Digital BW, The Print

Greg,

I share your scepticism about the absolute accuracy of such high
density (low L*) readings. No matter what the accuracy, however, both
the TR924 and the Datacolor 1005 show reasonable precision, and the
patches that result in density readings of, say, 2.6 appear darker
than patches that read 2.5, as long as you view them in good light.

We don't seem to have a problem with a lightmeter that claims a third
of a stop accuracy (equivalent to a step of 0.1 D) over a range of
eighteen stops, so why do we distrust a densitometer or spectrogizmo
over ten stops? The difference between 2.0 and 3.0 can be described as
the difference between 1% and 0.1% (hey, that's only 0.9%, surely the
meter isn't that accurate?) or as three stops.

Whatever the accurate numbers are, the effect is there to see:
impressively deep shadows that are either featureless or full of
detail, depending on what's there and how bright the viewing light is.

Both the TR924 and the DC 1005 are 45/0 instruments, they are not
measuring diffuse reflectance. I've just read the 'density' of my hand
mirror with the DC 1005: 2.33 (L* is 4.3).

Best,
Helen 

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Greg"
<dfaprinting@...> wrote:

> The Printfix Pro (Datacolor 1005) device, and just about all the 
> other spectrophotometers from Gretag and Xrite have a specification 
> of 2.5D, so when I start seeing numbers below that, I have to 
> question where they're coming from. Since I knew you had used 
> different densitometers, I thought you might have had something of 
> the quality to get accurate readings that low, and thought I would 
> ask. To check, I contacted Xrite this morning, and confirmed what I 
> thought, here is part of the reply:
> 
> "A density value of 0, means the all light is reflected from the 
> surface
> (again in a specific region of the visible spectrum), a density value 
> of
> 1.0 means that 10% of the light is reflected from the surface, a 
> density
> value of 2 indicates that 1% of the light is reflected from the 
> surface,
> a density value of 3 indicates that 0.1% of the light is reflected 
> from
> the surface. As such values over 2.0 are extremely rare. Because this
> metric is log based, when you go beyond 2.5, very small changes in
> reflected light will cause very large numerical shifts."
> 
> He also said that L*a*b* values can not be directly converted into 
> density values, I'll check this later for myself.
>