From: Jeffrey D. Mathias (jeffrey.d.mathias@worldnet.att.net)
Date: 05/26/00-03:34:19 PM Z
FotoDave@aol.com wrote:
> ... You originally asked how many zones are needed and if there was a study. Last
> night before I slept I suddenly remembered where I read about it. It was
> mentioned in Ctein's "Post Exposure." ...
> ...And that is how subliminal messages work. ...
It seems Katherine has corrected this.
> ... that is the natural, uncorrectedly response of the sensor. ...
Yes the detectors in the scanner record a linear relationship of current
per light flux.
> ...That is why it is better to scan positive than negative. ...
> If we have a 12-bit scanner, either the curve
> during scanning or curve using the 12-bit data can improve a little. ...
> I believe it will convince you that scanning negative can cause some trouble ...
This is somewhat true. However, if the original negative's maximum
density is within the 12-bit scanner's range of accurate sensitivity,
there is no problem with having enough signal (depth) in the data
(pixels). This is why I have adjusted original negatives slated for
digital processing.
The detectors in the scanner are only capable of producing a certain
minimum level of current before the signal blends into the base noise.
The key is to only allow a minimum signal which is clearly above the
noise level by a margin comparable with the pixel depth desired. (For
example: if the noise level is hypothetically 0.01 uA and accuracy must
be 0.0244 % (1/4095 or 12-bit), the minimum signal level would be 40.95
uA) For all signals and the resulting data above this minimum level,
there will be identical accuacy because the limitation becomes the depth
(12-bit in this example).
Thus, when the minimum signal level is kept, it will make NO difference
in data accuracy whether a negative or positive is scanned.
There is also a maximum signal level most likely limited by the
electronics measuring the current (impedance matching) before being
limited by the device itself. At higher current levels, the linear
relationship will cease as voltage is measured rather than current.
This may be able to be determined by scanning with alternate
illumination sources. (A cold light may be used which produces more
light flux than the scanner's light source.) Keep in mind that the
scanner designers probably already selected the brightest lamp they
could. But if more light flux can be used, then a negative with greater
density range could be scanned.
The easiest solution has been to limit the density range of the
negative. I have found this to be a bit less of a range than typically
used for Pt/Pd printing. When a scan is made on a block of maximum
density negative (which includes base + fog), the noise can be seen.
Scan blocks of decreasing density until no noise is seen and this will
indicate the minimum usable signal level of the scanner. A 21-step can
be used for this. (Do not confuse noise with dirt. With noise, scan
lines can be detected.)
It is important to get the original negative as close as practical so as
to avoid loosing information. I would also venture a guess that pyro
developed negatives may not fair as well when scanned, if they have less
of a total density range. This does not mean that they cannot be used,
only that they may not deliver as much information.
-- Jeffrey D. Mathias http://home.att.net/~jeffrey.d.mathias/
This archive was generated by hypermail 2b29 : 06/13/00-03:10:22 PM Z CST