ðòïåëôù 


  áòèé÷ 


Apache-Talk @lexa.ru 

Inet-Admins @info.east.ru 

Filmscanners @halftone.co.uk 

Security-alerts @yandex-team.ru 

nginx-ru @sysoev.ru 

  óôáôøé 


  ðåòóïîáìøîïå 


  ðòïçòáííù 



ðéûéôå
ðéóøíá












     áòèé÷ :: Filmscanners
Filmscanners mailing list archive (filmscanners@halftone.co.uk)

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[filmscanners] Re: scanner dmax discussion



Austin
You are right that the A/D and detectors themselves are not calibrated
beyond the creation of the LUT.  But we are talking about this in the
curious context of both just the A/D and detector and the context of it
being inside a scanner.

It is not unusual to calibrate/test a scanner by sticking in a step tablet.
Assuming a fine stepped enough step tablet, you could have a
tablet-step/bit.  The gotcha here is that you could end up with some very
odd effects in this sort of case.

Consider what would happen if the tablet-steps matched the incremental
detector steps exactly, one of three things would be observed of the output:
    1) you would see an exact 1 bit increment in the data coming out of the
scanner for each of the tablet steps scanned - IF the tablet step densities
aligned perfectly with the A/D converter LUT values being generated
    2) you would see 1/2 as many steps and incremental values of 2 binary
steps in the data if the step tablet values were exactly 'out of phase' for
the A/D converter LUT values - ie assume that the A/D generates an output of
3 for a density of .01-.02 and stepn generated a sensed value of .01 and
stepn+1 generated a sensed value of .02
    3) you would see odd patterns if the 'phase alignment' was neither
perfectly in or out of 'phase'.

Reality is that you would see something more along the lines of #3 anyway
because no detector is that precise/accurate.

This is why the assertion that you really need to model the incremental
density resolution as a ValueX +/-Error  where Error= Step/2 is correct.
----- Original Message -----
From: "Austin Franklin" <austin@darkroom.com>
To: <karlsch@earthlink.net>
Sent: Saturday, July 12, 2003 11:04 AM
Subject: [filmscanners] RE: scanner dmax discussion




> -----Original Message-----
> From: filmscanners_owner@halftone.co.uk
> [mailto:filmscanners_owner@halftone.co.uk]On Behalf Of Paul D. DeRocco
> Sent: Saturday, July 12, 2003 1:47 PM
> To: austin@darkroom.com
> Subject: [filmscanners] RE: scanner dmax discussion
>
>
> > From: Karl Schulmeisters
> >
> > Except Austin, that's not how these devices are calibrated or
> tested.  The
> > universal (I have yet to see another one used but I am willing to be
> > corrected) approach is to use a step tablet with density
> increments of 1/3
> > of a stop and the 'dynamic range' is measured by seeing where
> the scanner
> > ceases to differentiate between two adjacent steps.
>
> I would assume that that limit would always be on the dark end, and that
> instead of running out of bits, the steps would just disappear into the
> noise. So noise would be the limiting factor, not the A/D resolution. Am I
> wrong?

Paul,

That is true only if the scanner is designed with the A/D resolution being
greater than the noise, obviously.

What he is bringing up is how to measure the scanner density range...which
really has nothing to do with what we were talking about, so I'm not sure
why he brought it up.

Regards,

Austin

----------------------------------------------------------------------------
------------
Unsubscribe by mail to listserver@halftone.co.uk, with 'unsubscribe
filmscanners'
or 'unsubscribe filmscanners_digest' (as appropriate) in the message title
or body

----------------------------------------------------------------------------------------
Unsubscribe by mail to listserver@halftone.co.uk, with 'unsubscribe 
filmscanners'
or 'unsubscribe filmscanners_digest' (as appropriate) in the message title or 
body



 




Copyright © Lexa Software, 1996-2009.