Experience log: the Hasselblad H4D-40 and medium format

_8022716 copy

I shot almost only Leica M and a couple of lenses for the majority of 2009/10; it completely changed the way I approached photography. I no longer focused on getting everything – every nuance, every detail, every highlight, every shadow – instead, the limitations of the camera forcing me to focus on the moment and the subject, making them strong enough and isolated enough that the edges and everything else in the exposure were clearly of secondary importance. Every image I’ve shot thereafter has followed the same principles: the importance of subject and quality of light on the subject have become indelibly ingrained onto my visual cortex.

A6557673 copy
Night in the city. H4D-40, 80/2.8

The second shift came when I focused on both accuracy of color and its use to consciously influence perceptual impressions in the viewer of an image as part of my experiments into the cinematic style of reportage; this added another tool to my repertoire. A third, less major, shift came after the Nikon D800E forced me to shoot mostly tripod based: I would shoot with precision, with more depth of field, and even more care over lighting; control was – and is – now the aim of the game. It’s had two effects for me: I shoot less with available light, and more with constructed light; also, I’m now working slower and frequently with a tripod. It means I shoot street very differently; it’s now about people, situations and context – almost abstracted people – and less of the ‘stolen moment’ type of photography I used to practice.

A6557650 copy
Decay. H4D-40, 80/2.8

Part of the reason for this is my shifted focus towards commercial work, part of the reason is because I’ve come close to exhausting the possible available-light street/ reportage locations available to me locally. I’m hoped that by changing the way I work again, I’d once again challenge myself to add a new skill to my repertoire, and thus up my game a notch. In hindsight, the problem was that the medium format options were either too similar or too vastly different to the way I was (and am) currently working with the D800E.

I had two options: buy a V-series 500CM or 503CW, and shoot film; there was also an ex-demo CFV-39 back available to purchase, which would give me a long term digital solution if I decided it fit me. The second option was a second-hand H4D-40. Both would represent a significant investment even before counting lenses, and thus I managed to arrange an extended test drive with the local Hasselblad distributor, Shriro – for which I am extremely grateful. I want to say upfront that it was not the quest for more pixels or image quality that led me to consider medium format; if anything, my clients have been extremely happy with the quality of the files from my D800E. Anything over that would be a bonus (or a curse for some, who are already struggling in handling the enormous 36MP 16-bit TIFF files the camera produces). Rather, I was looking for something that would force me to change the way I saw the world through forcing me to change the way I worked.

A6557637bw copy
Stalls. H4D-40, 80/2.8

From the get-go, the 503CW and CFV-39 were just too different. It wasn’t the waist-level finder (though it did make it very difficult to see the frame and focus under bright situations); it wasn’t the reversed directions (the V series Hasselblad waist-level finders do not invert and correct the image like a normal SLR finder). It wasn’t the manual focusing, or the need to wind the camera to re-cock the shutter between each shot, or the shutter speeds being on the lens (leaf shutter). It wasn’t the square format. Of all things – and this might seem odd when you remember that I’m an experienced rangefinder user – it was the assortment of crop lines overlaid on top of the focusing screen. The whole screen showed the 6×6 film negative. The first set of lines showed the entire capture area of the sensor – ostensibly a 1.1x crop horizontally, but rather more vertically – and a second set showed a square crop of the sensor, which gave something along the lines of a 37x37mm sensor.

A6557667 copy
Overpass. H4D-40, 80/2.8

Ostensibly, if one used the CFV-39 as a square sensor, the way that camera was intended to be shot, you’d be looking at a 1.6 linear crop, and a sensor that was about 50% larger than the 35mm full frame. (Note that cropped square, a FX frame becomes 24x24mm, and about 40% of the area of the medium format-square sensor.) I think somewhere between the combination of the multiple crop lines, the left-right inversion and everything else that was different, my brain shut down. I just couldn’t see anything other than what should have fit into the square 6×6 frame; digital on the V series was simply not an option for me, and it wouldn’t make sense to buy one solely for use with film especially since the H series can take both film and digital backs.

A6557694 copy
Experimenting with 2.4:1 widescreen on the street, I. H4D-40, 80/2.8

That left behind the H4D. Available in multiple flavors, the 40 made the most sense to me – it balanced price, specification and resolution. And it was available second-hand, which brought the price to the same ballpark as the CFV-39 and V-seriesd combination. The 40MP, 44x33mm sensor has a crop factor of about 1.36x, giving it a little under twice the image area of a full frame DSLR, and a tiny bit more area than the Leica S2’s 45x30mm sensor. Here, we have an interesting three-way comparison on our hands: the Nikon D800E, with a 36x24mm sensor and 36MP; the Leica S2, with 45x30mm and 37MP; and finally, the Hasselblad H4D-40 with 44x33mm and 40MP. From a raw sensor resolution perspective, near enough as makes no difference. The lenses would of course be another story entirely.

A6557705 copy
Experimenting with 2.4:1 widescreen on the street, II. H4D-40, 80/2.8

I’ve used the Leica S2 in the past, and found that it handled very well indeed for a medium format camera – in fact, it handled very much like my D800E, with the exception of being a bit larger and heavier, and having an enormously better viewfinder and somewhat simplified control system. In practice though, I didn’t treat it any differently from the D800E except that it had a single center focusing point – necessitating a lot of focus-and-shift-to-recompose. Focusing was fast enough that you could still use it as a street camera in much the same fashion as a regular DSLR. The H4D-40 was a bit different, though: firstly, though it has both autofocus and a gyroscope-adjusted True Focus system (the camera uses the gyro to determine how much it was moved by, then adjust focus by an appropriate amount accordingly depending on the lens) it was a much, much slower camera to focus than either the S2 or D800E. In fact, it was so slow that you couldn’t really do any spontaneous shooting with it; prefocus was the best way to go. Part of the problem is that it wouldn’t always lock on to your intended subject, and there was no real consistency to this – there were times whens it would find focus just fine on a low contrast subject you’d expect to cause a miss, and other times when it wouldn’t lock on to something with high contrast. In that respect, it reminded me a lot of the Fuji X100 running first-generation firmware – and what made me dump that camera was the overall lack of responsiveness.

A6557627bw copy
Steamed tree. H4D-40, 80/2.8

So, the idea of using the H4D-40 as an available-light documentary camera – the majority of my personal work – was already severely handicapped from the beginning. Add to that a maximum lens speed of f2.2 (with the HC 2.2/100) and sensitivity limit of ISO 1600, and you’re going to run into trouble early with handholding and camera shake. Fortunately, it was possible to program mirror lockup onto one of the function buttons; a quick one-two to prefire the mirror then release the leaf shutter with the main shutter release solved the shake problem. In fact, the leaf shutter is so low-vibration that I found myself able to get critically sharp images handheld down to 1/50s with the standard 80mm lens.* It’s a shame that a similar system isn’t used in DSLRs; I suppose it’s understandable when the speed limits for a leaf shutter are both relatively low, and made lower as the distance it has to traverse increases – the leaf shutter is probably going to be at a huge disadvantage compared to a focal plane shutter for very fast lenses like an 85/1.4.

A6557613 copy
Building canyon shadows. H4D-40, 80/2.8

*If you fire the mirror at the same time, 1/2x focal length is about the bare minimum required for a 50% hit rate for pixel-level sharpness – similar to the D800E. I am now of the opinion that the minimum shutter speed required to handhold and achieve critical sharpness at 100% magnification is now dependent solely on the pixel density per angle of view, and independent of format and sensor size. The only thing that changes this equation is the presence or absence of image stabilisation systems. It appears that for up to about 20MP or so, 1/focal length is acceptable most of the time; anything above this requires 1/2x or even 1/3x. With non-stabilized lenses, the D800E, S2 and H4D-40 were all happiest at 1/3x and above; 1/2x still yielded visible shake on a fairly regular basis.

A6557704bw copy
Building reflections. H4D-40, 80/2.8

From a shooting experience point of view, it felt like an early-generation digital: nail the exposure and focus, and the results were surprising; you had to be patient, above all things, and finally, if you got it wrong, no amount of Photoshop was going to save the day; processing latitude – especially shadow recovery – was very low. It was slow, clunky, bulky, and a bit on the kludgy side with the UI (though I understand that many aspects have been improved substantially with the H5D series). That said, the H4D-40 had a number of big advantages in its favour.

A6557630 copy
The cloud. H4D-40, 80/2.8

Firstly, the viewfinder was absolutely fantastic. It’s bright and sharp, the focusing screen is snappy – making it easy to determine which bits of your frame are in focus – and truly enormous. It makes going back to my D800’s finder like the difference between looking out of a cardboard tube and a panoramic floor-to-ceiling window. Secondly, the image quality was bipolar in a good way: images that worked required almost zero work in Photoshop; a quick color adjustment, some very minor curves, and that was it. Native dynamic range was good (though similar to the D800E), with few scenes exhibiting blown highlights. The files are best described as natural; they render with a tonal map that I feel replicates the actual scene very closely, with punch in the midtones and beautiful rolloff in the shadows and highlights.

A6557596bw copy
Nadiah. H4D-40, 80/2.8 at ISO 800

On the high ISO front, compromised dynamic range and quite serious luminance noise started at ISO 800; I’d say at this point the H4D-40 was about a stop behind the D800E. By ISO 1600, two stops. Yet the quality of the noise was remarkably pleasant and almost completely monochromatic; the tight grain reminded me very much of a good roll of black and white film. I wouldn’t hesitate to use the camera at all sensitivities, which is probably a good thing seeing as you’d definitely have to for handheld work.

A6557638 copy
Urban camo. H4D-40, 80/2.8

Interestingly, the native tonal map means that it seems as though the dynamic range is already optimally allocated; the tonal response of the D800E’s sensor means that details are blocked up in the shadows and some recovery must be done in order to achieve pleasing tonality; similarly, there is more recoverable highlight headroom than with the H4D-40, but in the first place, you do have to recover it. If they were trying to replicate the tonal response of a good negative film, they managed admirably. Needless to say, there is no question over resolution. No medium format backs or cameras have anti-aliasing filters; as a result, the amount of detail captured is spectacular. Subjectively, I’d say the D800E and S2 are very, very close in terms of resolving power; the H4D-40’s files seem to be a slight notch above both, with a noticeable increase in acuity at the pixel level. Dynamic range on all three cameras is very similar, but the two CCD-based medium format cameras produce much more pleasing native tonality. Both of the medium format systems offer consistently better optics across the system than what’s natively available for the D800E, with the exception of the 2.2/100 for the H-system – it’s a portrait lens, and it shows.

A6557642 copy
Territoriality. H4D-40, 80/2.8

Most importantly though, the H4D-40 did make me shoot differently – slower, with an increased awareness of the quality of light; however, I was mostly drawn towards static subjects; I attempted street photography with it once and gave up; not only was it enormously conspicuous, but it simply wouldn’t focus fast enough – and I lacked the skills to do it manually. The trouble was, the kinds of things I was shooting I could do just fine with any of the cameras I currently owned – even the RX100. And at typical viewing sizes, there would not be any big differences, either. If I was that conscious of light, or constructing my own, then I could easily photograph scenes that fell within the native tonal range of the cameras I did have to hand. Perhaps the bigger problem was that I didn’t really like the kind of images I was producing; they were too static and lifeless; I didn’t really see an evolution path in there, either.

A6557640 copy
Wondering what happened to the owner of the bike? H4D-40, 80/2.8

Yet in a way, I found the whole experience far too similar to the way I shoot now with the D800E – operation of the camera was very DSLR-like, and I could do exactly the same thing, with almost identical image quality, without having to spend a single cent. Perhaps it might be different if I were to shoot landscapes more often, which would be better suited to a slow, meditative style; the trouble is, there aren’t exactly many such opportunities where I live, and the quality of light in the tropics doesn’t really lend itself to it, either. Unfortunately, after a week with the H4D-40, the only conclusion I can come to is that at this point, medium format is not for me: there is no commercial justification**, it doesn’t advance my personal work, or suit the kinds of subjects I typically encounter. The one thing that did stick in my mind, however, is how little work the files required – it makes me wonder why on earth can’t we get the same tonal quality out of our DSLRs? It certainly isn’t a dynamic range or bit depth limitation; the 16-bit native color on the H4D-40 does render slightly more pleasing colors than the 14-bit D800E, but you had to look for it. I suppose it should theoretically be possible to engineer a default tone curve or series of curves that would at least get the tonal output of the D800E to a similar baseline…food for thought. MT

**A full set of lenses to cover all of my requirements would be prohibitively expensive even after selling all of my current equipment; there would be no wideangle solution to match the Zeiss ZF.2 2.8/21 Distagon, and the maximum macro magnification would be nowhere near what I can easily achieve now with a couple of extension tubes. On top of that, I’d have to use a more powerful set of lights to deal with the smaller apertures I’d have to use to achieve equivalent depth of field on the smaller format, and I’d lose TTL flash metering and remote control. None of my clients need the extra image quality, and certainly wouldn’t pay extra for it.

The Hasselblad H4D-40 is available here from B&H


Visit our Teaching Store to up your photographic game – including Photoshop Workflow DVDs and customized Email School of Photography; or go mobile with the Photography Compendium for iPad. You can also get your gear from B&H and Amazon. Prices are the same as normal, however a small portion of your purchase value is referred back to me. Thanks!

Don’t forget to like us on Facebook and join the reader Flickr group!


Images and content copyright Ming Thein | mingthein.com 2012 onwards. All rights reserved