Could there be an ideal camera ? 16 April 2019

Many people might think that a silly question   but I think the idea is worth consideration.

Maybe the concept of an “ideal” camera is a bit ambitious but I think we will see a process of convergence towards something like that.

As I see it a camera needs to provide the user with an experience which can only be offered by a purpose designed device, not a device like a smart phone in which photographs are a secondary capability.

The camera will be used by someone for whom taking photos is the primary activity.

Considering this I see the market for cameras moving in certain directions:

1. All cameras will move upmarket. There will be no reason for budget compacts or budget interchangeable lens models to exist when a decent smartphone will do the job quite nicely and be easier to use.

2. DSLRs will disappear as soon as high end mirrorless interchangeable lens models have gained comparable high speed autofocus capability.

3. Mid range interchangeable lens models could struggle for survival as bridge types gain capability making them ‘good enough” for just about any purpose.

4. High end mirrorless interchangeable lens models will survive because professionals and enthusiast amateurs who demand and can afford “the best” will keep this end of the market viable. But prices will be high.

5. Leica has shown that there is an ongoing market for prestige models at a very high price point. Low volumes are offset by the high margin on each item.

My realisation of this is seen in Mockup 6.2 shown here. This is presented as a generic camera which could have interchangeable lenses or a fixed lens.

It is not tiny. It is geared towards other criteria. It has an anatomical handle and thumb support, an eye level electronic viewfinder, a fully articulated monitor screen, a zoom lens and a full set of controls to enable the experienced user to drive it like a sports car.

It is not for snapshooters or photo dilettantes. However beginners can set Auto mode while learning how to make full use of the camera’s extensive capabilities for photo and video or both simultaneously.

Most cameras have archaic image processing capability. I offer just two examples but could go on all day about this.

1. Why are we stuck with JPG and RAW ? 

RAW demands that the user become adept at using complex Raw conversion and image processing programmes in a computer or similar device. That might be all right for someone like me who has been using Photoshop for 20 years but for a newcomer to the world of cameras that is a daunting hill to climb.

On the other hand we have JPG,  a lossy output format created in 1992 which throws away half or more of the data captured by the sensor.

Neither is really suitable for the modern era.

Surely modern image science can do better and develop an output format which is not lossy (and therefore is responsive to post processing if desired) but which contains many in camera adjustments to color, tone and sharpness so most pictures require little or no work after output.

2. As an example of the issue above:  Consider that many modern cameras ask the user to make decisions about things with names like  Auto lighting optimiser, D-Range correction, Auto EV compensation, color space, Image control, Peripheral illumination correction…..and so on. The list in some models is extensive. I have been using digital cameras for almost 20 years and am still not quite sure what some of those options mean.

Surely the image sensor has all the information it needs to automatically balance various aspects of exposure and rendition. It just needs enough processing power and the right data processing capability to make it happen.

But most modern cameras are seriously lacking in processing power. As a result they cannot do things which smartphone users take for granted.

The menus on most current cameras look like an afterthought. Most are a muddled mess of disparate items arranged in no particular order. Many items are redundant or their purpose unclear. New items are added with new models, leaving old ones in place.

I think it is imperative that camera makers go back to a clean sheet on this and re-design menus which are user relevant.

Here is my suggested list of headings and subheadings:

This is indicative rather than prescriptive.

Consider the motor car. In the early days each make and model had quite different control systems. One might operate the brakes by pulling a hand lever, another by pressing a foot pedal.

Eventually most of these design differences underwent a process of convergence to the point today that I can hire a car anywhere in the world and expect the basic controls to be the same.

There could be various reasons for this, one of them being safety but I have no doubt that another reason is basic ergonomics. The control layout of most cars is efficient. It allows the vehicle to be operated with a small number of simple  actions.

Most cameras on the market today are the opposite. Their operation requires intensive study of a badly written 500 page owners manual.  They are festooned with numerous controls laid out in no predictable fashion.

If cars were designed like cameras the death toll would rise to catastrophic levels.

Like this……On this model when you stamp on that pedal the vehicle turns left when certain menu items are selected. Sorry about those pedestrians. It’s operator error you know, look, the manual describes this on page 355 in the footnotes to the “pedal function” heading under the “options” subheading the details of which are clearly outlined on page 476 under the “directional variations” tab.

You think that is a joke ?

That is exactly how most modern cameras work. Fortunately nobody is killed when things go wrong.

How did we get into this ridiculous situation ?

The interesting thing is….it was not ever thus.

For a considerable portion of the 20^th Century the dominant camera type for enthusiast and expert users was the single lens reflex (SLR). Most of them looked very similar and used the same controls.

The market supported many different makers most of which have since departed the scene.  You could pick up any camera from any maker and use it right away without having to read a weighty tome to figure out how to persuade it to work. Just load a roll of film.

But then the camera was hit by a quadruple whammy of disruptive changes.

First came autofocus which evolved in the 1970s and 1980s and is still being improved.  This put a lot of makers out of the game.

Then came digital capture which appeared on consumer cameras from the mid 1990s and became almost universal by the mid noughties.   This produced challenges which several manufacturers failed to meet.

In 2008 mirrorless interchangeable lens cameras came on the market putting the DSLR on notice from that point.

Then came the smartphone to render irrelevant the great majority of cameras most of which were utility compacts.

These changes have left camera makers scrambling for relevance in a declining market. The remaining players appear to be feverishly casting about for some kind of magic formula which they hope will keep them in business.

Their approach to this has mainly been via technological development. In the process the user interface has been neglected.

But the curious thing is that in many ways we are back at the situation which prevailed in the middle of the 20^th Century.

That is:  most cameras from most makers do most of the same things. Their capability is very similar as is their output as digital picture files.

So the main difference between cameras and the main area which desperately needs major improvement is the user interface in all its aspects including the experience of using the device.

The trend with many recent cameras is to throw buttons and dials at the thing like confetti at a wedding, presumably in the hope that somebody will find some of them useful for something.  Check out the just released Panasonic Lumix S1 and S1R models for instance. How on earth are ordinary humans going to remember the assigned functions of all those buttons, dials, levers and other little gizmos scattered all over the front, top, back and sides of the thing ?   The reality is, they won’t.

Cameras need to be getting more simple at the user interface not more complicated.

I recently discovered that my Sony RX10.4 became easier to use after I disabled 8 buttons with user assignable function and congregated those functions onto the single Fn button. Now I don’t have to remember which button carries which function and operation of the D-Pad is no longer affected by the sequence in which I press the cross keys.

Leica is one maker which has had several tries at keeping it simple  but with inconsistent results.

Their leitmotif is das wesentliche  which translates into English as the essentials, the nitty-gritty, the essence, the substance, the vitals….you get the idea.

Unfortunately the Leica boffins appear not to agree among themselves about what is essential. So we get models like the strange TL claiming the same conceptual basis as the old style M (messsucher) series.

Here is my contribution to camera design in the form of mockup 6.2 with a carefully considered shape, layout and set of controls.

Obviously being a wooden mockup this is all about the shape and the external controls not the inner workings, of which it has none.

This design is based on some underlying principles which are

1. Camera use can be usefully  considered in four Phases with Capture having three elements.

Setup

Prepare

Capture:   Holding/Viewing/Operating

Review

2. Each phase requires the user to carry out various Tasks.

Each task requires Actions.

3. The ideal camera can be operated efficiently with the fewest possible actions each of low complexity.  

By way of example some cameras like my Sony RX10.2 ask you to change the lens aperture by turning a ring on the lens barrel after setting the Mode Dial to the correct position. This is a complex action. It requires that the right hand be partly released from holding the handle to turn the Mode Dial then the whole of the left hand is required to turn the aperture ring.

Compare this with the way many Mode Dial + twin dial cameras work. You still have to turn the Mode Dial to the A (Av) position but then the right index finger can change aperture by rotating the front dial. The left hand remains in place supporting the camera. This sequence requires fewer, less complex actions.

4. If the camera has user interface modules in excess of requirements they are just clutter which does not improve usability.  They may actually have a negative effect on usability by interfering with each other as I discovered with the RX10.4.

The corollary to this is that the ideal camera should have enough direct user access points to exercise full control over the device in all phases of use but no more.

5. The next principle derives from my observation (which by the way anybody could make) that when moving from one usage situation to another, for instance general hand held to sport/action or hand held/street to tripod/landscape….you get the idea ….. we want to change groups of settings all at once.

For instance when moving from general hand held work to sport/action we need to change The 

Shooting Mode from Av to Tv, frame rate from single to burst, autofocus from single to continuous,  minimum shutter speed re-set to a much higher level, and so forth. Lots of things all at once.

One tested way to do this is to pre-load those groups of settings onto  Custom (User) Modes on the Mode Dial.

For this to work optimally the camera needs to be devoid of Prescribed Purpose Modules. Modern cameras often have these in the form of  a Drive Mode Dial, Focus Mode Dial and similar.

These are often touted by reviewers as a good thing but I have come to the view that they work against streamlined shifting from one usage situation to the next.

How?  Because in addition to moving the Mode Dial to one of the C(U) positions you have to move the prescribed purpose module(s) to a new setting. This means more actions each more complex.

6. Controls need to be grouped on the available camera real estate in accordance with

a) their place in the operating hierarchy, with controls for Capture Phase/Operation being on A-Level locations, controls for Prepare Phase on B-Level locations and controls for Setup and Review Phases on C-Level locations,  and

b) which finger is assigned to use the control.

You can see how this works in the picture of the mockup with captions.

7. The only controls on the lens barrel are fully circumferential. This is the only control type which works equally well with left hand under/left hand over, landscape or portrait orientation.  Zoom and focus would be the obvious candidates to locate on a lens barrel but as functions on modern cameras are mediated electronically any control could be located anywhere on or off the camera.