[Leica] Re: RE. Digital Characteristic Curves

["tlianza" <tlianza@sequelimaging.com>]

Hi to all,

Mark provide his data and I've looked at it and rescaled the data in a form
that is the same as that used in Zone system analysis. The big difference
between digital systems and film analysis is that the film information is
returned as DENSITY. To use Marks data (which was collected quite correctly)
to compare to a photographic system we must convert the digital values to a
VIRTUAL DENSITY VALUE.  Mark reflected his exposure, again, quite correctly,
as exposure shifts in f-stops.  If we wish to calculate a contrast index or
gamma from this data we must scale the f number shifts and convert  them
logrithmically as well.

To scale the digital values to form a virtual density using Excel we use the
following relationship

virtual density = -log10(nn/255)  where nn is the number we wish to scale
(assuming an 8 bit value)

to scale the fstop axis, as Mark has provided we use the following
relationship

log exposure = -log10(2^fshift / 2^maximum positive fshift)

The log exposure is the X axis of the characteristic curve
The virtual density is Y axis of the plot.

Now you have a plot that can be related in photgraphic terms to other
systems.
I will send a reduced spreadsheet and plot of the data Mark has posted here
to anyone who requests it.

here is the reduced data

      -5 5  2.40824 1.70757
      -4.5 6  2.257725 1.628389
      -4 9  2.10721 1.452298
      -3.5 13  1.956695 1.292597
      -3 20  1.80618 1.10551
      -2.5 24  1.655665 1.026329
      -2 32  1.50515 0.90139
      -1.5 43  1.354635 0.773072
      -1 56  1.20412 0.658352
      -0.5 78  1.053605 0.514446
      0 96  0.90309 0.424269
      0.5 128  0.752575 0.29933
      1 147  0.60206 0.239223
      1.5 177  0.451545 0.158567
      2 196  0.30103 0.114284
      2.5 231  0.150515 0.042928
      3 252  0 0.00514


The first two columns are the data as Mark sent. If you plot this data, you
will see a small shoulder, a rather short toe and a gradient of around .85
in the midtone region.  This is very agressive for display on a monitor, but
it mimics the gradient that you would get by scanning a transparency.  This
is not really optimimum at all.

The reason for my comment about archiving in NEF is that I honestly think
that Nikon understands that work will continuly improve in the color
management world.  Storage in Raw format will allow better algorithms to be
applied to the data AFTER capture.  Keep in mind that Nikon also allows
storage of NEF files from their scanner applications.  As our understanding
of the color science improves, we will be able to apply those advances to
archived images.

Tom Lianza
Technical Director
Sequel Imaging Inc.- A GretagMacbeth Company
25 Nashua Rd.
Londonderry, NH 03053

- ----- Original Message ----- 
From: "Mark Davison" <mark_e_davison@msn.com>
To: <tlianza@sequelimaging.com>; <leica-users@mejac.palo-alto.ca.us>
Sent: Thursday, July 31, 2003 10:32 PM
Subject: Re: RE. Digital Characteristic Curves


> A propos of T. Lianza's comments:
>
> I used the following methodology to measure the characteristic curve of
the
> D100:
>
> I set the D100 to RAW, Normal Tone Curve, spot metering, aperture
priority.
>
> I then took a series of photographs of a piece of siding on my house that
> has a very light texture, varying the exposure compensation in 1/2 stop
> increments.
>
> I brought in each NEF file to my computer, converted to Photoshop with no
> adjustments, selected the area I had metered on and used Image > Histogram
> to get the mean luminosity.
>
> This methodology mimics what is typically done in the film Zone System
when
> a step wedge is not available. Technically it is probing the
characteristic
> curve of the combined lens, camera and sensor system, rather than just the
> sensor itself. I used a 17-35 2.8 Nikkor zoom lens.
>
> Here's my data:
>
> Exp Comp   digital output
>
> -5 5
> -4.5 6
> -4 9
> -3.5 13
> -3 20
> -2.5 24
> -2 32
> -1.5 43
> -1 56
> -0.5 78
> 0 96
> 0.5 128
> 1 147
> 1.5 177
> 2 196
> 2.5 231
> 3 252
>
> I have plotted the data and there is no discernible shoulder.
>
> Mark Davison
>
> _________________________________________________________________
> Tired of spam? Get advanced junk mail protection with MSN 8.
> http://join.msn.com/?page=features/junkmail
>
>
>


- --
To unsubscribe, see http://mejac.palo-alto.ca.us/leica-users/unsub.html