Why use Pentax?

Started Mar 18, 2013 | Questions thread
bob5050 Regular Member • Posts: 234
Re: Why use Pentax?

Gerry Winterbourne wrote:

I'm not saying that such tests are useless, but they need to be combined with other things to get a full picture.

I feel a bit impertinent to be talking in the presence of my betters (I never fail to learn from your posts, Gerry), but actually I'd want to extend that point a bit.

Numerical measurements can obviously be totally useless. Under what circumstances is that the case?

1. The measurement is irrelevant because indistinguishable in actual use. The situation in modern digital photography reminds me quite a bit of the situation with audio gear in the late '70s: different speaker and amp manufacturers would publish frequency-response measurements up to 50kHz, when in fact 'normal' human hearing generally ranges from 16-20kHz, and vary rarely exceeds 30-35kHz (young adults with great hearing, generally higher with women than men). Measurements of flat, unbiased reproduction are the standard and basis for comparisons, but in practice, above 30kHz susch measurements are highly suspect in importance, and above 40kHz they're worthless if you're buying the system for a human audience to experience.

2. To matter, the aspect of the system under test must be related to the dominant system constraint (see Theory of Constraints ). Otherwise the measurement is, at best, unactionable and irrelevant, and at worst actively misleading. The smart phone ads in this week's BestBuy flyer are all touting 8MP cameras. Does anyone here believe that sensor size--rather than, say, lens size, build quality and precision, or lack of focal control--is really the active constraint on the photographic capabilities of these devices? To use the audio analogy again, the ability of one set of speakers to produce less distortion than another at 100w/channel doesn't really matter if your amp only provides 20w. To improve a system, you have to measure and improve the weakest link.

3. The attempt to isolate the constraining factor for measurement produces measurement conditions totally at odds with real-world conditions. For example, comparing the high ISO noise directly between two P&S cameras is totally artificial if one comes with an f/1.4 lens and the other with an f/3.5. In the real world, under the same shooting conditions, the cameras are simply never going to be at the same ISO. So the direct comparison does not illuminate reality, it actively misrepresents it.

4. The test doesn't actually measure what you think it does. This can actually be a very subtle question, the most famous examples of which are IQ tests. But in photography, high ISO noise is again a great example. What really matters, I'd argue, is low light performance. A high ISO noise measurement doesn't actually test that any more than an IQ test objectively tests 'native intelligence.' It tests one isolated aspect of a much more complex issue, and not necessarily the constraining one for any particular system.

In evaluative summation, then, test results always have to be themselves tested against criteria of relevance and reliability, since

- because something can be easily and objectively measured doesn't mean it matters.

- because something has no generally available objective measurement doesn't mean it doesn't matter

- single dimension tests of interacting multi-dimensional systems are especially suspect.


 bob5050's gear list:bob5050's gear list
Pentax K-3 Pentax smc DA 18-135mm F3.5-5.6ED AL [IF] DC WR Sigma 17-50mm F2.8 EX DC OS HSM Pentax smc DA 18-55mm F3.5-5.6 AL WR Pentax smc DA 35mm F2.4 AL +3 more
Post (hide subjects) Posted by
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark MMy threads
Color scheme? Blue / Yellow