Photo Enlargement
RKII
baseball team picture. It's the size of a baseball card itself about, I
wanted to know if there is ANY way I can blow the pic up bigger without it
getting all pixelated and then it looks like crap. Any solutions would
greatly be appreciated. :) Thanks in advance! :)
RKII
TacitR
>baseball team picture. It's the size of a baseball card itself about, I
>wanted to know if there is ANY way I can blow the pic up bigger without
>it getting all pixelated and then it looks like crap.
Nothing can enlarge the image without degredation. The computer cannot make up
detail that was lost in the original scan.
In general, enlargements of greater than 120-140% will become highly noticable
in Photoshop; your best bet is to re-scan the original at a higher resolution.
Sorry...
------
Onyx, the game of sexual exploration; Xero, the industrial magazine
of art, fiction and photography; and online photo gallery--all at
http://www.xeromag.com/franklin.html
zdiver
point your silly little browser over to:
http://www.genuinefractals.com
and download the free demo.
You'll then see if you like the quality.
If you don't want to pay for the program,
contact me and we'll see if we can come
to an arrangment.
> Nothing can enlarge the image without degredation. The computer cannot make up detail that was lost in the original scan.
This is true and I'm not arguing that it's magic, but I'm silly
impressed with the ability to take a 900K jpeg at 1600 x 1200 from a
nikon digital and blow it up to 17" x 25" at photo quality.
> In general, enlargements of greater than 120-140% will become highly noticable
> in Photoshop;
This is true. Via interpolation, expect to not like your results.
your best bet is to re-scan the original at a higher resolution.
This is also true. But depending on your ability to scan at high res,
you could output your "team picture" to anywhere from 8.5 x 11 to 17 x
44. (Bigger if you want to go to Kinko's or some other large-format
output place and pay the big bucks).
>
> Sorry...
Don't be - you didn't know.
Adam Yalonetsky
wrote:
>Hey all...I just got done practicing the rubber stamp tool on a classic
>baseball team picture. It's the size of a baseball card itself about, I
>wanted to know if there is ANY way I can blow the pic up bigger without it
>getting all pixelated and then it looks like crap. Any solutions would
>greatly be appreciated. :) Thanks in advance! :)
>
>RKII
>
How much larger are you wanting to go? You can get away with double
usually. Enlarge it, then try experimenting with blur and then
sharpen filters.
Adam Yalonetsky
ad...@interlog.com
br...@iterated.com
zdiver <zdi...@sticky.buds.com> wrote:
> though shalt now be amazed, young grasshopper.
> point your silly little browser over to:
>
> http://www.genuinefractals.com
>
> and download the free demo.
> > Nothing can enlarge the image without degredation. The
> > computer cannot make up detail that was lost in the
> > original scan.
>
> This is true and I'm not arguing that it's magic, but I'm silly
> impressed with the ability to take a 900K jpeg at 1600 x 1200 from a
> nikon digital and blow it up to 17" x 25" at photo quality.
Yes. While the technology in Genuine Fractals can't "make up detail
that was lost in the scan", it does an impressive job of accurately
rendering the existing detail in an image to any desired output
resolution. Assuming adequate source detail -- I'm talking 800% or
more.
People tend to confuse large output resolution requirements with
requisite levels of detail. For example, most people would agree that
a good 35mm slide contains a fair amount of detail. But even if we're
talking very fine-grain film (ISO 25-64) there's not much point in
scanning that 35mm slide at more than about 3000 pixels per inch. This
yields an RGB image of about 36MB.
If you then need 600MB of pixels for a large output print job, I submit
that the fractal software can do a better job of rendering that 600MB
(from the 36MB representing accurate image detail) than any scanner can
from the original slide.
-- brian
Sent via Deja.com http://www.deja.com/
Before you buy.
zdiver
I don't have a crack for this - I bought it.
I was only offering to do the blowup work for a fee.
-Z
Mike K.
for doubling an image's pixels:
1. Layer > Duplicate Layer
2. Filter > Noise > Despeckle
3. Play with the Duplicate Layers mode and opacity
4. Image > Image Size > grow the image twice-fold
5. Layer > Duplicate Layer
6. Filter > Blur > Smart Blur; play with the settings
7. Play with the Duplicate Layer's mode and opacity
8. Layer > Duplicate Layer
9. Filter > Sharpen > Unsharpen Mask; play with the settings
10. Play with the Duplicate Layer's mode and opacity
Adam Yalonetsky <ad...@interlog.com> wrote in message news:37f91f2c....@news.psi.ca...
Patrick Brinton
need the same kind of resolution in the enlarged image as they had in
the smaller one. This is generally not true. To take an extreme
example, billboards are often made from 35mm slides. There is no
special interpolation happening, and yet the result looks sharp. This
is because of the viewing distance; when the apparant size of the
picture is the same as an 8 x 10 print held at arm's length, then the
apparant sharpness will be the same too. (If that is unclear, imagine
holding the small print out in front of you and moving with respect to
the billboard until the pictures seem to be the same size; obviously
you will be some distance from the billboard.)
On the other hand if you stood at arm's length from the billboard you
would not even be able to make out an image at all, just a bunch of big
blurry dots. If the billboard were rendered at (say) 1200 dpi, and we
could magically interpolate the pixels accurately, the apparent
resolution at normal viewing distance would not improve very much. You
simply cannot make out all that detail at that distance.
The same effect works at smaller enlargements too. The normal viewing
distance for a mural is much greater than for a 16 x 20 print. Another
way to confirm this effect is to project a slide; at the distance you
view it the resolution looks fine. If you go right up to the screen you
see how poor it really is. Again, there is no interpolation going on.
So, folks, go ahead and blow up your prints without fear; they will
still look good when viewed in a normal fashion.
Patrick Brinton
pbri...@wco.com
http://www.unusualgraphics.com
images and outrageous opinions. Come visit.
Jan Steinman -- jan AT bytesmiths DOT com [remove .gov]
(TacitR) wrote:
> Nothing can enlarge the image without degredation. The computer cannot make up
> detail that was lost in the original scan.
But, but, but, I was watching "Profiler" the other night, and they got
some video from an airport security camera (gee, that had to be ~250 line
video, right?), and they blew it up and did some computer enhancement, and
you could read the airline ticket the guy had in his hand? You mean that's
just Hollywood? :-)
There are two things I am unwilling to suspend disbelief for at the
movies: 1) sparks coming out of electronic equipment (ever heard of
fuses?), and 2) infinite zooming of limited bandwidth source material (did
someone repeal Shannon and Nyquist?)
Re: Genuine Fractals: in general, I like what it does better than what
bicubic interpolation does, but I did a blind test. I took about a 100
pixel square portion of an 800x600 digital camera image, and up-sampled it
ten times, using both Photoshop's bicubic, and GF. I then printed it, and
showed it around. Although people said the GF one looked more "natural,"
and less "computer-like," they unanimously preferred the bicubic
interpolation.
In my judgement, the GF one had more contrast, but that seemed to be
because it "threw away" more detail. The bicubic interpolation had more
detail, but looked pixellated and blocky. The bicubic looked like a
computer blow-up, the GF looked like abstract art. These were people's
faces, which might also explain the unanimous preference. With nature
scenes, GF might have gotten more votes.
In short, Shannon has not been repealed, and There Is No Free lunch. My
simple test seems to indicate GF is more hype than reality, although I
still prefer the look it gives to moderate -- two to four times --
upsampling.
--
: Jan Steinman -- Jan AT Bytesmiths DOT com
: Bytesmiths -- digital artistry <http://www.bytesmiths.com/Art_Gallery>
: +1 503 635 3229
z_diver
I have also done similar tests to try to show people the difference in
image quality. These small samples that I have taken and then blown up
via interpolation and GF don't seem to do the same thing. For one -
they are just that, small samples. I usually am not amazed until I
output something to 17 x 22, then stand back in awe of the clarity in
the image. Try this test: Take an image of 1600 x 1200 resolution (or
whatever) and blow it up to poster-size and print it. I don't care what
it looks like on your screen, we're talking printed output here. I've
done the same with most of the files at my disposal, and I'll take the
GF copy every time.
Z