View Single Post
  #123  
Old February 17th 17, 06:00 PM posted to rec.photo.digital,microsoft.public.windowsxp.general
nospam
external usenet poster
 
Posts: 4,718
Default Why exactly does Paint.NET make bigger files than Irfanview?

In article , Mayayana
wrote:


| "Adding canvas" would be accomplished by
| painting a bitmap onto a second larger bitmap, just
| because that's how Windows graphics works.
|
| If you do it that way it is a bit of a kludge. Bitmap is a very old
| Windows way of thinking. There are far simpler and efficient ways of
| doing what you have described.
|

You're doing it that way, too. You just don't
know it. (RAW is more complicated, in ways
I don't fully understand, but once you get into
raster graphics it's all the same.)

Now that I know what's meant by
canvas I don't see any problem with it. I'm
curious how I never noticed "canvas" before.
But it's just a GUI convention. It's not functionality.

I don't see any inconvenience with copy/paste
onto a new image. That makes it easy for me to
move around one or more images on a larger
background and operate on various layers until I'm
ready to merge them. That's a more flexible method
than canvas. If I want a text area I'll usually just
paint on a rectangle. But canvas seems fine for
adding a regular, defined border.

What you're doing with "canvas" is using a
conceptual device to imagine working on a picture
that's on a background plate. You then swap out
for a bigger background plate when desired.

That's actually an unnecessary device if you
think in terms of what the image really is and let
go of the irrelevant, concrete-world limitations
like easels and stretched canvas over a frame.

Raster graphics deals in rectangular bitmaps,
which are arrays of byte values defining pixels, with
the display of those pixels defined by width, height,
color-depth and orientation specs included with the
byte array. Thus, a 24-bit bitmap starting at upper
left with a size of 200x100 will mean that to create
the image the bytes are read in groups of 3 to
light 200 pixels across the screen. 600 bytes for
one line of pixels. Then the next line is read out.
It always gets down to simple numbers at some
point when you're using computers. (I don't know
whether Apple uses the term bitmap, but they're
doing it the same way if they're doing raster graphics.)

Before it's
painted to the screen it's just that byte array.
When you enlarge your canvas, what goes on
behind the scenes is that the system defines a
second device-independent bitmap and then paints
your existing bitmap onto it once you decide where
in the image you want your original pasted and
what color you want the background. It has
nothing to do with Windows and nothing to
do with any "old way of thinking". Look up raster
graphics. The same is true of things like brightening,
color pencil plugins, borders, etc. It's all mathematical
formulae applied to an array of byte values that
represent a pixel grid. So if you enlarge your
canvas what actually happens is that your byte
array gets enlarged and rebuilt. The software might
do that directly, or more likely it will do it by
defining a second bitmap and using system API to
paint one to the other, then retrieve the resulting
byte array. In that case the software is letting the
system do the grunt work. Either way, what goes
on underneath hasn't changed. Only the GUI changes.


mostly correct, but entirely irrelevant.

none of that helps in making better photos.

That's the trouble with only using Apple tools.
You end up thinking that a microwave is the way
cooking happens and when someone talks about
making spaghetti sauce from scratch on their stove
you think they're either ignorant or lying.


you were doing reasonably well up until that.
Ads