ðòïåëôù 


  áòèé÷ 


Apache-Talk @lexa.ru 

Inet-Admins @info.east.ru 

Filmscanners @halftone.co.uk 

Security-alerts @yandex-team.ru 

nginx-ru @sysoev.ru 

  óôáôøé 


  ðåòóïîáìøîïå 


  ðòïçòáííù 



ðéûéôå
ðéóøíá












     áòèé÷ :: Filmscanners
Filmscanners mailing list archive (filmscanners@halftone.co.uk)

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: filmscanners: pixels, printer dots, etc



Hi Ken,

One thing I've learned long ago is that simple questions rarely create 
simple answers.

The first problem is that of definition of terms.  There are many types 
of printers used with computers, and there are many different types of 
each subclass, each using different methods of achieving their aims.

But, let's just speak about color inkjet printers, since that is what 
most people in this forum probably use.

The strategy is "simple", that being to attempt to reproduce as many 
colors as possible with the least number of inks and complexity.

As you probably know, most inkjet printers used by consumers fall into 
two basic categories.  Ones which use a CMYK inkset, and ones which use 
a CcMmYk set.  Commercial inkjet printer often use other ink sets which 
might include green and orange, or yet other combinations.

Let's take a quick look at how colors are produced with a standard 
inkjet printer. CMYK.

The only "real" ink colors these printers can produce are:

C = cyan
M = magenta
y = yellow
K = black
CM = blue
CY = green
MY = red
CYM = process "black" (more a muddy dark brown usually

Since adding K (Black) to any of the colors, pretty much overprints the 
color, those combinations don't produce any real "colors" to speak of.

Now, not only is the above list pretty limited, but it also is 
indicating only one specific shade of each of those colors.  Not very 
helpful when making a photographic image.

So, here is where a trick of the eye gets involved.  By using a mixture 
of a grouping of different colored dots, the space around them (the 
paper) and in some cases, a variation of dot size, the printer creates 
the illusion, using a dithering pattern of some sort, to trick our eyes 
into seeing a lot of colors.  This is not all that different from how 
offset presses mix printers inks to create magazine images, which can, 
when properly printed, rival photographs themselves.

The type of "matrix" used to create a wide variety of colors via an 
inkjet printer varies with the manufacturer, ink types and color 
densities, and whether the printer overprints or not.  There are some 
basic formulas for determining how many "cells" or dots of ink one needs 
to produce different bit depth interpretations, but frankly, I don't 
find it all that useful.

In the case of an image which is downsampled to some ridiculously low
resolution, the printer is printing a series of "cells" together all 
with similar information to make up the large "pixel" representation.

What is important to understand, however, through all of this is that if 
a printer claims 300 dpi, or 600, or 720, or 1440 or even 2400 or 2880, 
that does not translate to resolution of one color pixel (well unless it 
was a perfectly translatable ink color as shown above (cyan, magenta, 
etc,.))

If you read any printing list about inkjets you will see many 
disagreements as to what the maximum resolution a printer can use as 
source material before it becomes overkill.  Epson, for instance, used 
to recommend 240 dpi or ppi input for their 720 x 720 dpi printers.

Most people will tell you that there 1440 by 720 printers show little 
improvement beyond 340 dpi.  As to if anything more is achieved or 
achievable in their 2880 x 720 dpi printer, is hard to determine.

What you need to know is that all inkjet printers require many less 
individual pixels as input in relationship to their advertised 
resolution at output.  That number usually simply means that is the 
resolution that the printer can address one specific location on the 
paper, but it only means that the printer can deposit one drop of ink at 
that site, not that the printer can produce any specific color at that 
site.  So you don't want to scan at 1440 dpi at full size to deliver a 
file to an Epson 1440 dpi printer, the driver will just toss all that 
extra info, and slow down the processing, require a bigger printer spool 
space, and make for ghastly huge files.

In regard to your question about printer driver resolution, again each 
printer operates differently.  Some produce less, but larger drops of 
ink in lower modes, and therefore may produce a lower number of 
perceived colors, or a coarser look.  I'm not sure a lot of ink is 
saved, but printing time is usually faster.  The truth of the matter of 
  how many drops are needed to produce a specific color, depends upon 
the color one is trying to create.

Art


Ken During wrote:

> Ok, I have I think I simple question, stemming out of my study of
> Wayne Fulton's scanning tips.  Just getting clear, so forgive me if
> it's a stupid question. 
> 
> On page 67 of that book he shows a tiny 32 pixel image scaled to 5
> dpi.  It's printed as a 6.4 inch graphic with pixels that are,
> obviously, 1/5 inch in size.  My question is, what does this say about
> print resolution?  The printer is obviously using a certain number of
> dots to produce one pixel.  Is this number of dots specified simply by
> selecting the print resolution in the printer driver menu?  
> 
> Thanks
> 
> 
> Ken Durling
> 
> 
> 
> Photo.net portfolio: 
> 
> http://www.photo.net/shared/community-member?user_id=402251
> 
> .
> 
> 






 




Copyright © Lexa Software, 1996-2009.