[Leica] Tina and Digital (was, Enlarger

[bquinn at sgi.com (Barney Quinn)]

Marc,

I'll give this a try. There are, in general, two ways to store an image on a
computer. One is called vector format. In vector format you don't store the
image, you store the instructions for making the image. In the old days a
vector, in basic physics, was a line which had both magnitude and direction.
In vector graphics you store an image as a set of instructions which say
things like draw an orange line from this point to that point. This is how
postscript works. It is a set of instructions which describe how to make the
image in question. Postscript files are plain text files and if you want you
can open one up with a text editor and actually look at the instructions. If
you are really adventurous, and if you know postscript very well, you can
even take a text editor or a word processor and write out all the
instructions to make an image out by hand. But, all you need to remember is
that in vector graphics you don't store the image, you store the instructions
for making the image. It's the recipe, but not yet the dinner you are going
to serve.

The other way to store or represent an image as as what is called a bit map.
This works the same way as a picture in a newspaper does. The image is broken
down in to a bunch of dots. Each dot is stored as a number, or a series of
numbers, which represent the things you need to know about the dot such as
its color. These dots are usually called pixels. You could store a very crude
image by breaking it down into a series of dots which were either black or
white. If you represented each black dot with a zero and each white dot as a
1 then you could store the entire image by storing the 1's and 0's which
represented the image as a file on your computer. This is exactly how digital
photography works, only more information is stored because you need more
information other black or whit to make a continuos tone image. If you wanted
to compare an image made with just ones and zeros to conventional photography
it would look like something printed on litho paper because there would be no
grays. In a slightly more sophisticated system rather than describing each
dot or pixel with a 0 for black or a 1 for which we could describe it with a
number from 0 to 7 where the numeric value told us how gray the dot was,
sorta like the zone system.

There is a tug of war going on between these two approaches. Vector format is
useful for some things, such as line drawings, of a page of text, because it
takes less space to store the instructions to make the image that in would to
store the actual image. But, as an image becomes more and more complex you
need more and more instructions in order to store it in vector format. In the
case of something like a photograph the amount of space you would need to
store the instructions to make the image can get to be more than the amount
of space needed to store the image as a bit map. What this means is that it
takes less space to actually store the image than it does to store the
instructions for making the image. There is a slight complication here, which
we will not go into at this point. Imagine a black and white image with a
high white cloud in it. That is a lot of white, and it takes a lot of numbers
to describe it.

There are techniques, called data compression, which will let you store the
image of the cloud without actually having to store all the ones and zeros
which make up the cloud. This is called data compression. But, here's an idea
of how it works. Let's say you wanted to draw a line which ran vertically up
and down the page which was ten inches long, and let's say that you made your
image by storing 300 pixels, or dots, for every inch the line was long. In
the real world computers store information using groups of eight 1's and 0's
called bytes. In the most basic real world graphic systems one dot, or pixel,
is stored as a byte or group of eight ones or zeros. To represent the ten
inch line would take 10 times 300 pixels, or 3,000 bytes of space. It takes a
lot less space to say start here, and go north for 3,000 dots. The first is
bit format, the second is vector format, and the difference can be thought of
as data compression.

Here's where the rubber hits the road. A bit mapped image is just a bunch of
ones and zeros stored inside your computer. A vector image is nothing more
that a set of instructions showing how to draw the image. Neither is an
image. Both are just a lot of data which is unintelligible to a human. To
actually get a picture on the wall it has to be printed. This is done with
something called a rendering engine. Computer geeks use the term engine for a
program which does the same task over, and over.

The easiest way I know to answer your question about the meaning of the term
RIP is to consider what happens when a line drawing is printed out on a piece
of paper. Let's say that you are using a printer which prints its images at
300 dots to the inch. What that means is that a black square one inch by one
inch would be made up of three hundred dots across, running from left to
right, and this would happen 300 times running down the page. When the
printer printed the image it would start and put 300 black dots on the page
running from right to left.  Then it would finish running across the page.
The print head which made the dots would move back to the left hand side of
the page, and page would be advanced 1/300 of an inch, and the print head
would make 300 more black dots. This process would repeat it self until both
the one inch by one inch square and the page had been completed. This should
also give us a hint about why this is am important question. A black square
an inch by an inch at 300 dots to the inch is made up of 300 x 300 or 90,000
dots. The amount of information you have to store and manage goes up by
squares as the size of the image increases.

A TV set makes an image by sweeping a beam of electrons across the screen,
causing the appropriate "dots" to "light" up as it goes. When it is done with
the first line it skips down a line and does another. It works its way down
the page first doing all the odd numbered lines and then comes back up to the
top and does all the even numbered lines. This is called interlacing, and we
aren't going to go into why it's done that way today. This pattern or across,
down, across, down, across is called a raster. I have no idea why. A printer
works the same way, only it doesn't skip lines. It does the first, then the
second, then the third line. Because of the similarity of how a printer and a
TV work people started calling an image inside a printer a raster, too. RIP
has three meanings which I know of. The first is Rest in Peace. Most people
by now are probably wishing I had done that five paragraphs ago. The second
is Raster Image Processor, and the third is Raster Image Profile.

All printers have A RIP, or raster image processor built into them. Let's say
that you want to print an image which is stored in vector format. First you
have to convert the instructions for making the image into ones and zeros or
pixels. This is done by a piece of software called a rendering engine. In the
case of postscript the instructions which make up the image are rendered by
something called a postscript interpreter. In the case of a compressed bit
map image the image has to be uncompressed before it can be printed. The
resulting ones and zeros are now handed to something called a RIP, or raster
image processor. The RIP converts the ones and zeros into the back and forth
and up and down instructions that the printer needs to actually put the image
on paper. Every engineer on the list is probably cringing at this crude
generalization, but it is the basic idea.

Now, to the other meaning of RIP. Raster Image Profile. The world is not an
ideal place, Physical systems are not ideal. Corrections need to be made to
the information stored in the computer which represents the image in order to
get it to look "right" when it is printer out. You might, for instance, have
to make the red's a little bluer in order to get them to look right. The best
way I can explain a Raster Image Profile is that it is the set of corrections
which must be made to get the image to look "right" in the real world using a
specific printer and piece of paper. If is very much like the orange mask in
a color negative, or the correction pack which you use during color printing
to compensate for things to get the image to look proper.

May God forgive me for this unending string of over simplifications. I hope
it helps.

Barney



Marc James Small wrote:

> At 05:25 PM 2/6/06 -0800, Frank Filippone wrote:
> >Think.... RIP= Ripped O..
> >
> >That is pretty close, and 60's Lingo, so that should work for you....
>
> I am even more lost.  "Ripped Off Plums"?  I am really at sea here and the
> boat is sinking rapidly.  I do not need hints.  I need reality.
>
> You folks are not being much help at all.
>
> Marc
>
> msmall@aya.yale.edu
> Cha robh b?s fir gun ghr?s fir!
>
> NEW FAX NUMBER:  +540-343-8505
>
> _______________________________________________
> Leica Users Group.
> See http://leica-users.org/mailman/listinfo/lug for more information

--
Barney Quinn, Jr.
(301) 688-1982 (O)
(240) 535-3036 (C)
(877) 220-0981 (P)