Paint Shop Pro support - Specified: how to get a uniform appearance

Author

Specified: how to get a uniform appearance

Siegfried Schoberth

2004-10-29, 7:14 am

Ok, Trev and Fred, here is my question corrected:

just as an example to demonstrate my point:
working with an image Im with the resolution of 72 dpi,
I did copy a small area A, created from this copy of A a new
image File Im2, and changed the resolution of Im2 from 72 dpi to
18 dpi. Then I copied this low resolved Im2 to A, which yielded
in a duplicate object O of the original area A. After resizing O to
the same size as A, O looked definitely less resolved than A and the
remainder of image Im, O2 appearing blurred and soft compared with
the remainder of the image.

How can I make the total image Im (that now contains the
low resolved O2) having the same look, such that one could not
tell that O2 was pasted into it?

Hope this clear ? (Guess not)

Siggi

"Siggi" <smusnmr@yahoo.com> schrieb im Newsbeitrag
news:2ud9l0F29h982U1@uni-berlin.de...
> Hi all,
>
> I bet this is the wrong espression but I am an absolute beginner in
digital
> photography and image processing, let alone PSP9 trial. Maybe my question
> could be answered by scavenging more thoroughly throug PSP9's help, but
> trial time is limited!
>
> QUESTION: after having put together different layers with different pixels
> per inch each, can I average or equalize the total pixels?
>
> E.g. I have a background of X pixels per inch, paste to it my person cut
out
> from a different picture with Y pixels per inch, and add another cutout
from
> another source of Z pixels per inch (Z being the lowest resolution). Is
> there a magic command "total pixels be an average of X, Y, Z!", or "total
> pixels be all Z!"?
>
> Otherwise the image would look kind of strange, with different image
> elements having different resolutions!?
>
> Thanks,
>
> Siggi
>
>

Kiri

2004-10-29, 7:14 am

Do you change the resolution to change the size? If so, don't. Resize dimensions using smart size.

Kiri

"Siegfried Schoberth" <smusnmr@yahoo.com> schreef in bericht news:2uebf8F29ucm7U1@uni-berlin.de...
> Ok, Trev and Fred, here is my question corrected:
>
> just as an example to demonstrate my point:
> working with an image Im with the resolution of 72 dpi,
> I did copy a small area A, created from this copy of A a new
> image File Im2, and changed the resolution of Im2 from 72 dpi to
> 18 dpi. Then I copied this low resolved Im2 to A, which yielded
> in a duplicate object O of the original area A. After resizing O to
> the same size as A, O looked definitely less resolved than A and the
> remainder of image Im, O2 appearing blurred and soft compared with
> the remainder of the image.
>
> How can I make the total image Im (that now contains the
> low resolved O2) having the same look, such that one could not
> tell that O2 was pasted into it?
>
> Hope this clear ? (Guess not)
>
> Siggi
>
> "Siggi" <smusnmr@yahoo.com> schrieb im Newsbeitrag
> news:2ud9l0F29h982U1@uni-berlin.de...
> digital
> out
> from
>
>

Fred Hiltz

2004-10-29, 12:14 pm

Siegfried Schoberth wrote:
> Ok, Trev and Fred, here is my question corrected:
>
> just as an example to demonstrate my point:
> working with an image Im with the resolution of 72 dpi,
> I did copy a small area A, created from this copy of A a new
> image File Im2, and changed the resolution of Im2 from 72 dpi to
> 18 dpi. Then I copied this low resolved Im2 to A, which yielded
> in a duplicate object O of the original area A. After resizing
> O to the same size as A, O looked definitely less resolved than A
> and the remainder of image Im, O2 appearing blurred and soft
> compared with the remainder of the image.
>
> How can I make the total image Im (that now contains the
> low resolved O2) having the same look, such that one could not
> tell that O2 was pasted into it?
>
> Hope this clear ? (Guess not)

No, I am afraid it is not clear. Please read again the posts from
Trev and me. We both began by trying to correct your
misunderstanding that makes it not clear. Pixels per inch (PPI),
sometimes misnamed dots per inch, DOES NOT MATTER. It comes into
play only when scanning or printing.

Now, please rewrite your question giving the sizes of the images and
the small copied area in pixels (PSP shows them at the right end of
the status bar). Then we can offer some suggestions.

I will offer a guess. When you enlarge an image a great deal, it
will look blurred and soft because the original detail is spread
over more pixels. For example, resizing by 200% multiplies the
number of pixels by four. PSP has to invent the content for three of
every four pixels by interpolating from neighboring values.

Try Adjust > Sharpness > Sharpen on the blurred layer to add the
illusion of detail.

You will get more specific suggestions when we know the sizes of the
images. Tell us about their intended use--web site, photo album,
gallery print, billboard--because that matters too.
--
Fred Hiltz, fhiltz at yahoo dot com

Kris Zaklika

2004-10-29, 12:14 pm

Siegfried Schoberth wrote:
>
> Ok, Trev and Fred, here is my question corrected:
>
> just as an example to demonstrate my point:
> working with an image Im with the resolution of 72 dpi,
> I did copy a small area A, created from this copy of A a new
> image File Im2, and changed the resolution of Im2 from 72 dpi to
> 18 dpi. Then I copied this low resolved Im2 to A, which yielded
> in a duplicate object O of the original area A. After resizing O to
> the same size as A, O looked definitely less resolved than A and the
> remainder of image Im, O2 appearing blurred and soft compared with
> the remainder of the image.

Siggi, let's bypass your experiments for the time being
because they are ill-defined in some respects. Instead
let us focus on what things mean and how they work.

Resolution has two meanings. In optics it refers to
resolving power, for instance the ability of a lens to
separate and distinguish two tiny objects next to each
other. High resolutions is therefore good. In image
editing, resolution normally means how many pixels you
print for every inch of paper. You need no more pixels
printed than your eye can make out because you can't
see anything smaller. At normal viewing distances this
turns out to be about 300 pixels per inch and no more.
More pixels in a print is pointless.

These two kinds of resolution are completely different
concepts but marketeers like to mix them up since it is
then easy to say "more is better", without stopping to
ask "more of what?". The basic information carrying unit
of a digital image is a colored dot called the pixel.
The size of a digital image is measured in pixels. (A
physical size like inches only comes into play when
you print.) Since the pixel is the information unit in
an image, if you want the image to contain more detail
or more information, you must have more pixels.
Changing the print resolution only determines how many
image pixels print on every inch of paper. It does not
affect your digital image at all.

There is one basic fact to accept about image size in
pixels and the information content of the image. If you
you are using many pixels at the time you capture or
create an image, the image can (potentially) contain a
lot of detail and information. However, if you capture
or create an image using few pixels, increasing the
number of pixels after the fact won't magically create
missing detail. It will only make things bigger. Here
are a couple of things to ponder. Suppose I have a
10 x 10 pixel image, for a total of 100 pixels. I now
increase its size to 20 x 20 pixels, for a total of
400 pixels. We started with 100 units of information
and we need an extra 300 from somewhere. These extra
300 pixels (or 75% of the larger image) are completely
invented from the original 100 pixels using a process
called resampling or interpolation. Or to put it another
way, if you started with a 1 x 1 pixel image, would
you really expect to be able to enlarge it to fill the
screen and see a complete scene on your screen? Much
the same ideas operate when you decrease the size of
an image. If you go from a 20 x 20 pixel image down
to a 10 x 10 pixel image, you have discarded 75% of
the original image information and the small image is
necessarily less detailed or of lower quality.

To return to your experiment, I'll note that you made
the usual error of not telling us whether you were
thinking about the sizes of your images in physical
inch units (wrong) or in pixel units (right). Go back
and do the experiment over again. At each point ask
yourself the question: in the last step, did I reduce
the number of pixels and did I throw away image
information? If at any point you are unsure about the
size in pixels of a layer do Edit > Copy and Edit >
Paste > Paste As New Image and either look at the
size of this new image in the status bar or do Image >
Image Information. You can close the image afterwards
without saving.

As an additional example, let's suppose you have an
image where a face is 100 pixels across. You paste
this face as a new layer into another image already
containing a face 1000 pixels across. The pasted face
will look fuzzy compared to the large face because
it is represented with fewer pixels and does not have
the same amount of information as the large face. The
only way to get the appearance of the faces to match
is to discard information from the large face until
it is as fuzzy as the small 100 pixel one. This is
because there is no way to magically make from nothing
the information that we would like to have but is not
present in the small face.

> How can I make the total image Im (that now contains the
> low resolved O2) having the same look, such that one could not
> tell that O2 was pasted into it?

Well, now you know the answer, don't you? Make the
content with many pixels have few pixels instead so it
matches the other content that already has few pixels.
Going the other way is not possible, though we would all
like to do that kind of magic. It would, for example, be
wonderful magic to apply to a bank balance :)

> Hope this clear ? (Guess not)

Here is another thing that will help you if you are not
used to thinking of image sizes in pixels. Suppose you have
several images you want to assemble into one image. For
every one of these images do Image > Resize. Uncheck Resample
Using... (this is critical) and type in some Resolution
value. The value is not important (use 200 pixels per inch
for example) but it is important that every image gets the
same resolution value. In this way, all your image sizes
in inches or cm will be proportional to the image sizes
in pixels. If you now line up all your images in the PSP
workspace at 100% zoom you can immediately see which are
going to look tiny after pasting and which will look
large when pasted into the new image. Any that are tiny,
and which you have to stretch with the Deform tool to
match other sizes in your composite image, will look
fuzzy.

> Siggi