Peter Lik and medium/large format photography

[glocke12]

Just got introduced to this guys work via his show on The Weather Channel...

He seems like a pretty talented, dedicated, ambitious, hardworking photographer. Have to admit though that some of his shots seem to me to be only average, while some are absolutely killer.

I was also shocked to discover how just expensive his gear is...I think in one segment of the show he mentioned the camera he was using was something like 20k, and I also read somewhere else that someone is making him a 75 mp camera that costs 150k.

Pretty shocked at those prices!

What exactly are the advantages of shooting medium/large formats? Higher dynamic range? Ability to make larger prints??

[firestarter]

If you want to shoot film, neither medium or large format has to be particularly expensive.

Medium format is (was) defined as 'rollfilm' cameras - cameras using '120' format film, with no sprockets and a paper backing (later 220 film was developed, with no paper backing). Images are 6cm tall, and a variable width depending on format - from 4.5cm up to 17cm and beyond. The advantages were it's cheapness and simplicity and area of film used (maximising resolution), the ease of correction (since the film was large you could scratch or paint edits on), plus it was big enough to print a contact-sheet and see the images clearly.

Some of these advantages still hold today. Image sensor noise is all about physics, and there's no escaping the advantage that a much larger sensor with larger photo receptors capturing more photons gives you. Typically MF digital is extremely high resolution, high dynamic range (16 bits of data per sensor too) and very sharp as anti aliasing (blurring) filters usually aren't used in front of the sensor.

Creatively, MF can allow very low depth of field photography (again, using the physics of the large sensor), and the modular nature of the camera means you can often use specialist lenses (leaf shutter lenses, with flash sync at all speeds).

I used to use MF all the time (I had a Hasselblad 503CW, a Fuji GW670II and a Mamiya RZ67 Pro II)

Large format is typified by the use of sheet film - usually 4x5 inches and up. Again, the larger surface area gives a big benefit in resolution and noise. Some really large format cameras (20x24inches) can produce final images directly (you can get polaroid film cameras at this size).

The big advantage with LF is usually the flexible nature of the camera itself. Since the back and the lens are usually separated by flexible material, the lens (the 'front standard') can usually be moved separately to the back. This lets you use tilts, shifts and swings to alter perspective and plane of focus. You can easily correct the converging verticals of a tall building, or use the Scheimpflug principle to give you front to back image sharpness by tilting the lens forwards.

[wilsonlaidlaw]

There is no cheap way to go MF digital. With DSLR's becoming so good and DRF's also, unless you are a dedicated professional or architectural photographer, the cost/benefit equations don't really work. The cheapest satisfactory route is a good second hand Hasselblad or Contax MF camera plus a second hand 30 to 40 MP back. That is going to set you back around $15,000+ by the time you have a few lenses. For new, double that and more for a Phase One, Hassy HDx or Leica S2 camera plus lenses or a bit less for Mamiya 645 AFD. If you want to go Alpa, Horseman or the incredible 160 MP Seitz - rob a bank first!

For much less money you could go with a Canon 5DII, Canon 1DS3, Nikon 3DX or D700 plus there are lots of second hand lenses out there. As a smaller alternative, you could use what I do, a Leica M8 and/or M9. The M8 (10.5 MP) is now plentiful second hand at around $2500 and there are 60 years of lenses to choose from. The M9 (18 MP) is just about now being seen, albeit few and far between, on the second hand market at around $6000. The M8 will print with a good rip engine to 18 x 27 inches and the M9 to 30 x 45 inches. Do you need more?

Also bear in mind that MF lenses are often less good than 35mm ones because they didn't need to be. This can be a limitation, when you are going for higher MP backs. The latest MF lenses have been specially optimised for high MP digital backs and there is an according price penalty. For example, the highly thought of Zeiss lenses for the older Hassleblads and Contax 645 cameras are thought to gain little benefit from a back of much above 35 MP, simply because they cannot resolve detail of any finer than this. The latest generation Hassleblad (Fuji) and Mamiya AFD II lenses (also used on Phase cameras) are said to be capable of resolving detail for up to an 80 MP back.

[snberk103]

Just to expand on what wilsonlaidlaw and firestarter have posted. There are advantages to MF digital because of the size of the sensor. I have a Mamiya/PhaseOne camera/data-back and I can see the advantages.

However, the primary reason I went this route is because I already had a large investment in lenses for a film-based Mamiya camera. To duplicate the glass in a DSLR (35mm sized - since Mamiya/PhaseOne is also a DSLR technically) ... anyway, to duplicate the glass would have cost just about the same as buying this very high-end MF camera and data-back.
Some of the older lenses have some issues, but they work "good enough" for what I do. And now I can add modern lenses (Both Mamiya and SK now).

My point is this, I guess. If you can leverage a MF purchase by res-using old glass then it may be worth while. If you are starting from scratch.... then unless you are charging a higher than average rate for your work, and you do a lot of work, then MF is probably not the way to go.

One other advantage of MF system is the modular nature of the camera/data-back. One can have more than one data-back for a camera (or more than one camera for a data-back) to take advantage of certain features that are found in more than one model. For example PhaseOne is selling a B&W only data-back. Or you can choose between data-backs that offer huge files, huge dynamic ranges, or the ability to expose an image for an hour. Or you can have multiple bodies (for back up purposes or for features) without duplicating the sensor part of a camera.

Cheers

[merkinmuffley]

I visit the Peter Lik gallery in Waikiki often, it's about a block away from a Starbucks I enjoy. Given his prints sell for $5k up to around $75k, he can afford any equipment he desires.

[carlgo]

It will be interesting to see what happens when sensor costs come down and 80 or 100mb (or perhaps even far higher) cameras are affordable.

Even now I am pretty much convinced that FX sensors are only fractionally more expensive to make than DX ones and so on down the line.

In any case, I don't buy into the lens resolution argument nor the one that you need 5' prints to see a difference. I'll bet you don't. And at galleries I don't see people stepping back to view photos, I see them pixel peeping.

I remember seeing professional 8X10s printed from Leicas and thinking they were sharp until I got to see 8X10 contact prints...wow, an entirely different thing.

[wilsonlaidlaw]

Originally Posted by carlgo

It will be interesting to see what happens when sensor costs come down and 80 or 100mb (or perhaps even far higher) cameras are affordable.

Even now I am pretty much convinced that FX sensors are only fractionally more expensive to make than DX ones and so on down the line.

In any case, I don't buy into the lens resolution argument nor the one that you need 5' prints to see a difference. I'll bet you don't. And at galleries I don't see people stepping back to view photos, I see them pixel peeping.

I remember seeing professional 8X10s printed from Leicas and thinking they were sharp until I got to see 8X10 contact prints...wow, an entirely different thing.

Sensor costs go up more than geometrically in relation to pixel counts. An FX sensor is about 6 to 8 times the cost of a DX. I believe the cost of a 50MP Kodak sensor is between $3 and $4k. One issue is that the fault rate on manufacturing of large silicon wafers is high. On a small sensor you just throw away the affected sensor once you have cut up the wafer. Commonly only one or two MF sensors are made per wafer. Costs will come down but less than you might hope. What will happen is that new technologies will come along and FX sensors will attain the resolving power and low noise that MF sensors have at present.