Close, but no cigar: how to design mirrorless right

_7R2_DSC3092 copy_8058003 copy_5022857 copy
Too large/expensive; too slow and unresponsive, power hungry; no finder or IS

_1000207 copy_4D08961 copy_G004935 copy
Limited sensor resolution; overambitious image quality and fragile feel; too many steps to get shooting

_8058455 copy_2M00250 copyIMG_6800b copy
Fixed lens; great UI with terrible ergonomics; classical controls don’t work for digital, sensor limits

IMG_6957b copy_8B25854 copy_8038907 copy
Ergonomic and workflow challenges; IQ limitations from sensor size; needed two years to fix FW

And this is barely half of the mirrorless cameras I’ve used and reviewed on this site in the last couple of years. I still have not found a complete replacement for the DSLR, and I suspect there are many other photographers in the same situation. It isn’t for want of trying or stubbornness; it’s because the product simply does not exist. We’re not asking for the unicorn here, either: there are ergonomic/UI/UX/engineering solutions that have already been implemented and received well in other cameras – just not in the same one. And to clarify (since judging by email and comments, many are missing the point): this post is not to complain mirrorless isn’t a DSLR. It’s recognising that mirrorless is the future for so many reasons – but we are still suffering from stupid design that has already been solved. All of these problems beg the question: just how difficult is it to get it right?

Important: Read this first.

[Read more…]

In search of the unicorn

_8B03666 copy
Nope, that’s not it.

The ideal [insert your obsession of choice here] doesn’t exist.

We all like to think ‘if only…’ and it might. Whether it’s cameras, clients, light or partners, there’s always something that could be better. Perhaps this is a reflection of the consumerist and entitled nature of modern society as a whole, or perhaps it merely shows that we as people are always changing. Ironically, it is this very ‘if only’ that keeps things interesting: if you were to make the ideal image (in your own mind, and subject to the constraints of personal bias) of whatever you framed whenever you pressed the shutter, you’d quickly run out of possible subjects. It is not a bad thing at all that a) everybody has different opinions and b) we ourselves are in a state of constant flux. I know for certain that I approach familiar subjects like family or watches very differently now than from when I did previously. But there is perhaps such a thing as ‘good enough’ – better than 80/20, certainly – and we should probably know when to appreciate it. Today’s post is going to be looking at the business side of photography.

[Read more…]

The ideal camera

I’m going to disappoint you upfront: it doesn’t exist, at least not yet. But what if we did a hypothetical – yet realistic – exercise to see what a product like this would look like?

First disclaimer: I, and a group of partners, were very close to building this – but we were let down partway through by a key technology partner, and found funding challenging – despite the continuing growth of the photographic industry, especially at the advanced amateur and entry-professional segments. Alas, it never materialized. In any case, the idea has matured a bit since then, both through the benefit of time, and ideas gleaned from intervening new product releases.

Second disclaimer: This is what I would like as a photographer, and may not necessarily apply to everybody.

I’m not going to suggest a 100-megapixel, 25fps, point-and-shoot sized body with a full frame sensor, and built in 10-600/1.4 zoom. That’s stupid, and the preserve of equipment masturbators on camera forums who don’t actually use their equipment.

Right now, I use four systems: full frame Leica M digital, for photojournalism, reportage, portraiture and travel work; full frame Nikon FX for extreme low light, macro and controlled lighting work; Micro Four Thirds for personal work/ carry where photography isn’t the primary objective of the outing, and where anything larger might be obtrusive; and finally, a number of compacts for social documentary use. I’ve never been told by a client that I have insufficient resolution; and I’ve delivered images as small as 5MP before. Why? Because I’m a stickler for practicing shot discipline; that means even though I’m ‘only’ giving you 5MP, they’re 5 extremely clean MP at the pixel level, well exposed, perfectly focused, not posterized, and will print cleanly at A2. More MP – say 18 from the M9 – just means you can make larger prints before fine detail breaks down at close inspection distances. That’s all. I’ve only come up against the low light limits of the Nikon system once – and that was photographing a wedding in near darkness (only candles) in rural Nepal; but I was still able to get good quality images with ISO 12,800 and f1.4 glass. I don’t need ISO 512,000.

Let’s take the best parts of each system, and see what our bastard-camera would look like:

1. Sensor size: This is a tough one, actually. There are times when I want a smaller sensor for more DOF (think very compressed perspectives) – I will often use a compact for this. There are times I want a larger sensor for even shallower DOF for a given field of view – this is the main attraction of systems like the Leica S2. Let’s go with full frame, because we can always have a 2x crop factor (it’s easier to multiply out than 1.5x, and when you want a smaller sensor it’s generally because you need more reach or more DOF and more reach.)

2. Sensor spec: 48MP. Whoa, really? What happened to I don’t need huge MP, you’re wondering? Simple: pixel binning. 48 million photo sites – with RGB + Luminance values coming from individual photo sites, would mean amazing detail, fantastically accurate color, and low noise. Output file resolution of 12MP – but actual pixel detail of 12MP. With this technique and today’s sensor architecture, I don’t think it’d be a problem to get a clean ISO 12,800 file. And that’s enough for me. Add a Bayer algorithm mode to give a 12MP crop, and that takes care of my telephoto/ compressed perspective/ high DOF needs.

3. Speed: Doesn’t have to be fast – 5fps is plenty. I work just fine with the single-shot M9-P; most of the time bursts are for stability rather than action.

4. Video modes: I don’t use it, but I do like to experiment – so perhaps a 1080P/ variable frame rate mode would be nice, with full manual controls, of course. But no rolling shutter!

5. Autofocus: Let’s take the best there is now and match it: CAM3500FX in the Nikon D700 is eerily psychic at tracking moving subjects; however, it could use more frame coverage – expand the points to cover out to say 80% of the frame linearly, and we’re in business.

6. Firmware features: Lots of scope for customization: I don’t mind going through a lengthy setup process to ensure the camera does exactly what I want. I want to be able to hide things in menus and reassign all of my buttons, and have custom modes that set all camera behavior – including base exposure parameters – at once. And of course the ability to remember, save and transfer to another camera body.

7. Lenses: AF, fast aperture – at least f2, preferably f1.4. Useable wide open, of course. Physical aperture rings. I don’t need zooms, or extreme focal lengths; 18-24-28-35-50-85-200 and a short tele macro (maybe the 200 could double as a macro) will do me just fine. It could even be reduced to 24-50-85 in a pinch, so long as one of those was a 1:1 macro with capability to take a bellows or extension tubes.

8. Body: Robust, thick-gauge metal (tactility is important) and weather sealed; plenty of sensible physical controls; an optical viewfinder or excellent EVF – I think the new Sony OLED unit in the NEX-7 or 5N is about the minimum. But I much prefer the enormous, bright finder of say the Leica S2. I think M9-size is about perfect; big enough to have some heft and solidity, sufficient space to house all of those controls, and (yes, I’m shallow – or at least some clients are) impressively professional-looking. That would have to have an EVF to maintain the body thinness and accurate framing. If it’s an optical finder, we could probably make it small-SLR size – I don’t see why a full frame sensor won’t fit inside a D7000-sized body; the flange distance is FX-sized, and there were film SLRs of similar dimensions. Ergonomic comfort is very, very important – especially if you’re going to use it for 12 hours a day (or sometimes even more). Plenty of sticky rubber and thumb/ finger hooks, please. Buttons big enough to use with gloves, and wheels/ switches that don’t move unless you want them to – think locking buttons or very stiff detents.

9. Other considerations: File format: Compressed 16-bit DNG. It’s future proof, past-proof, and easy to handle and archive. Don’t bother with JPEG because I don’t use it. Live view would be nice, but not critical; it can be useful for awkward angles but is greatly enhanced by a swivel screen (which frankly I hate because they’re flimsy and compromise body integrity). A vertical grip is nice, but not necessary. What is necessary is a high magnification, high-eyepoint (I, and a lot of other photographers, are poorly sighted thanks to plenty of shooting into bright point light sources and so need to wear glasses) and with a nice big rubber eyecup. A high-resolution, color-accurate LCD is important for review. Most important is operating speed – menus and such – the camera should never feel laggy. This is the #1 reason why I’ve given up on a piece of equipment in the past.

Okay, so who’s going to make me one of these things? MT