[filmscanners] RE: 3 year wait

["Laurie Solomon" <laurie@advancenet.net>]

Alas Arthur, the quotes in your post are attributed to me; but they are not
my words. Since I did not write them although I did raise some questions
concerning the question of 4K RES used to define film recorder resolutions
as referring to LPI and not ppi which would mean that the resolution in ppi
would be 1.5 - 2 times less than 4K, I am not sure if your post is addressed
to me since we apparently are saying the same thing be it LPI or line pairs
or if the post is directed to others.

The key practical point is that many if not most film recorders do not
actually operate at 4K despite the written specs so their resolution in ppi
is going to be far less than the 2000 ppi - 2666.66 ppi that it would be if
the 4K lpi was an actual operating resolution. I understand how you derived
your 2849 ppi figure; but my math says that to get a 4000 lpi line screen
you would have to multiply by a factor of 1.4 (most printers suggest
multiplying the ppi by a factor of 1.5 - 2.0 to derive the halftone line
screen (lpi).

-----Original Message-----
From: filmscanners_owner@halftone.co.uk
[mailto:filmscanners_owner@halftone.co.uk]On Behalf Of Arthur Entlich
Sent: Saturday, May 11, 2002 6:07 AM
To: laurie@advancenet.net
Subject: [filmscanners] Re: 3 year wait




Laurie Solomon wrote:


>
>>"4K" simply means 4000 (and 96) pixels across the 36mm film chip.
>>Actually, 2889.9ppi.
>>
>


The problem above is the direction of the film being measured.

A film recorder refers to the longer dimension as 4K, so the 4096 pixels
across, represents the approximate 1 and 7/16th inch wide frame, (4096
pixels divided by 1.4375" equals 2849 ppi or so)

On the other hand, most film scanners scan the narrower width of the
film frame, so a 4000 ppi scanner needs to have a apparent sensor
density of about 3750 elements across the film frame in the short direction.


>>I'd say that 11-12MP of true pixel info IS pretty
>>much what (Ektachrome,
>>at least) film can resolve.
>>
> [What is a MP and is that a standard abbreviation?]
>
>


MP is Mega-pixel, as usually used in digital camera sensors. One MP is
1024K pixels (in an array).  So a one megapixel array would be about
1200 x 800 pixels.  However, the reason the reference is "true pixel" is
because other than the new Foveon chip, digital cameras use a CCD or MOS
array which can only record one color (RGor B) per pixel location, and
therefore an array is used (called a Bayer Pattern) which uses
interpolation to take a matrix of 25% R and B and 50% G filtered pixels
to make up the color information, and therefore the accuracy is
lessened, and interference patterns (moiré) develops with some color
combinations.


>>T-Max 100 has a resolution rating of around 200 line pair/mm, that's over
>>10k samples per inch, and would be a file of APPROXIMATELY FOR EXAMPLE
SAKE
>>(since you are being anal about arithmetic ;-) ~10k x ~15k or ~150M pixels
>>
>
>


I'll leave someone else to explain this part, but the main point is that
you cannot measure to a single line because there needs to be a contrast
developed, meaning a series of lines, such as black line, white line or
blank space, black line, etc, to measure. That is why the term is line
pairs, since a minimum of two lines (or the space between them) is
required to read the result.

Art


----------------------------------------------------------------------------
------------
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