F Stops

Started Aug 11, 2012 | Discussions thread
Shop cameras & lenses ▾
dosdan Regular Member • Posts: 488
Re: Um, no.

First off, as I understand it, each manufacturer is free to choose a metering point for each scene for what they consider to be a "correct" exposure. Since scenes are not normally completely uniform in either illumination or reflectivity, AE can be set up to allow a certain percentage of blown sensels and still be considered a "correct" exposure.

Also metering can vary in offset for another reason. Manufacturers can choose a different "K" (reflected-light meter constant). So for the same scene, one manufacturer can have a "hotter" exposure than others. For example, Pentax uses K=14. which is 0.17 stops hotter than Canon or Nikon's K=12.5 and 0.3 stops hotter than APEX's original K=11.4

This is explained in the link to the APEX article:


Accurate absolute light level determination is not possible because you're not sure of exactly what the camera you have in your hands is calibrated to (there might be minor variations during manufacturing). And you have the issue of illumination colour and spectra (if it's fluoro lighting).

But it may well be possible to use the camera to accurately determine relative light levels e.g. the relative change in exposure after a "9-stops" ND filter is attached to the camera.

Then there is different forms of exposure metering. Matrix/Evaluative metering weighting is propriety and methods used to pick which weighting will be used for which distribution of scene luminance are not published by any manufacturers. So matrix meter is inapplicable for what you're trying to do.

That leaves Centre-Weighted and Spot metering. If may be difficult to get a surface uniform in both illumination and reflectivity so that the minor jiggling of the camera on the tripod as you add or remove an ND filter doesn't vary the metering by more than 1/6- or 1/3-stop. So I'd suggest first trying pointing the camera at a white wall and jiggling it slightly in Av mode to see if shutter speed varies much. (If the camera allows, set it to the smallest shutter speed step size.)

If the shutter speed jumps around with that test, repeat using Centre-Weighted.

Once you know the metering is stable enough, it comes down to the metering linearity (we're dealing with log units - "stops", but you know what I mean). 9=stops = 2^9 = 512. So the metering nears to be linear to within say 1/6-stop for a light level variation range of 1 to 512, from the starting point of your initial illumination. Let's assume it is.

Now you need to be in Av mode. You will be varying the light of a range of 1 to 512, so the shutter speed should be set in Av to allow this, e.g.
1/1000s without ND filter
0.5s with ND filter

The LV calculator link in my first msg has drop-down list of 1/2- and 1/3-stop steps of ISO, shutter speed and f-stops. If, with the ND filter, shutter speed drops from 1/1000s to 0.5s, the filter is closer to ND512. But if it drops to 0.4s, it's closer to ND400, and 0.3s, it's closer to ND320 (ND320/400/512 are approx. 1/3-stop steps).

But the linearity the changing of the shutter speed may make even this uncertain.


 dosdan's gear list:dosdan's gear list
Pentax K100D Super Pentax K20D Pentax K-5 Pentax K-01 Pentax K-3 +1 more
Post (hide subjects) Posted by
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark MMy threads
Color scheme? Blue / Yellow