Re: 4mp doubled = 16mp!?

Today Thomas T. Veldhouse commented courteously on the
subject at hand

All Things Mopar <nunofyour@xxxxxxxx> wrote:

This isn't what I'm talking about. A highly detailed,
daylight image will have more information that goes into a
"good" print than, say, a night shot which has a generally
narrow histogram. For the latter, I'd guess that banding
would be a bigger issue than pixelation.

I know what you were talking about. You were saying that
the camera you used had such horrible jpeg compression
issues that a higher MP sensor would not benefit the final
image. We have the Kodak Easyshare (3.1 or 3.2MP) and know
people who have the 2MP and 4PM versions. The highest
quality JPEG settings did indeed make more detailed images
with the higher MP.

Little Kodaks use terms like "good", "better", and "best" to
me 1 MP, 2 MP, 3 MP, but generally /not/ what "good" cameras
refer to for JPEG such as "basic", "normal", and "fine". So,
my 2 Kodaks have no ability whatsoever to alter compression,
just mega pixels. But, my Rebel XT gives me 3 resolutions (2,
4, 8 MP) and "normal" and "fine" compression.

I really don't know why you think
otherwise.

I think otherwise because JPEG has no-thing to do with fine
detail and sharpness except that it can destroy it. But, I
assume you are already shooting with the lowest compression
possible in your camera and using low compression when you
save from your fav graphics editor.

Either way, the image resolution was indeed
higher with the higher MP, reguardless of what the image
processing did to it after the exposure was captured. JPEG
issues like that are NOT anything common to be sure and not
worthy of consideration with the average purchase decision.

You're still mixing these two issues up. One is an example of
little Kodak P&S's that over compress and the other is the
need for image "information" that /you/ seem to require to do
large prints from as large an image in mega pixels as you can
afford. For the latter, I was saying that a night scene is
"simpler" than a highly detailed daylight shot, say of a car
or architecture or still life.

Lower JPEG compression doesn't "create" detail, it simply
doesn't destroy what is already there by using its lossy
algoritithm to throw out pixels. As I said earlier, the main
issue with JPEG is the "mushing" of fine detail through the
creation of JPEG artifacts, which most people define as
streaks, blobs, spots, and other easily visible areas of large
expanses of color or in areas of fine detail.

JPEG compression /is/ a very important purchasing decision
criterion. If it ain't got it, you ain't gonna get decent
images. Period. That's not an issue with a Rebel or D70 but is
for cheap, high MP P&S.

What do you mean by "artifacts"? The common use of the
term is for image damage due to JPEG over compression and
has nothing to do with inherent image quality put out by
your D70.

I meant pixellation.

That isn't what "artifact" means. See above. Pixelation occurs
from trying to force a small number of pixels into a large
area linearly, such as using too small an image to print -
which is /your/ point.

So, if your definition of image degradation visible in a print
is pixelation, then, yes, more mega pixels /should/ yield less
damage. I say "should" because all those other pesky factors
such as lens, image processing, lighting, your skill as a
photographer, and what you do post-processing the image are
just as important, if not more so, than a simple increase in
mega pixels. Again, "all other things being equal, but they
seldom are."

Back to printing, most/all printer drivers do simple pixel
resizes to "create" pixels to cover tha paper size a graphics
app and user specify, no matter what the claimed DPIs at
various quality levels in printers you have to choose from. If
the information isn't in the pixels sent to the printer, it
will print /something/, which /may/ result in pixelation. It
can also result in other problems, such as banding.

The rest of your message seemed directed at the D70. I was
indicating that I own a D70, and not the OP.

I didn't accurately track who said what, so I apologize. The
comments are flying in this thread with several misconceptions
along the way.

--
ATM, aka Jerry

"Whether You Think You CAN Or CAN'T, You're Right." ? Henry
Ford
.

Relevant Pages

  • Re: Sharpness in Tiff Scans
    ... Tiff vs. JPEG has nothing to do with resolution (number of pixels). ... Nor, necessarily, with sharpness as long as you don't get carried away in the degree of compression in the JPEG format. ...
    (comp.periphs.scanners)
  • Re: JPG compression - yet another question !
    ... That is 4,915,200 pixels which would be about 15MB for the in- ... the space besides just the pixels, the header info, and EXIF. ... However, once JPEG does it's magic, all bets are off. ... that it just isn't possible to recover the compression factor ...
    (rec.photo.digital)
  • Re: Avoiding pirating of DVD images
    ... of horizontal and vertical pixel density in digital images? ... many folks know the difference between pixels and dots. ... And, JPEG uses a unitless 1-100 scale, although several ... percentage compression because kids get taught in junior high that ...
    (rec.photo.digital)
  • Re: JPG compression - yet another question !
    ... That is 4,915,200 pixels which would be about 15MB for the in- ... the space besides just the pixels, the header info, and EXIF. ... However, once JPEG does it's magic, all bets are off. ... that it just isn't possible to recover the compression factor ...
    (rec.photo.digital)
  • Re: OT: Early Fall
    ... Since you are not processing the jpg and then saving the file, the compression and decompression will not introduce noise into the photo. ... TIFF pictures store a single raster image at any color depth. ... Not all TIFF files are compatible with all programs that support the baseline TIFF standard. ... JPEG pictures store a single raster image in 24-bit color. ...
    (rec.boats)