RE: [APML] Picture Window Pro vs. PhotoShop

["Thor Olson" <Thor.Olson@EFI.COM>]

Just a few offhand comments on this interesting discussion...

... If you have an 8 
> > bit image on 
> > the monitor next to a 16 bit image, they are both being 
> output to the 
> > monitor at 8 bits, that is all the monitor will work with.
> 
> I don't think I understand it either.  Maybe PW is doing 
> something with
> the data, but it's always been more or less obvious to me 


I like to characterize human vision as being "9-bits sensitive".
There's alot of auto-gain adjust and dynamic compensation,
but a discriminating observer (e.g. someone who may have spent
a few hours at an eyepiece scrutinizing low level detail in
an obscure deep sky object), will be able to distinguish image
features beyond an 8-bit representation.  It takes practice,
and the right equipment, conditions, and environment, but it
is measurable.  This may explain what is observed if the
"right 8 bits" are selected and presented to the user from a
16-bit deep image.  Jerry is right though, the display itself
is really only an 8-bit device, and if the same 8-bit frame
is displayed from a 16-bit original, there is no physical
distinction.


>  
> > Likewise, you may have a 16 bit image originally and may 
> > correct it in 16 
> > bits in PW, but before you output it to a printer, it has to 
> > be converted 
> > to 8 bits. Printers only work in 8 bits.
> 
> True, and the difference in prints made at each bit-depth is much less
> noticeable, but I guess I've always attributed that (without any hard
> technical cause to do so) to the way I think ink moves when it hits
> paper.

The issue here is most likely dynamic range of the print versus
the monitor.  Many prints have a tough time making a 100:1 contrast
range from black to white.  CRTs are considered a 100:1 device, and
projected slides in a dark room can approach 1000:1.  Obviously with
a smaller contrast range, fewer levels (bits) are needed to smoothly
span the range.

>  
> > Photoshop, btw, only really operates in 15 bits, not 16, they 
> > are actually 
> > pulling a trick.

I've not heard of this, can you elaborate?

Regards,
Thor.

Thor Olson                            
web-res astrophotos at:             http:\\www.nightscapes.net

   "Man has ten, computers have two, Nature has e fingers."



--  APML Archives at <http://astro.umsystem.edu/apml/>  ---
             Unsubscribe at <majordomo@seds.org>