1.12.2016

There is a lot of talk on the web about the new Phase One camera. 100 megabytes sounds good...



Nearly eight years ago I used to write articles and reviews for Studio Photography Magazine (another casualty of the digital transition) and in my role as a freelance technology writer I got asked to review some pretty cool stuff. Sometimes interesting boxes would appear at my doorstep. One day I was shocked and amazed when I got a box from Leica containing the brand new M8.2 and four brand new lenses. The Fedex or UPS driver left the box, unsigned for, right at the front door. I thought that was crazy and risky until I had a similar thing happen with a box from Phase One that contained their latest 40 megapixel, medium format camera and a trio of really nice lenses. $40,000 worth of cool camera gear just hanging out in front of my studio and thunderclouds looming in the northwestern sky....

But, as usual, I digress. What I really wanted to talk about today is the announcement of the latest Phase One camera, with 100 megapixels. The underlying question: Is this camera at all relevant to working photographers or enthusiasts today? I know there is a market for it but for photographers who are not employees of large corporations or museums; does this camera make any sense at all?

When I reviewed the 40 megapixel Phase One camera (based on a Mamiya body) the one thing I really liked about it was somewhat independent from the pixel count. I like the way the larger sensor emulated the look for medium format film cameras by allowing the use of longer lenses for the same angle of view as shorter lenses would give on smaller format cameras. Even a fast 80mm lens gave a different look when coupled with the larger sensor. One could say the benefit of the system was the way it drew images. 

While I understand the benefits of higher megapixel counts coupled with low noise CMOS sensors I think that the designers of medium format cameras (and sensors) have been going in the wrong direction. They should have moderated the pursuit
of higher and higher megapixel counts and pushed harder to reduce the cost of making larger sized sensors. If the selling point (to me and to other previous users of film medium format systems) is the way the longer lenses draw images (a faster ramp from focus to out of focus for equal angles of view) then the goal should always have been toward achieving a real, full frame 6x6 cm sensor and lenses to match. I wouldn't much care if the "dream" sensor had 40 or 50 megapixels of resolution instead of 100 if it offered a look that's not really achievable by any other technical or operational parameter --- and I feel that serious practitioners of portraiture and still life photography would readily agree.

The Phase One backs, the Leica S series and the Hasselblads of all variety are still hobbled with much smaller sensors than the film gate sizes of our previous commercial favorites like the 6x6 Hasselblads and the 6x7 Mamiya RZ 67 cameras of yore.  While the newest crop of sensors is far superior to film as regards noise I'm not sure that was a problem that most of us were looking to get solved. 

The latest Phase One Camera (post Mamiya buy out) is an amazing technical achievement that will appeal to enterprises that require enormous amounts of detail in their photo files. But the cost of ownership and operation is, for all intents and purposes, far beyond rational for a small photography business that works in the open market today. Beyond the $50,000+ price of acquisition (camera and accessories) and the $20K to $30K investment in lenses real thought has to be given to handling the enormous files generated. And much of the quality will be lost if the target for the images doesn't match the state of the art of the camera system. How do the advertised 15 stops of dynamic range map to large, high speed inkjet printers limited gamuts? What monitor will you use to glean the results of the 11,000+ long row of pixels generated in each file? Will most of the process require proxy files or will clients also update their systems in order to receive and work on these files?

For a closed loop operation (say an in-house photography resource for a  furniture maker) you can justify a certain cost of doing business, but suppose you are a working freelance photographer with regular clients whose biggest stretch is the creation of tradeshow graphics. If I shoot with a camera that generates enormous and hit bit depth images which create 400-600 megapixel tiff files just how many of my corporate clients (sitting in front of minimum spec Dell system) or my typical medium size ad agency art directors (sitting in front of 27 inch Apple iMacs) will be able to pull those files into their systems and work with the images in InDesign or even PhotoShop in an efficient and meaningful way? Damn few. 

While the Phase One's are truly aspirational tools the reality of using them and the real inflection of their value is suspect. I think we are approaching the point where film makers are with Panavision cameras and Arriflex's top cameras live; in the rental houses or in the company's rental inventory. Used only on the biggest projects and then only sparingly. 

I don't work in the arena of "spare no expenses" clients. I could never justify the benefits of this camera in the current market or even the current milieu. In fact, I am still more or less blown away by the detail I see coming off the D810, the Sony A7R2 and the Canon 5DsR. I live in fear of always running out of hard disk space. My clients already ask for "dumbed down" files that work on their systems.

I think it's wonderful that a company continues to push the envelope and I think it must be exhilarating for the photographer swho buy them to work with such impressive tools. In a way it helps create a new top of the market for users and camera makers. I'll pass on this revolution though; until they make a sensible, full frame 6x6 or 6x7 (I can always crop...) sensor that makes use of the miraculous optics that have already been created and are standing by to be rediscovered in their intended glory. 

How about you? Rushing out to buy the 100 megapixel Phase One? I'd love to hear your rationale....
To be shocked back into the reality of actual, visible improvements in use,  look no further than Kevin Raber's  from the camera over on Luminous-Landscape. Then look fondly at whatever camera you happen to have sitting next to you and smile, knowing it would have done just as good a job.

blog note: between 2008 and 2009 I spent about three months of studio time shooting with MF cameras from Phase One, Leaf, and Mamiya. I enjoyed the process and enjoyed the learning but was more than able to resist their charms. I have recently been playing with a friend's Leica S camera and while I like the feel and quality just fine, in many instances, with the exception of focus ramp differentials, I can't see much difference in file quality between that camera and my D810. Actually using and testing the cameras always trumps informed conjecture.

18 comments:

Marshall said...

Hm. Well, I suppose, if I end up as the sole winner of $1.4B (minus taxes, etc.), I might consider getting one, trying to take one great picture to fill that giant wall in my mildly oversized new house, and then realize that I probably don't need another picture I can print that large very soon.

Bill Pierce said...

That quick fall off in depth of field is exactly the reason I have kept my 8x10 view camera. Of course, I used to use it at f/64 and now use it wide open. It sits in the studio, too big for an old person to carry around, and the darkroom is nothing more than 4 small trays in a dark room and scanning on a relatively economical flat bed. Only one problem, I've been shooting APS C digital with an f/1.2 lens wide open and it looks sort of the same.

Kirk Tuck said...

Funny Bill. I'm having the same experience with a cheap, full frame Nikon and an old f1.4. The last view camera went away last year. A tinge of sadness.

amolitor said...

I *think* the big back from Phase is basically 645. The press release says "full frame" and everyone just copied that, but if you dig into the tech specs I think it comes out to around a 6x4.5 frame. Which is a heck of a large chip.

I think there's a basic problem with digital that every step is really just a small increment over the previous one, so we're not really noticing the changes in format. 645 is wildly more huge than an APS-C and M43 sensor, and will provide a wildly different look. But people won't compare it with that, they're likely to compare it to DSL full-frame, or maybe a 40Mpixel MF back, and they'll shrug.

One wonders if they'll offer up a 24Mpixel pixel-binned format? Even better color, manageable file sizes, medium-format look. You can do it in post, but you're going to wind up dragging these gigantic files around farther than is necessary, and you really OUGHT to demosaic with the binning in mind, I think.

Anonymous said...

Just like photographers. Big eyes. Small stomachs. Looking at the whole buffet table.

Kirk Tuck said...

We've been told for decades now that the limitation of sensor size has to do with wafer/die yields. More small sensors equals less failure rate equals more profit. Screw that. The engineers have had ample time to make big, dopey, non-complicated sensors on 300mm wafers. They just need to feel the lash of market demand. :-)

aurèle said...

Recently i started to use the Pentax 6x7 and the 105mm f2,4. The rendering is what drew me to it. And to be honest, i shoot, then scan it in a 36 mpx picture. It's what i found to be the Best intermediate size between détails, and manageable size. If one Day, someone produce a 6x7 with "only" 40 mpx or so, i would consider ponying up to buy it. More megapickles is to my hobby, Just too much. Even if i do some 30x45 cm and 45x60 prints out of it.

I really love that big neg rendering, and if i had a True 6x7 digital caméra, i would bring it everywhere to shoot. Despite thé size and weight.

Cpt Kent said...

I always wondered if the appear of the MF 'look' had something to do with the fact that our own eyes are ~6cm apart. Not smart enough to figure it out.

amolitor said...

Some subjects you can just stitch up the MF look, but not all.

Kirk Tuck said...

I"m not calling it stitching any more. I call it quilting.

Richard Leacock said...

Would I contemplate a nice 6x6 or 6x7 medium format sensor based in a classic styled body like a Mamiya 6 or 7, 3 or 4 leaf shutter lenses for the system with the technology today? Mmmmm...Yes. Yes I would. So with some spare cash in hand which manufacturer should I contact ; )

Anonymous said...

Buying the new Phase back and trying to justify it is like buying a Ferrari to use in going to the grocery store when a myriad of other cars will perform the duty very well.

I've owned several versions of the Phase One gear mainly using the backs with my Cambo technical camera. I cannot justify it but it is simply great fun to shoot this gear. Yes, I've ordered the new back and I already have the new body. I'm an amateur. I don't sell my work, enter photo competitions, exhibit my work, or have a website. It is simply me chasing light for the personal challenge of doing so.

I also own a D810 that I simply love. I used to shoot Canon gear but the D810 changed that quickly. The D810 is a magnificent camera that I plan to own and shoot for quite some time.

All the best,
Jeffrey L.

Noons said...

When it comes to size of image, I'm convinced these 100Mpixel backs are almost always unnecessary. Sure, they have some - very restricted - applications.
As you well point out, very few clients have the gear and facilities to deal with the resulting monster files.
And quite frankly, quilting (I agree, sounds better than "stitching") can handle a lot of situations that would otherwise require such gear.
I'm always faced with the "no movement" objection when I mention quilting. Really? Have a look at below, it was done with a D200, 2 images (who said quilting has to always be horizontal?) and been reduced in size so it shows reasonably well in my Samsung Galaxy S2 pad (original is much larger):
https://photos.google.com/photo/AF1QipMoOIyr7fj6J8gVG3xGUI6G3jP_ZIuinhpwveVH
How's that for "no movement"?

Mark Davidson said...

When I started using a Pentax 6x7 back in the 80's I loved the results I got just shooting snaps of my kids. Even 4x5 inch prints rendered a quality unobtainable with 35.

I have long felt that a6x7 sensor with 40 or 50 MP would make images of such beauty that we would finally stop lamenting the passing of film. Until we wanted 4x5 sensors....

Mark said...

> Screw that. The engineers have had ample time to make big, dopey, non-complicated sensors on 300mm wafers.

Canon's policy has been to stick with the most mature (some might less charitably call it outdated) process technology, presumably because they prioritise sensor yield and manufacturing costs over performance. Yet somehow here their cameras are not any cheaper than comparable Sony/Nikon bodies.

It is getting quite difficult to improve sensor performance - modern sensors are astonishingly efficient and are starting to run in to hard limits of physics. However, there is still a lot that can be done to dramatically reduce cost...

Anonymous said...

Mark > "presumably because they prioritise sensor yield and manufacturing costs over performance. Yet somehow here their cameras are not any cheaper than comparable Sony/Nikon bodies."

Lower costs = higher profits.

Daniel Walker said...

What happen to art

Kirk Tuck said...

Art seems fine. We had lunch together last week. He stuck me with the check...