Oh, dear... This requires a stern coffee...
OK. We know what a bitmap or raster graphic is. (Those who don't please
look it up in WikiPedia and come back...)
A "Vector" graphic is a graphic in which the pieces are all comprised of
mathematical formulae. We all know what a straight line is. It's in fact a
"curve" that has no curvature. So a straight line and a curve are created
from a mathematical formula that specifies the starting point, the length,
the angle, and the rate of change. A "square" is just a broad line. A
circle is a broad line with no length. And so it goes...
There are two kinds of vector illustrations. "Metafiles" and "Page
Description Languages". The most common version of Page Description
Language is PostScript. PDF is a flavour of it, and so is EPS. Internally,
the code is PostScript.
In it, the mathematical formulae describe the colour and position of marks
in a "space". When you print or display one, the printer or display driver
tells the computer how big the page is, and what the device's resolution is
(say 96 dpi for a screen, or 1200 dpi for a printer). The formulae can then
all be scaled to place the objects in the correct place.
This has the advantage that the resolution of the picture is infinite,
limited only by the device. A picture may look pretty fuzzy on a mobile
phone screen, but print the same picture to a Sputtering Etcher at 9,600 DPI
and your eye will not be able to see the imperfections. A "feature" of page
description languages is that not only their resolution is infinite, so is
the "complexity" of the images they can describe. (A sputtering etcher is
one of the devices high-end printing publishers use to make printing plates
from digital images, avoiding the "dottiness" that comes from using a screen
for photo-etching).
This infinite resolution doesn't cause much trouble at home, but you should
hear the bad language that a Master Printer (person!) is capable of when
their Raster Image Processor runs out of memory and crashes trying to print
a 28-layer engineering drawing of a skyscraper in seven colours
I was
there when one happened. I never would have believed Donna knew all those
words...
In the real world, we needed a solution that was a lot more compact.
Particularly back in the days when a "fast" computer had an 8 MHz processor,
and 640 kb was a lot of memory...
So was born the "Metafile". Common examples today are PICT, WMF, and EMF.
But they all have their roots in CGM - Computer Graphics Meta Files. CGM
has just been ratified as the ISO standard, so we'll see a lot more of it in
the future. The name "Metafile" comes from the fact that instead of
containing the actual drawing objects, the file contains only "information
about" the drawing objects. The drawing objects themselves are pre-defined
and stored in the computer that is trying to process the file.
I found this a really difficult concept to grasp, so let's see if I can
provide an example. Let's take a "Line". It has a position, a length, a
direction, a line weight, a colour, a brightness, and a shading gradient.
In a page description language, ALL of this information is accurately
specified in the file. In a metafile, only the position, length and
gradient are specified accurately. The rest of the information is sent as a
series of "menu choices": Line weight = 5, Colour = 32, Brightness = 5, and
"gradient" doesn't exist. Because these are choices from a limited menu,
they can all be described in a single binary word, saving a LOT of space.
Which is one reason that a metafile can be one hundredth of the size of the
equivalent PostScript file. Obviously, the amount of memory taken to
process it is also dramatically less, and so are the number of CPU cycles it
takes to decode it. And because a metafile contains a large number of
repetitions of predefined "things", it is easy to compress it by a factor of
20 without losing any information.
Yet the result, when you print it, will be exactly the same as it would have
been if you had printed an unlimited resolution PostScript file. Because
the quantities are mathematic, they can be scaled at the maximum precision
of the output device.
However, you don't get anything for nothing in this world
There are
distinct, absolute, specified limits to the picture elements any given
metafile format can describe. In the early versions of metafile, these
limits were often in the area of "precision" and "colour". WMF is one of
the few formats still in use where you can see this: print a WMF image on a
very high resolution printer and you will see that it is comprised entirely
of straight lines
Take a magnifying glass to a small feature and you
will notice a distinct jagginess to the outline (this may not happen in
text, that is often sent as the original font, so the printer can restore
the resolution by using its internal fonts). Another area of economisation
was in the colour table. A WMF can describe only 16,000 distinct colours:
lovely shading gradients turn into nasty zebras of banding.
Right: Now your head is hurting some, it's time to get technical
Each metafile starts with a "menu" of the "widgets" it can describe. In
Microsoft Office, they're called "shapes". A "shape" certainly has a
visible shape, but it can also have other properties: "behaviours". Since
we know what this thing is, we can thus know how it behaves. You see this
in the Connector lines in PowerPoint. They not only know that they're
lines, with a thickness and a colour, they also know they have "elbows" and
"sticky points". So you can "glue" a connector to another drawing object,
and as that drawing object moves, the connector will stretch to remain
connected.
To answer one of your questions, a "Shape" in Microsoft Office is a drawing
object that has many properties, such as associated text, embedded data, and
a set of behaviours.
Now: To answer your other question: "Why can't I mix shapes and PICT?"
Well, you "can", but ...
Competition is a wonderful thing. Except when it happens on a computer near
you... Competition has caused the companies that made our computers to
strive to build them as good as they are today. But it also means that they
don't want to do things the same way as each other: not only is each
convinced that their way is "better", but each would sue the other if they
got half a chance.
So: Remember that PICT is a metafile? It relies on an internal menu of
"things" that can be described, such as line weights, colours, and
behaviours. Internally, Microsoft uses EMF, "Enhanced Metafile". Like
PICT, it grew from CGM. However, whereas PICT is optimised for the graphics
engine contained in Mac OS, EMF is optimised for the one in Windows.
In Mac Office, the drawing packages operate with an internal language based
upon EMF, and convert their output to PICT when they embed it in the
document. (That's an oversimplification, but stay with me...)
Let's assume that PICT can have line weights of 1, 3, 5, 10, 50 and 100.
EMF can have line weights of 0.1, 0.3, 0.5, 1, 10, 50, 100 and 1000. (This
is a fictitious example...) When you convert from EMF to PICT, you have to
take the "nearest available equivalent". For graphic attributes, it's
unlikely to be a problem because (with the exception of line weights) both
formats are very similar. It used to be exciting going from PICT to WMF,
because PICT had a 32-bit colour table, so badness happened. However, both
PICT and EMF are 32-bit formats, so provided you only want to display
things, the conversion is very accurate.
Not so when you get to "behaviours". PICT supports only 'some' of the
behaviours EMF can describe, and vice versa.
So the answer to your question is: "You can mix PICT drawings and shapes,
provided that the shape you want is a simple one whose behaviours are
supported on both platforms." Regrettably, most Microsoft Office Shapes are
not that simple. And that's why.
When you edit a metafile, the result is expressed in the capabilities of the
program that did the editing. If you edit a PICT in Microsoft Office and
paste it into a high-end graphics program, the graphics program won't have
much trouble with it. If you paste into something like Canvas, a lot
depends on what is in the picture. There are things Canvas supports that
Word can't, and vice-versa. Same with PC Word: Unless the PC user has
QuickTime 6 or 7 installed, they won't be able to handle the full range of
things that Microsoft Office 2004 can put in a picture.
As to exporting your pictures: try it from PowerPoint. PowerPoint has a
greater ability to export things to PICT or import them from EMF/WMF than
the other applications do. Copy your picture into PowerPoint and export the
slide.
There: Now I need another cup of coffee...
Hope this helps
To add more confusion, there are .pict "vector" graphics which come
from inserted .pict files and Word "vector" graphics made in Word with
the graphics tool bar or whatever it is. Why is this and is there a way
to make them the same ? I haven't been able to mix the two.
Also, editing a .pict "vector" graphic and saving it sometimes saves
it as a graphical object or a Word object, which can't then be edited
again, or copied and pasted into a graphics program such as Canvas.
And then if the Word file is copied to a PC, some vesions of PC Word
won't allow you to edit a .pict "vector" graphic.
Could you comment on and explain all these confusing different forms
of a "vector" graphic ?
Thankss
--
Please reply to the newsgroup to maintain the thread. Please do not email
me unless I ask you to.
John McGhie <
[email protected]>
Microsoft MVP, Word and Word for Macintosh. Consultant Technical Writer
Sydney, Australia +61 (0) 4 1209 1410
