Re: Step Wedge

From: Etienne Garbaux ^lt;>
Date: 11/10/04-10:20:01 PM Z
Message-id: <p05210600bdb87617660e@[]>

Sandy wrote:

> The problem with generating curves exposing through the camera is
> flare. It varies a lot according to lens, bellows position and
> lighting conditions which would probably result in considerable
> inconsistency in your curves.

Gregory wrote:

> You can always use the camera in the shade I focus approximately 7
> inches from a grey card so the card fills the exposure area. An
> overcast day is (Ideal "IMOP") for film testing.

Sandy wrote:

> If flare were consistent and predictable we could account for it, but
> it is not. It varies considerably with bellows position, lens, subject
> lighting, and type of lens hood used. And, unfortunately, even small
> differences in flare will distort curves significantly, so much so
> that the method makes comparison of effective film speed of different
> films virtually useless.
> Second, the accuracy of most shutter systems does not provide the
> degree of precision necessary to compare different films and determine
> EFS. You really need some type of light integration for this.
> The BTZS system of testing, which eliminates most factors other than
> the characteristics of the film material itself, is a much superior
> method of testing.

Precisely because flare renders curves determined in a no-flare environment
meaningless in real life, one needs to test film in the camera, exposing
through the lens, in order to have any real idea what tonal values you are
going to get in the shadows when you expose film in the real world.

I have not had the difficulties that Sandy notes as far as differences in
flare with any one lens under different shooting conditions. This may be
the result of two routine practices: I never use a lens that is less than
scrupulously clean and free of scratches on all surfaces (including the
internal surfaces), and I do not allow bright light to fall on the lens
itself, except in rare instances when the source of illumination is in the
picture. [Note that just a small amount of internal haze in a lens wreaks
havoc with shadow contrast, much more so than a lack of optical coating.
To check for this, shine a penlight through the lens and look in the other
end, moving the penlight around so it illuminates the glass but doesn't
shine directly in your eye.]

However, there is a great deal of difference from lens to lens, at least if
you have a selection of old and new lenses of widely varying optical
formulae, as I do. So, the bad news is that you need to run curves for
every lens if you really want to be able to previsualize your results.

There are several other problems with the contact-printing method. First,
it is usually done under an enlarger, and the color temperature of the
illumination is something other than 5500 Kelvin. This introduces errors
because the sensitivity and curve shape of film emulsions varies by
wavelength. Further, it is very difficult to accurately time enlarger
exposures as short as typical camera exposures unless you have a shutter on
the enlarger. The on-off times of incandescent lamps are variable and are
much, much longer than typical camera exposure times, and cold lights do
not fire up repeatably or come to full brightnes during such short
exposures. On the other hand, if you lengthen the exposures to make them
manageable, you face errors due to reciprocity failure that affect both
absolute sensitivity (i.e., apparent film speed) and curve shape.

Here's my method for avoiding these errors: I use an enlarger color head
(removed from the enlarger), which gives me a bright, evenly illuminated
surface that I can adjust to 5500 Kelvin. I put a step wedge on this
surface, then shoot it with the camera. This gives me the full range of
tones on one sheet of film, provides a typical flare environment, and
allows me to use the same shutter speed every time I test, for consistency
(I can dial in additional neutral density with the color controls to
control exposure). The only disadvantage I see is that the image must be
in focus, which means that I have to account for the bellows factor. (If
you are using Gregory's method, you can leave the lens focused at infinity,
as long as the target is considerably larger than the camera's field of

Best regards,

Received on Wed Nov 10 22:20:42 2004

This archive was generated by hypermail 2.1.8 : 12/08/04-10:51:32 AM Z CST