates the un-sharpness of the original image. Indeed the Nikon files looks better here.

I mentioned above that the SD1 Merrill resolves much more detail than the D800E in areas of homogenous colours, such as the surface of the small tiles presented in the samples above; so I thought I'd post a 100% crop of an area of the original image that illustrates the point pretty well:

The above crops come from files at original resolution; we are talking about 36 MP for the Nikon D800E vs 15 MP for the Sigma SD1 Merrill, and yet - as you can see - the SD1 Merrill's files shows much more detail in this area of the image than the D800E's. It is clear to me, both from this and from hundreds more pictures I have shot so far with the Foveon sensor vs those thousands I have shot with Bayer sensor cameras, that when demosaicing Bayer images of such homogenous areas the algorithms struggle to render the fine details that the Foveon sees with ease.

The D800E is a better low-light camera than the SD1 Merrill; even at base ISO, the SD1 Merrill shows more noise in the shadows than the D800E, and I wouldn't recommend the SD1 Merrill for critical colour work at ISO above 400; if you convert your files to black & white using the Monochrome mode in SPP, on the other hand, you can get very clean images up to ISO 1600 and even 3200 if you expose carefully.
The Nikon D800E is a more forgiving camera when it comes to metering; its files have more room to play and recover, and more dynamic range immediately available for you to play with. With the Sigma SD1 Merrill, on the other hand, you'll have to be much more careful. Blowing highlights is not a good idea, since there is little to no room for recovery there; that said, I'd definitely recommend to expose to the right, to get as much light as you can to the shadows without blowing the highlights. If you'll do so, you'll be able to squeeze a lot of dynamic range from your files, almost as much as what you get out of the D800E's; however, the SD1 Merrill is much more unforgiving towards those who don't pay attention - behaving more like a digital back in this respect.

Now let's see how these two camera / lens combinations behave as far as vignetting:

As previously noted, the 50mm Nikon vignettes much more than the 35mm Sigma; of course, as mentioned above, using a full-frame lens such as the 35mm f/1.4 Sigma on an APS-C sensor camera such as the SD1 Merrill - thus using the centre area of the lens only - helps reducing vignette. Still, if the lack of vignetting is important for you, the Sigma outputs a cleaner file than the Nikon. Again, you can observe here the SD1 Merrill's behaviour when it comes to colour differences between shots.

Finally, let's see how these camera / lens combinations deal with out-of-focus areas (focus is on the ColorChecker card):

Judging the rendering of out-of-focus areas and how a particular lens draws the world are very personal matters that depends on one's artistic vision and taste; therefore, I will not comment and leave it to the viewer to decide which lens draws closer to one's own vision. Technically, you can notice how the Sigma images show some "Bokeh fringing", that is the green / magenta cast you can see at f1.4 & f2 on the books on the far top left (green on the left, magenta on the right of some books' edges). You can correct that to some extent in Sigma Photo Pro or in Photoshop; of course, the Nikon ones do not show this behaviour to begin with, and do not need any extra steps to correct it. As well, you can check out the ColorChecker card in the images above to see how the two cameras render colours: to make things as "equal" as possible, I did a spot WB on all images on the third grey square from the top, and used both SPP's and Nikon Capture NX 2's "Standard" colour profiles.

For your convenience, CLICK HERE to download all the test images above and examine them at your leisure on your computer.

In short: either cameras (using proper lenses and proper technique, of course) will enable you to print up to 17" wide getting really wonderful, impressive results that needs to be seen to be believed; you can print up to 24" wide with great results, if you know what you are doing. Of course, talking about prints on the internet makes little sense, one literally needs to see them to believe; so I am afraid you'll have to take my word for it when I say that I'll be equally comfortable printing and selling artwork based on files from either cameras.

The Sigma SD1 Merrill is a 15 MP APS-C camera selling for less than $2.000 US; the Nikon D800E is today's resolution champion of Full-Frame DSLRs and sells for more than $1.000 US more than the SD1 Merrill. Nikon lenses are considered among the best for 35mm DSLR, while Sigma has a somehow spottier reputation; Nikon lenses of comparable focal length and speed cost generally at least 1.5 times Sigma's offer. However, Sigma uses a Foveon sensor which reads full colour information at each photosite vs the classic Bayer sensor of the D800E that needs to interpolate colour information, and this is what makes it unique.
By itself, just being able to compare files from the Sigma SD1 Merrill against the Nikon D800E's tells a lot about the quality of the SD1 Merrill; while the D800E wins in sheer resolution, the SD1 Merrill produce files that have exceptional micro-contrast and fine detail, especially in areas of the image of homogeneous colours (see samples above). The two cameras' colour rendering is also very different: while colours can be manipulated and adjusted in the digital darkroom to no end, of course someone will prefer the one vs the other; however, both cameras produce very good files with very pleasant colours to my eye, files that are perfectly usable as a base to create wonderful images.
The Nikon D800E is a more forgiving camera when it comes to metering, while the Sigma SD1 Merrill will punish you with shadow noise and reduced dynamic range if you expose carelessly. However, if you treat it well, you'll be able to get fantastic images from the Sigma as well.

Which camera would I recommend for Landscape / Fine Art work?

Both, actually. Definitely the SD1 Merrill is the best bang for your buck if you are starting from scratch, or if you want to step up your game and haven't already invested heavily into one particular system; however, if you already have a D800E, I'd not sell it to get the Sigma unless you felt hopelessly in love with the Foveon look. If you shoot stuff other than Landscape / Fine Arts and you can't afford to maintain two systems, the D800E is definitely the choice that makes more sense of the two; however, if you are a romantic like me and do not mind spending a bit of time to get your images and to process them carefully one by one, for Landscape and Fine Art work the Foveon look has something special to it that to me is worth working with - no matter the headaches that Sigma's software will give you to get there.

I will keep the Sigma SD1 Merrill for now and bring it with me to my next outing in Scotland (leaving the D800E at home this time over), to work with it exclusively for about a month and see what I can make it do.

I have more lens comparisons in the works involving the D800E & the SD1 Merrill, so stay tuned!

Now for the technical stuff: the quick and dirty product shot at the beginning of this review has been taken with the Sigma DP3 Merrill; light has been provided by one Metz Mecablitz 20 C-2 on camera, sided by 2 Nikon SB-700 Speedlight used as remote slaves on manual. All other photographs data are as noted in the comment under each image.