Re: [APML] CRT vs Flat Panel for display of astrophotos

[Dave Hutchison <davehutch@flash.net>]

Quoting Rick Kellogg <rkellog1@twcny.rr.com>:

> Then (IMHO) this a pretty meaningless measurement.
> For a CRT, under no ambient light, in a black room that absorbs all light
> from the CRT, I would expect to get very high contrast ratios by just
> turning down the brightness until black areas emit no light (other than what
> is scattered internally).  But that does not mean that all CRTs with similar
> brightest white to black ratios can display a good gray scale.

I would aggree with you that contrast ratio does not determine the ability of a
device to display a good gray scale.  It does however give an indication of how
good a black is.  The problem is that it does not account for ambient lighting
conditions so one is always left with the question of "what contrast ratio is
good enough?".  CRTs tend to have better contrast ratio when compared to LCDs. 
But in the bright lights of a Best Buy store, this might not be a very
noticable.

> 
> I did not mean to imply that bits have much to do with it (yet I knew that
> is how it would be taken).  An analog display driven by a voltage, does not
> have any bits, but has an equivalent signal/noise level that limits the
> dynamic range possible.  Neither bits nor S/N should be the limiting factor,
> yet 8 bits does not seem to be enough on high end broadcast video recorders
> to eliminate the perception of banding on gradients of small color range.
> But this is probably a red herring on my part anyway, and not worth any
> further thought.  Of the 10 bits, it is likely only 7 1/2 - 9  are signal.
> A lot of these gradient issues are from computer created images, not signals
> from cameras, and thus have artificialy high dynamic range, that can band
> when limited to 8 bits.
> 

You nailed it.  8 bits is not enough on high end video.  Furthermore, video is
transmitted with a non-linear transfer function (gamma corrected).  When you
consider that many of the digital display devices are linear in nature, then
internal representation of the display data needs to be greater than 10 bits. 


Dave
_______________________________________________
Astro-Photo mailing list
Astro-Photo@seds.org
http://seds.org/mailman/listinfo/astro-photo