The Sigma Bigprint Unofficial National Tour

So I mentioned in the previous post that Sigma digital camera devotees are a small but hardy band. Many of them hang out on the Dpreview Sigma Forum, withstanding a regular barrage of slings and arrows from trolls. And the members of the group (the sincere ones, not the trolls) are friendly and supportive of each other—some of them have even met up for photography outings and such.

Quick background, even though it's well known: Sigma has the exclusive rights to the Foveon sensor, which works like color film by layering the photodiodes one on top of another instead of placing them side-by-side as in a more conventional Bayer-type sensor. This allows each photosite to record RGB color information and gives Foveon sensors higher resolution than Bayer-array sensors. Unfortunately, this has led to a semipermanent dispute: Sigma likes to say that each photosite counts for three pixels, since it detects all three colors whereas a Bayer array only detects one; so it would call a sensor with five million photosites, for example, a "15 megapixel" camera. Competitors and their followers call foul on this, feeling that a photosite should equal a pixel, and claiming that a 5-MP Foveon sensor doesn't actually quite equal the resolution of a 15-MP Bayer-array sensor.

Arguments ensue.

The latest and greatest Foveon sensor debuted in the Sigma SD1 in 2010. It has a 3200x4800, 15.4-megapixel Foveon X3 sensor, which by Sigma's nomenclature means it's a 46-megapixel camera. (See preceeding paragraph.)

The Sigma Merrill Foveon X3 direct image sensor.

So anyway, let's let that sleeping dog lie where it lays, and move on. Recently, two denizens of the Dpreview Sigma Forum, Rick Decker and Kendall Helmstetter Gelner, had some 2.5x4-foot (30x48 in., ~76x121 cm) prints made from Sigma SD1 files, loaded them into a big cardboard box, and asked who would like to see them. The prints are currently on a circuit of the nation, wending their way from the home of one American Sigma Forumer to the next. (There might be a European tour as well, eventually.)

TOP reader and occasional contributor Jim Kofron was second or third on the list. Since he's in Kenosha, Wisconsin, not far from me, he kindly invited me down to have a look.

Jim Kofron holding one of the touring prints from the SD1.

We met in a deserted elementary school lunchroom; Jim's wife Jennifer is a drama teacher, and a play (which Jim photographed) had just ended. We spent a pleasant hour unrolling the prints and inspecting and discussing them.

The indiputable conclusion: SD1 files are a lot closer to the output of 46-MP Bayer-array sensors than to that of 15.4 MP ones. Foveon sensors have a particularly beautiful output; they have a sort of vividness and clarity that, when everything clicks, can be quite lovely. Some of the prints, even at that size, were really good—just as photographic prints, never mind the technical forensics.

Really big prints are a curious subgenre of photography, one that's been given undue prominence of late because of the conflation of size with status in the art world. Art photographers now almost have to print big. For normal photographers, huge prints are problematic. Most of us don't have places to store them (a dirty little secret of the gallery world: a lot of collectors don't have places to store huge prints either, and some photographers whose work was actually intended to be larger have had to provide 16x20s to galleries because that's the biggest size some collectors want to buy). There are a number of aesthetic considerations that apply as well; some pictures work well big, others really don't. Technically, the bigger the print, the smaller the technical flaw it will expose or exaggerate.

The one Jim's holding works well at the very large size. The eye is drawn past the foreground rooftops and spire and in to the docked sailboats, which are presented in rich, satisfying detail.

Of the twelve or so prints Jim showed me (sorry, I didn't count), a couple of the pictures didn't "work for me" at the size; one picture "fell apart" technically because it just didn't have enough detail where it needed it (detail was, in general, in no shortage in the prints as a whole); one was relatively poorly printed (these weren't fancy custom prints, according to Jim, but inexpensive fixed-price ones); and one showed a serious lens flaw—decentering—with admirable, almost textbook clarity.

My favorites were a western rock face with petroglyphs visible, with two tiny people climbing around on the rocks; the picture Jim's holding, and a similar one of European rooftops; and—the best argument for the Foveon sensor, I thought—a woman sitting in a garden. About that last one: I think it's just impossible to have a reasonable complaint with it. It looks great, even at what I consider an excessive size. As an enlargement it beats medium format film fairly handily, in my judgement, and I'd even say that I've seen plenty of prints from 4x5 that aren't as rich in either color or detail. I've never seen a print from the A900 quite that big, but it at least holds its own with the oversized A900 prints I have seen.

The prints look rich rather than hyper-resolved—there's plenty of detail—detail enough, certainly, even from eight inches with my reading glasses on—but the detail doesn't dominate. It integrates as a whole with the picture.

A slightly-bigger-than-lifesize portrait worked wonderfully, too.

With any camera, some pictures will print better big than others. But with the normal and ordinary caveat that the picture needs to be one that "works" big, I'd have no trouble at all, personally, making four-foot-wide prints from the Sigma Merrill sensor.

The Sigma SD1 Merrill is available as of this month—as mentioned in the previous post, for a much more real-world price than what it cost when it first came out. And what might be, to some, even better news: Sigma plans to put the same 15.4 X3 (15.4/46 MP) Sigma Merrill sensor in a version of both the DP1 an the DP2—to be called, naturally enough, the DP1 Merrill and DP2 Merrill. (At the link, scroll down.) (The difference between the two is the lenses—one has a 19mm [28mm-e] lens, the other is 30mm [45mm-e].)

Many thanks to Jim for sharing the touring prints with me, and to Rick and Kendall for their generosity to their fellow Sigmaphiles. It's always nice to encounter a bunch of photographers with such a sense of community.

Mike

Send this post to a friend

Please help support TOP by patronizing our sponsors B&H Photo and Amazon

Note: Links in this post may be to our affiliates; sales through affiliate links may benefit this site. More...
Original contents copyright 2012 by Michael C. Johnston and/or the bylined author. All Rights Reserved.

Featured Comment by Rick Decker: "Mike: Re the picture Jim is holding. You may not have realized that you can see the top of a cruise ship which is in the harbor and then behind it in the distance the houses you see are on the other side of the bay. Also, which you probably noticed, you can read the company name on the crane on the right side. And this from a zoom lens. Thanks for the review. And your three favorites match mine."

Featured Comment by David Paterson: "I owned a Sigma DP1 compact for a short time. There was lots wrong with it but the colour it produced was way the best I have ever seen from a digital camera. The green part of the spectrum—usually difficult—was particularly beautifully and richly reproduced. It had a 5-MP sensor which competed pretty equally, detail-wise, with my two other cameras at that time—Nikon D2X and D300 DSLRs, both 12 MP. If the Sigma SD1 Merrill could take Nikon lenses, I'd be very interested."

Mike replies: Okay, but beware of those write times. Note that the above is not a camera review—it's a print review. As I said in my shortest-ever camera review of the DP2, "Shooting raw with the DP2 is like walking holding hands with a three-year-old: it's just gonna slow you down." I'm told if you shoot Raw with the SD1, write times are very slow by today's standards.

Featured Comment by Jim Kofron: "Mike, I really enjoyed our session with these prints. I've been a Foveon/Sigma fan for a long time, but I was surprised how good many of these prints looked. European Sigma fans should know that Kendall and Rick have apparently sent a batch of the prints over to Amsterdam, so the European tour has commenced!"

Comments

[Stan B.]

I'm sure it's a more than capable performer- pity it's so damn fugly.

[John Robison]

Yup, you really can't do math. a 3.5X4-foot print would be 42X48 inches, not 30X48 inches.

[Pedro Estarque]

We have so many choices for small cameras these days I feel almost spoiled. What a stark change from 5 years ago or so.
I gotta find myself another pocket for this new DP2.

[Stephen Best]

I'm more interested in better, rather than larger, prints. The DP2M looks very tasty, though I haven't seen pricing on this. Let's hope the dice Sigma uses for this purpose lands at the lower end.

[Mike Johnston]

"Yup, you really can't do math. a 3.5X4-foot print would be 42X48 inches, not 30X48 inches."

Sigh...thanks.

I hate typos.

Mike

[Aashish Sharma]

I wonder what Ctein would have to say about it (the Foveon sensor, its 'merits'/potential, etc). On a second thought, he may already have in some post much earlier. Would anyone please point me to such a post if there is any?

Thanks,
Aashish.

[Jonathan]

As a non-American I think that's the first time I've seen Kenosha mentioned outside of (probably) the best novel in the English language.

"You never did the Kenosha Kid"

Thanks for the reminder!

[Andrew Hughes]

Oh how I wish Sigma could see the value in this technology. Sony has made a great business of selling sensors (to Nikon among others) without damaging its photographic business. Sigma is hardly a player in DSLRs and it should concentrate on three things:

1. Licensing Foveon to others for mainstream DSLRs (preferably Canon but that is just my bias), especiallly as a larger than APS-C sensor. Can you imagine a 24x36 mm Foveon with 30 million photosites or more?

2. Making great DMCs with this technology. This seems to be on the way if they can get it right.

3. Making great value high quality lenses available in multiple mounts. Sigma are already pretty good at this.

Sure, make a DSLR for their fan base - it is great to see the credit (not refund) programme - but this is never going to make them any serious money. The three businesses described above would certainly do so.

Andrew from Addis

[MuTaka]

Quote from http://sigma-sd.com/SD1Merrill/highresolution.html
[i]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.[/i]
Although I think the camera is excellent low iso performer, Sigma are not doing themselves favor with this 46 Mp marketing. They remind me a lot of the early 12Mp Fuji marketing.

[Kenneth Tanaka]

It's a small tragedy that such a once-promising imaging technology had such ham-handed business management that it became relegated exclusively to Sigma. This "big prints" tour by some true believers is a bit sad. Foveon's window for big impact is passed.

[Mark Janness]

Mike,

There is a market for super large prints although it is some what specialized and tough to enter.

Commercial art placement in hospitals, clinics, and hotels often have larger wall spaces suitable for larger than average size prints.

I know the Princeton Express fitness clubs have installed supersized photos on their walls.

[James]

I hadn't heard about the pocket cams. If they are reasonably priced I would be interested

[James]

You're going to get me in trouble... *sigh* - I like the idea of the sensor. I just found out an interesting tidbit: SIgma uses a clone of the EOS AF protocol and have the same flange distances. So, people are successfully converting their sigma cameras to AF with EOS lenses.

[Paris]

Like Ctein, I prefer normal and longer lenses, so the DP2 looks really interesting. Semi-fast fixed normal lens, big Foveon sensor, small size. Definitely worth a look, as long as they don't introduce it at $7,700.

[hamondian]

Re the dimensions of the prints: Your text says 2.5x4 feet and the picture looks more like that sort of proportion, not 3.5x4 feet.

[Mark Janness]

Kenosha, Wisconsin: Former home of American motors. Birthplace of the Matador, Gremlin and other fine automobiles.

[Daniel]

When the original foveon sensor came out several years ago I was fascinated by the technology. I downloaded every file I could find and did a lot of test prints. Recently when I was going through some of my older prints I stumbled across several of these. They still hold up quite admirably and they still have a very unique quality. I remember that after considerable experimentation I came to the conclusion that they were comparable detail wise to files twice the size from bayer sensor cameras. Not quite the three times comparison that is made by the company but quite impressive nonetheless. The color has a clarity to it as well that seems to be a bit more defined somehow. Admittedly this is anecdotal but still, at twice the resolution from the latest sensor with that beautiful color I’m definitely intrigued.

[John Krumm]

If you live in a tourist town where cruise ships dock all summer, you can see super-sized prints becoming more common in the galleries (often garish and badly reproduced). I'd love to try the camera in our rainforest here in Juneau. I often feel like the lush greens aren't fully represented in some of my shots. But I'd like to see a comparison with the D800 before making any decisions. Perhaps the folks over at The Luminous Landscape will tackle that.

[ctein]

Dear Aashish,

I have never tested one of the Foveon cameras, and I don't engage in lengthy speculation about the performance of existing equipment I've never worked with. The odds of making a fool of oneself are simply too high.

That said, from what I do know technically about the sensor (quite a lot) I think people might be surprised at the results if you actually put the Sigma up against one of the Sony or Nikon 24 megapixel cameras (or wait for the new Nikon 36 megapixel). When all the metrics of image quality get considered, not just resolution but the actual structure of fine detail, color accuracy, exposure range, noise as a function of exposure, and so on, it's a complicated picture, and I doubt the Foveon would win. But it's not a very strong doubt, and I wouldn't even risk a two-bit bet on the outcome. I could be surprised.

That's not to say the look of the Foveon photos might not be substantially different. But then that boils down to what you happen to like, it's not amenable to any kind of objective analysis or testing. It's more like people who loved the look of Kodachrome over, say, Provia. One is entitled to like what one likes.

pax \ Ctein
[ Please excuse any word-salad. MacSpeech in training! ]
======================================
-- Ctein's Online Gallery http://ctein.com 
-- Digital Restorations http://photo-repair.com 
======================================

[ctein]

Dear Mike,

This whole pixel thing is not merely an argument between advertisers, although Sigma would like to portray it that way. There's been a long-standing, generally accepted usage for pixel as a technical term (which it is), which means a geometric point in the image. There have been corner cases where the term is used differently, and then it is carefully explained that the usage is different, but they are rare. It's broadly understood this is what a pixel means. Pixels can have any bit depth from one (like a fax machine) to thousands (hyperspectral imaging). It's still one pixel. Quite simply, Sigma is indulging in a lie, a deception.

There are many honest ways Sigma could have distinguished themselves in advertising like, “One of our pixels is worth three of theirs” or “The first honest pixel.” (Thank you Round Table!) You could even support those claims, at least within the realm of reasonably plausible advertising (after all, we are talking about advertising, here, not scientifically precise accuracy). Nice pixel-peeping comparison photos of “our” pixels vs. “there's” would be compelling.

Instead, they just decided to lie about it. And that was done with the full support of both Mead and Merrill, who were scientists of some standing who did know better.

Yes, I have a bit of a hate on about this, but tell me, how do you feel about the German lens makers deciding to come up with the trade term "APO" to fool people into thinking the lenses they were buying were apochromats? Betcha that one annoyed you, didn't it?

pax \ Ctein
[ Please excuse any word-salad. MacSpeech in training! ]
======================================
-- Ctein's Online Gallery http://ctein.com 
-- Digital Restorations http://photo-repair.com 
======================================

[anurag agnihotri]

This is off topic perhaps: Sigma has some really compelling (and cleverly thought out) lenses on offer. 150/2.8 macro, 85/1.4, 24/1.8, 30/1.4, 8-16....come to mind. What's more, they seem to be updating their key offerings with the new high end glass (flurite, whatever??) which is actually very very good. 17-50 2.8 OS and the 8-16 have amazing resolving power as tested by photozone.de and lenstip.

Now, there're so many horror stories about their quality control, their lenses front focussing or back focussing in particular, that people may be hesitant to test the waters with them.

Sometime i wonder weather this aspect, their QC not being in place, is a internet folklore or more real in nature...

[KeithB]

Someone should pull a "world's smallest drill returned with a hole drilled in it" and sneak in a print from a bayered DSLR.

8^)

[James B]

A few years back I actually read one of those threads on some website about whether the Sigma was one pixel or could be counted three times. Lord, I must have been bored or more probably sitting on one of those interminable global sales calls my company does while the rep from New Zealand goes on and on in exquisite detail about some sales prospect on the other side of the world. Sigma's technology or counting convention is certainly of no technical interest to me. However, comments were flaming back and forth.

Eventually one commenter made a statement (I paraphrase from memory) " If you hold up three coloured filters in front of your eyes at the same time, and look through them all, how many eyes do you have?"

Whether he or she was right or wrong, it pretty much brought the flame war to a rapid conclusion.

[Burt]

the dp1 and dp2 look like they could compete with the likes of the X100, although it's too bad they're f/2.8 and not f/2. sigma will have to improve ISO performance and have a good price point to make it stick... i guess we'll see how it goes.

[R Erickson]

Curiousity got me a DP2x when the price dumped. Despite all negative comments I couldn't let it pass. It's a very neat and tidy camera, small but easy to hold, fast to turn on, not noisy at all, fast to write to the card, and - the images are as advertised. They have more oomph. The Sigma app SPP is sluggish, but it does the job with raw files. I find it hard to imagine what the coming DP2-M will produce.

[Ludovic]

So basically, the DP1M is the thing I should have held out for instead of buying the X100?

[Antony Shepherd]

I've had a DP-1 since just after they came out, and while it's often an awkward camera to use, I've always loved the image quality.

The idea that they're producing new compacts using this fantastic new sensor sounds brilliant, although if I was going for a Merrill I'd choose the DP-2M.

[Joe Decker]

Horribly off-topic, I know, but I... visual memory is an amazing thing at times: I'm next to certain that that's Alesund, Norway, and I've only been there once, for a handful of hours.

[Andre]

I can't easily rationalize buying into another camera system, so the SD1 is out for me. The DP2m is on my short list though. I've been on the fence about the original DP2 since it came out, but the idea of having a pocket camera that's in the same ballpark as a full frame pro SLR is just too tempting.

[titus]

The autofocus of my DP2 is as fast as my twenty years old Volkswagen. The manual focus option is like Michael Schumachers Formula One Mercedes Benz. Clever and fast. The 1:2,8 / 41mm lens is superb. The Resolution of 5/15 MP are enough. Maybe the new compacts are faster and better for bigger prints. At the moment i am old schoolish. I will work side by side with my gear until one of us will die. God bless this decisive moment.

All the best.

[Sigma Photographer NY]

Not surprised at all, I have a sigma dp2 and just had some 20x30s printed up. They look just amazing. I'm seriously considering the merril sd1. I just wish sigma had some better glass available.

[Laopai]

I have a Pentax 645D that has a 40MP Bayer sensor. The largest landscape print size I got from it is 2m*1.5m after careful reviews of the RAW files by a local print house I trust.

I am waiting for the delivery of my SD1m now, upgrading from SD14. If SD14 can be a reference resolution wise, SD1 should be around 30MP equivalent. And this means print size up to 1.8m*1.2m should be very doable.

my $0.02.

[Luca]

I've just brought a DP2X, isn't a DMD, but is a lot better than original DP2.
I love it, but it's still unusable for fast action shooting, like all others P&S. But if you have time, the images are great an magics