Crystal ball gazing – have we reached a plateau?

The split, redux: all will be explained in the article…

In the late 90s/early 2000s, film photography arguably reached its zenith in many ways: you could get all sorts of hardware in all sorts of form factors; emulsion technology peaked in both proliferation and quality, and it was easy to get anything developed and printed, and developed well. There were high end pro compacts, super fast DSLRs, consumer megazooms, large format folders, sub-frame cameras…films varying in speed, look, positive/negative, and even crossover-types like C41 process black and white. I’d even argue that since then, film emulsions have not really improved (undoubtedly due to the vanishingly poor business proposition created by the emergence of digital) – and we’ve lost most of the major manufacturers and choices. (To say nothing of the labs.) The core technology reached a balanced plateau: lenses were matching emulsions in resolving power; AF systems were matching the rest of the system in precision required to consistently deliver the aforementioned resolution. On film, there’s not much difference between one of the better 50mms of the time (say a C/Y 1.7/50 MM, A Leica 50/2 Summicron, or a ZF.2 2/50 Makro-Planar) and arguably the best of today – the Zeiss 1.4/55 Otus APO-Distagon. I tried this experiment on an F6 some time back, with Fuji Acros: I couldn’t really see much of a difference in resolving power. Drawing style, yes, but not resolving power. Your ability to focus made far more of a difference. And running the same film through my 1979 F2 Titan or the 2005 F6 made no difference at all, of course. Ultimately, during the film era: image quality was proportional to format size. How is this relevant to now?

Digital has played an out of sync catchup game between formats over the last decade and a half. Smaller sensors improved faster because the economics worked in their favour: smaller = much cheaper = volume = R&D budget to improve (plus competitive factors). But as sensor costs fell, we saw this slowly move up the chain: ever larger sensors fell below lower absolute dollar amounts; to say nothing of inflation. First APS-C, then full frame. And we’re now on the verge of seeing the same thing happen with the smaller medium format (44x33mm). What’s happening, of course, is the result of two things: pixel-level architecture scalability, and improvements in quality control on both the production process and the silicon blanks* have lead to improving yields and falling sensor costs – or at least convergence to a rough cost per square area model. Sound familiar? It’s not unlike film.

*Impurities in the silicon wafers lead to errors which lead to rejections; a fixed impurity rate is fine for small chips because at 100 chips/wafer and 5 impurities/wafer, at most, you reject five chips. But with the same impurity rate and 4 chips/wafer, you may land up rejecting all of them – or none; the impurities could fall outside the lithographic area. With larger sensors, it isn’t the silicon that’s expensive: it’s throwing away the duds. Up til not that long ago, reject rates on very large sensors might be over 50%.

There are a lot of complex interactions in the whole imaging chain that may affect performance: the higher the pixel pitch, the steeper diminishing returns get as you start to have lens-quality effects, alignment, focusing step limits, dynamic range etc. It’s quite possible that if the assembly tolerances aren’t high enough and cause an image that’s just slightly defocused beyond the circle of confusion for that pixel pitch, a significantly lower resolution (remember: true resolving power goes by linear pixel count, not area) sensor may well produce visually better results, and not much difference quantitatively, either. This limit is significant because the biggest factor in most cases is the user – not the hardware. A little sloppiness in technique or not calibrating lenses/ checking focus can throw away all of the difference between 24 and 50MP you paid a lot of money for.

Oddly enough, I believe on the technical side, human physiology is the cause of the limits. Firstly, no matter how good your stabilisation system, nobody can be perfectly still; this means move to a tripod or be prepared to have astronomically high shutter speeds. Having shot tens of thousands of frames with 100MP for the last eighteen months, I can tell you that deploying the difference handheld and consistently is not trivial. My preference strongly swings towards working off a tripod in this case. Secondly, and more importantly, is the practical output limit: whilst we can have enormous displays and prints and enormously high density displays and (Ultra)prints, perfect eyesight at a best-case 20cm viewing distance tops out around 1000px/in, or 500lp/in. This applies in the real world, too, and at distance. No matter what you’re looking at: your eyes cannot resolve more than this. At that distance, I can’t see more than a 13″ monitor, even at the edges of my peripheral vision (which resolve much, much lower). 13×8″ is about 104MP. Even given the benefit of the doubt of bayer limits vs true pixel color, oversampling etc. I don’t see us being physically able to absorb more than 200MP of information at one go in a still. Area by area viewing is of course another thing entirely, but we’re back to the practical limitations and implementation of big printing etc.

it’s not just the technical side that’s limited by us: it’s also the creative and consumptive side. The camera only makes what we tell it to; this much has always been clear. But I don’t think most people’s images are improved by the added technical challenge of maximising output of their equipment when they haven’t maximised their visual vocabulary yet – it’s just one more distraction. Analogy: learn the language to be able to clearly express your idea before you start shouting at the top of your voice. Yes, more technology opens up greater shooting envelopes – but I’d argue we’re not far off the useful limits here, too – when I can shoot moving people in a very dark cave, handheld, and produce image quality like this – I’m not sure there’s that many commonly-encountered situations that demand more.

In my mind, the biggest challenges and changes are going to come on the output/ content consumption side. There’s simply so much content these days that we are overloaded with visuals; forget standing out, just being seen is already difficult enough. And with the usual social media channels moving to a very advertising-heavy presentation, people get both increasingly dismissive and simply turned off looking. Instagram was interesting for a year or two when there were a lot of good photographers posting content not filtered to blindness-inducing color and minimal to no unexpected visuals; now I’d say one in five or so posts I see are ads for things that are not interesting, irrelevant to me, and worse, often visually ugly – think random people promoting their horrible blurry selfies or cat photos. I can say my interest in this platform has declined significantly as a result, as with others I’ve spoken to. But given the monopoly they’ve had on visuals – what’s left? The smaller sites never reached critical user or financial mass, many have closed down or remained largely in stasis as they’ve changed hands (Flickr) and others are simply not very good for serious image sharing (Facebook, Snapchat, Tinder, Flubber, Blobme, Globulitis etc).

The net impact across the photographic industry has been one of contraction across the board – hardware progress represents such diminishingly small returns for most users that the volume tends to be at the consumer/prosumer end trading up slightly, rather than the mad rush to the top of the latest and greatest we’ve seen in the last five to ten years. The formerly very serious hobbyists have either dropped out now that gear doesn’t provide much visible improvement and actual effort is required, or they’ve gotten very very serious and really dug in – I’ve seen a very clear divergence in the readers of this site (and also through speaking to other publishers). In a strange way, despite the relative stasis (or hugely diminishing returns for most) on the hardware side, and the contraction in audience of the serious output side – the situation is actually beneficial for the serious photographers, professional and amateur alike. We (and our clients) are no longer being distracted by dabblers, the lure of the next model, or chasing social media approval; it’s time to get back to making pictures. Until the next thing… MT

__________________

Visit the Teaching Store to up your photographic game – including workshop videos, and the individual Email School of Photography. You can also support the site by purchasing from B&H and Amazon – thanks!

We are also on Facebook and there is a curated reader Flickr pool.