Talk:Sensitometry

From Wikipedia, the free encyclopedia
Jump to: navigation, search
WikiProject Film (Rated Start-class)
WikiProject icon This article is within the scope of WikiProject Film. If you would like to participate, please visit the project page, where you can join the discussion and see lists of open tasks and regional and topical task forces. To use this banner, please refer to the documentation. To improve this article, please refer to the guidelines.
Start-Class article Start  This article has been rated as Start-Class on the project's quality scale.
Taskforce icon
This article is supported by the Filmmaking task force.
 

Confusion over HD[edit]

Judging by the terminology and edit comment, 68.51.193.244 (talk) seems to have confused "H-D curves" with the sort used in high definition video. Too bad Hurter and Driffield did not transpose the order of their names to prevent confusion. --Adoniscik(t, c) 22:47, 1 April 2008 (UTC)

Interpreting characteristic curves[edit]

One common problem for photographers interpreting characterstic curves - whether for film or video - is converting the log10(H) illuminance scale to log2 F-stops or exposure values. Luckily this is quite easy because 1EV ≈ log2(H) ≈ 0.301 * log10(H) and so 1*log10(H) ≈ 3.32 EV. --Redbobblehat (talk)

Responsivity vs Gamma[edit]

(Ray 2000 p.424-5) uses a "characteristic curve" (Fig.25.13) to depict video Responsivity, and defines "Responsivity, R : This is the slope of the straight line section of the output–input characteristic curve. Possible units are volts/joules cm–2, volts/quanta μm–2. etc." which sounds very much like good old Gamma[1] ... is there a difference ? --Redbobblehat (talk) 02:42, 15 February 2009 (UTC)

Yes, it's completely unrelated. Responsivity is the slope of a linear response, and has units, output units per input unit. Gamma is the slope of a log-log curve, characterizing an approximate power-law nonlinearity, and is dimensionless. Dicklyon (talk) 03:40, 15 February 2009 (UTC)

Different types of sensitometric curves ?[edit]

Is it incorrect or unwise to say that:

  1. a characteristic curve describes a film's "responsivity" to exposure ?
  2. a Hurter-Driffield curve describes a CCD's "responsivity" to exposure ?
  3. a Hurter-Driffield curve is a "responsivity curve" ?
  4. a Hurter-Driffield curve is a "gamma curve" ?
  5. when gamma is given as a number, it approximates the slope of the "gamma curve" (ie toe + gamma-curve + shoulder = HD curve) ?
  6. "gamma curve" is a misnomer, because gamma is by definition is the (acceptably) straight section of an HD curve ?
  7. you can apply gamma-correction to the toe or shoulder in order to retrieve gamma-compressed detail.
  8. "Crushed blacks" are literally the gamma-compressed tones in the toe of the curve, and "blown-highlights" are the gamma-compressed tones in the shoulder of the curve. These terms may be properly applied to film but not video. Video, having no toe or shoulder, exhibits "black clip" and "white clip" instead.
It's probably not a good idea to mix up a linear sensor response (responsivity) with a nonlinear film response (HD curve). Where did you find the term "gamma curve"? I don't see it around here, but it is possible that it's a misnomer in some contexts, as in describing an HD curve. Applying "gamma correction" to linearize a response in a toe or shoulder reason would be called linearization, in some contexts at least, but it's typically not done. I think your use of "gamma compressed" is a misnomer for the toe and shoulder compression. Where do you get the term "crushed blacks"? And what's with the extra hyphen in blown highlights? And video does have toe and shoulder, typically, added by the image processing. Dicklyon (talk) 03:47, 15 February 2009 (UTC)

Sensitivity[edit]

  • In electronics[2], Responsivity is distinguished from Sensitivity; the latter being reserved for the minimum level of stimulus to produce a valid response. The same distinction is made in photography; where the sensitivity of a given light-sensitive device (eg usage "a quantitative measure of the sensitivity of a metering system by quoting the minimum EV that can be measured for a film speed of ISO 100" - Ray 2000 p.318) represents the minimum useful exposure (Ray 2000 p.316).

*Film and video sensitivity are measured differently; and expressed as "film speed" and "SNR" respectively ... but in practice they both represent the minimum useful exposure, which is also synonymous with "exposure index", "exposure sensitivity" and "speed point".

  • The minimum useful exposure is measured differently for film and video; and commonly expressed as "Speed rating" or "Sensitivity" respectively.
  • The Arithmetic "ISO speed" is the (most widely recognised -ref?) index of photographic sensitivity. It is defined (Ray 2000 p.305 + a little reverse engineering) by a minimum useful exposure of Hm = 0.8/S, where S is the ISO speed; so for ISO 100: 0.8/100 = 0.008 lux seconds. Because it is so well known, alternative "exposure indices" are frequently compared to it, and the "ISO equivalent" (Ray 2000 p.121) has become a convenient measure of sensitivity. A "200 ISO equivalent" is claiming a minimum useful exposure of 0.004 lux seconds.
  • "Speed point" is used for special (high-contrast, etc) films (Ray 2000 p.306). I'd be happy for this to go into the film speed article because it's for special films which can't be adequately represented by an proper ISO speed rating.

Is this correct ? --Redbobblehat (talk) 04:46, 16 February 2009 (UTC)(updated)

Sort of. But SNR is not a sensitivity measure; sensitivity can be quantified for example as the minimum exposure that will yield a certain threshold SNR; that's how digitial camera ISO speed (Film_speed#The_ISO_12232:2006_standard) is defined (by the noise method, that is); there's also the method they call Standard Output Specification, which is not actually a sensitivity at all, but a responsivity based definition. They also have an "exposure index" definition, which is even more arbitrary. I've never heard of "exposure sensitivity"; "speed point" is a step in the process of measuring ISO based on image quality. These terms each have their own meanings, but they're related in various ways. Dicklyon (talk) 03:58, 15 February 2009 (UTC)
Thanks for the pointers :). "SNR" was shorthand, sorry ;) Does ISO 12232:2006 make a distinction between "digital still" and "video" sensitivity? Specs I'm thinking of always use "Sensitivity" followed by a-pack-of-lies dB ;). "Exposure sensitivity" is just a belt & braces tautology. I've tried to include your other points in the re-draft (above) but my brain is melting. --Redbobblehat (talk) 04:46, 16 February 2009 (UTC)
I'm pretty sure that standard is only for still cameras. The dB thing is nonsense of one sort or another, usually. I've still never heard of "exposure sensitivity". Dicklyon (talk) 05:37, 16 February 2009 (UTC)

ways of writing H–D[edit]

Do we really need to include all of "Hurter–Driffield curves, H–D curves, HD curves, H & D curves"? Aren’t at least the last three pretty much the same thing? –jacobolus (t) 03:47, 14 March 2011 (UTC)

I thought it would be useful to have sourced examples of the various things they're called. If you'd like to figure of which ones are too rare to bother with, go for it. Dicklyon (talk) 03:51, 14 March 2011 (UTC)