Raster graphics have a fundamental problem: They are made up of
raster points. Raster points have different sizes on almost every
output device. If you created a raster graphic in one resolution, but
want to print it in another, the raster points must be enlarged or
shrunk. Unfortunately discrete values – which is exactly what
raster points are – cannot be reproduced smoothly on another,
equally discrete raster. The problem is that in the digital domain one
can only talk of yes and no, on and off (the analogy to the expression
a little bit pregnant
being impossible comes to mind). The
whole thing should be made clearer with an example: Imagine you are
looking at a greatly enlarged raster that has two white raster points:

So you have two raster points. Let's say we want to increase the width of the raster by a factor of 1.5. As a result, we will have three raster points in a row:

As long as the points are all black or all white (as above), there is no problem, but what happens if one point is black and the other white – how does the enlarged raster look?

The first point of the enlarged raster is certainly black and the third white, but what about the second point? That is the first problem. It manifests itself in the fact that some pixels will be doubled up, others not. The result is often ugly, because the graduations become regular and therefore noticeable.
The second problem occurs when shrinking raster graphics. Let's say you want to shrink this graphic

to a quarter of its size in both directions. As a result, you will have only one point instead of the original four. Depending on where you start to count, you will always get a black or always a white point. The subtle raster disappears, and you end up with large black or white areas. In conjunction with the first problem of rounding off, both enlargement and shrinking can create a Moiré effect, a series of streaks in the middle of your picture. These Moiré effects may be familiar to you as they may be seen also when travelling by car or train. Drive under a motorway bridge and look at the overlapping of the two parapets. Parts will be quite dark, in other areas you will then see just one of the ballustrades. The analogy may be a little hard to understand, but the effect is really the same. The basic cause is optical interference of one pattern with another.
Back to the raster graphics function Optimize frame size for
screen
. This function enlarges or shrinks your raster graphic so
that in the Actual size (1:1) display each raster point (pixel)
corresponds exactly with one pixel (or with a multiple of that
pixel) on the screen. If the Optimize enlargement
function (see
below) is not active during this, the graphic will always appear in
its original size, each image pixel corresponding with a screen pixel.
If the Optimize enlargement
function is active, the enlargement
factor will be rounded to the next highest whole number. In either
case the image will appear in its original aspect ratio, that is its
proportion of length to width is unchanged. Optimization avoids the
distortion created when resizing a raster graphic. We repeat the most
important factor: You will only get a 1:1 (or x:1) display on screen
when you have selected the Actual size (1:1) icon in the Top row.