I recently was commissioned to produce a status update of sorts and small vignettes of documentary covering work in progress for the construction of the HK-Guangzhou-Shenzhen high speed rail link. The vast majority of the Hong Kong portion of this lies underground, which makes sense given both the lack of space and need to have a terminus somewhere centrally downtown. Fortunately, Hong Kong’s underlying geology is very friendly to tunnelling – I’ve always had the impression a good chunk of the island and Kowloon peninsula is really hollowed out given the number of subways, tunnels, malls and utilities hidden underground.
In this series, we see the work somewhere in the final third: the tunnels are complete, the walls are lined and most of the large-scale construction is complete. Supporting slabs underpinning the tracks are in the process of being laid. The tracks themselves have not yet been completed, and full electrical cabling and passenger-facing facilities and amenities are not yet installed. It’s actually a good time to do project coverage from a client standpoint as there has been enough work done to give a sense of the scope of the project, with long ‘gallery’ views and men visibly working to give a sense of scale – but not so much that it looks complete. I had also visited the various sites of the project much earlier on – but there was really not a lot to see on a larger scale as when tunnelling is in progress, the machine occupies most of the tunnel and it’s impossible to photographically capture a sense of scale since most of the cavity itself is filled in, and of course there are no men near the working face for safety reasons.
I also find that images tend to be a bit disorientating because the walls and floor of the excavation can often be covered in irregular mud, rock, or soil – all of which are fractal-type objects and do not really offer cues to physical size. On top of that, lighting is either harsh and very directional/dark, or uniform – depending on the kind of work taking place and the requirements of the workers – which further makes it difficult to fully appreciate other than at very large scale output.
These kinds of assignments are amongst the most challenging I do for several reasons. Firstly, they’re physically demanding: none of the mechanised access is ready or installed (lifts, elevators) so everything must be gotten to on foot. Access points are usually either out of the way and/or deep underground, necessitating a lot of walking from point to point even within the same site. All equipment must of course be carried in from the staging point, and you can’t really put anything down because the floors are often quite muddy, or worse, wet concrete. Humidity is extremely high both because the denser, moister air tends to settle lower down, and because of the concrete drying/setting – not only is this process putting more moisture into the ambient atmosphere, but there’s a huge amount of heat being generated, too*. In summer, the conditions are really tough: think 40-45C and 90+% humidity that condenses over everything and anything; walls are dripping wet, and your cameras tend to be, too. Even in winter, temperatures may be a little lower, but humidity remains quite high. And then there are the mosquitoes…
*Historical tidbit: the Hoover Dam had so much concrete in it, hollow steel pipes were used both as reinforcement but also to carry cooling water to accelerate setting.
On top of that, given the documentary nature of some of the work – you want to both capture authenticity and not interrupt the process (some processes, like tunnel boring, can’t be interrupted) – which means you need to be able to react quickly. Even though there are times when you’re working in a slower, more ‘architectural’ style, off a tripod and with camera movements – sometimes interesting things take place nearby and you’ve got to be able to switch between one way of working and the other on very short notice. I find that having an assistant for this kind of thing doesn’t really help because I want all the gear on me for as fast access as possible. This can often amount to as much as 20kg if I’m working with a full setup (two H bodies, five lenses, HTS, spares, sufficiently heavy tripod, Cube, water safety harness and helmet, etc.).
There’s also the general lack of light to contend with, and huge variance in color temperature of ambient sources – light is light and whatever is handy is used by the workers; that could be older sodium vapour stuff, fluorescent installations that are to be a permanent part of the construction, or modern portable battery-powered LEDs. You’re constantly having to do a lot of post-compensation for strange mixes of color temperature to be perceptually acceptable (even if what was captured was accurate), which means that the more dynamic range and color latitude possible, the better. For the documentary situations, a tripod isn’t so useful anyway because you need a sufficiently high shutter speed to freeze action – that means working at ISO 3200 or ISO 6400 a lot of the time. Whilst smaller sensors can still give you clean results, what you start losing very quickly is dynamic range and color differentiation – I think this actually remains the biggest advantage of larger sensors.
Here are two examples of what I’d consider to be exemplary performance under really difficult conditions. Both are presented here as 100% crops, with no noise reduction. The crop of the train man portrait was shot at ISO 1600, under strong fluorescent light; note that most of the information sits in the yellow channel, and he’s behind a layer of glass. For the other worker, I had to push the shadows 1.5 stops because he was backlit by a sodium vapour light (color corrected in post). This was on top of being shot at ISO 6400 – for an effective ~ISO19K rating. Note again surprising accuracy of skin tone and fine detail preservation.
At the time of shooting, I was also debugging a very late version of the X1D firmware; for this shoot, it replaced the one of the H cameras as primary body, and I went a little lighter on the load out – H5D-50c body, tripod, 24mm, 35-90mm, 150mm and 90mm on the X. I found that the H-X pair allowed me to have the H5D set up for tripod work, and the X1D always ready to go for documentary grabs; whilst I could do the same with two H bodies, having one extra one swinging around your neck at all times isn’t really practical (and tends to bash into whatever is in front of you). The X1D stayed out of the way, and provided a lot of the the opportunistic closeups you see here. I could have done the same with two X1Ds and one set up for tripod work with H lenses and the H-X adaptor, but a second camera wasn’t available at the time – not to mention balance with the larger lenses like the 35-90 and tripod mount on the body being a bit too front-heavy. In future, it’d definitely be much easier physically to go with a pair of X-bodies.
I admit that my biggest concern when switching to medium format was actually this kind of work – traditionally the preserve of the pro 35FF DSLR. But having successfully completed several jobs of this type now, I can safely say that those worries were unfounded – and there’s a level of image quality I hadn’t been able to achieve previously. MT
Images used with kind permission of Chun Wo Construction Holdings Limited, Hong Kong. Shot with a Hasselblad H5D-50c, X1D-50c, H24, H35-90, H150II and X90mm lenses, and post processed with Photoshop Workflow III.
Ultraprints from this series are available on request here
More info on Hasselblad cameras and lenses can be found here.
Visit the Teaching Store to up your photographic game – including workshop and Photoshop Workflow videos and the customized Email School of Photography. You can also support the site by purchasing from B&H and Amazon – thanks!
Images and content copyright Ming Thein | mingthein.com 2012 onwards. All rights reserved