On Mon, 4 May 1998 09:20:17 +0200, "Johannes Hubert"
<jhu### [at] algonetse> wrote:

>Hi Tony,

Hi Joh! You are all wrong! Please, don't beat me!
But I beat the drum, next TreeDesigner will have an L-system inbuilt!
That's amazing news!

>
>I think we more or less say the same things with different words.
>
>And while I tried to explain in an abstract way why images as such don't
>have any DPI size (and why it is therefor futile to ask POV-Ray to render a
>certain DPI size), I did not discuss anything concerning the quality of the
>printed output. It is difficult enough to explain the very messed up concept
>of screen-resoultion and printing-resolution in a simple news-group post so
>I didn't want to start into explaining why a 600x600 pixel image looks
>better when printed on a 2x2 inch area (300DPI) than on a 1x1 inc area
>(600DPI). For that I would have to start explaining how color-printers work,
>how they mix their colors etc.

Here we go! And there comes in the LPI, that's lines per inch, and
that's the value for the printer, how many colors he can actually mix
at the available dpi.

>
>In the light of the original post it was probably a good idea of you to add
>the quality-concerned view of the problem.

And that's what he meant there, correct me, if I am wrong, Tony.
On a bubblejet printer, the effect of the rasterized print is achieved
by smaller or bigger points of the color 'bubbled'. Now, if one would
print a 24bit color image delivered to the printer at its maximum
resolution, the space available between the dots would only allow
for no shading or, with a bad printer driver that does not care, a
waste of ink, as there would not be any distinguishable image, just a
sea of ink. So what the windows driver does, for instance, he
calculates down, I do not now to how many lpi the given bitmap as per
the user settings, contrast prefered, saturation prefered etc.
Some years ago a 360dpi bublejet at a setting of the maximum 100 lpi
could deliver 9 shades of one color. So that would make a realworld
3bit COLOR resolution print. At a setting of 60 lpi this increased to
25 shades, which made 4 and a half bits and so forth. Now, it is
clear, there can be different methods for a driver, to treat the given
input resolution to meet the given demand for quality. He could blow
up, lets say a 800x600pixel input bitmap to a 3200x2400 dots print and
has 3 of 4 dots as space for dot variation available, he really
overprints his dots then also, compared to a thin b/w print. He
divides his print resolution compared to the b/w print the same time
by for, this would be a 80 lpi print for a 360 dpi printer. For a 1440
dpi printer, it would be a 360 lpi color print, I don't think, they do
this, they will give more space, which will give a higher color
resolution.
So, I have not checked them really out, but take a 21" monitor
at a screen resolution of 1024x768, has ~60 pixels/inch. With the
modern 1440 dpi color printers, if they really do not call some
dithering method resolution, and as the color resolution roughly
quadruples with halfing the lpi resolution, a ~1 to 1 pixel/dot print,
would give (60 lpi) a color print of 225 shades of one color, 7 three
quarter bits, almost true color.
Don't ask me, how they call it, enlargement or whatever, but that's
what the driver and printer have to do to master color print.
I know other platforms, where ordinary printer drivers let set you lpi
directly, but that's another thread.
Of course, there are also different dithering methods involved,
even the raster angle, but that's to much now. And, of course, you
know it anyway, Johannes.
Just bounced into the thread last.

>Tony Vigil wrote in message <354CC40F.3F47F503@gte.net>...
>>I am sorry if I sounded a bit harsh.  My bad!
>
>
>Apology accepted, though I didn't really feel the need to get one :-) I
>wasn't that pissed... (you can always tell I'm not, when I litter my
>postings with smileys... :-))
>
>Johannes.
>
>

Hei! Wanted to make my first smiley ever, now my Free Agent shows it
blue!

:-))
,
,:-)), aha!

steff

Post a reply to this message