Re: [APML] Matt's S&T Cover

[Craig Utter <cmutter@bellatlantic.net>]

Glenn,Robert,et al

I just wanted to set a few things straight, The S&T cover is printed with a 150 line screen, although we follow the 1.5x rule, images for the cover have to have a minimum resolution of 250 dpi. The images for the interior of the magazine are printed with a 133 line screen which allows a minimum resolution of 200 dpi(which I only let pass if the image is printing small).
2184x1472 at 250dpi only results in an image of about 6"x9" which isn't even close to filling a cover and your right, we usually do crop in.That's the biggest limitation for using a CCD image on the cover, because the image is physically smaller than 35mm film it's not always possible to get it to the size needed, it can be done but it's not always easy.
Now if SBIG starts using Kodak's new 14 megapixel CCD that has the same dimensions as 35mm film we might have a whole different story!

Craig Michael Utter

"Glenn E. Mitchell II" wrote:

Robert:

First, the ST10 is not a typical CCD camera, nor are its images typical
in resolution. For every ST10, there are probably 50 ST7's and ST8's out
there. So, for the few 2184 x 1472 images out there, there are lots and
lots of images at 765 x 510 and 1530 x 1020.

With film, actual resolution captured, grain of the film, and grain of
the paper determines how well an image will tolerate enlargement.
Digital images are different from photographic enlargement, however.

Your unbinned ST10 image gives you sufficient resolution at 8"x10" for
the 175 screen Sean Walker claims. Barely. No cropping. Anything else
requires that you increase the resolution in the image, be that with
something like Fractal Graphics or the bicubic interpolation in
PhotoShop. I'm sure you know, that there's no free lunch when you do
that. You introduce noise and soften the image. You can sharpen the
enlargement, and then you introduce more noise.

ST8 and ST7 images, which are far more common and typical, need
considerable digital enlargement to get to 180 screen for an 8x10 image.
The unbinned ST8 has sufficient information for just 125 screen at 8x10.
The unbinned ST7 can't even muster up 65 dpi, less than the resolution
of the typical monitor when you try to print them as an 8x10 picture.

No matter how you slice it, your ST10 provides 3.2 million pixels of
information. Run of the mill magazines use 175 or 180 screen. That's not
what most people would consider photo quality. Even high quality
magazines, like National Geographic and Audobon, are not satisfied with
175 screen. They use 300 or 240. That leaves a 3 mega-pixel image enough
print resolution for a 4"x6". Period.

Now, you can get some lovely CCD images with an ST10. Same with an ST7
and ST8 (if you don’t push the digital enlargement too far). That's
owing to the small pixel size of the CCD imager in the ST10. At 6.8
microns each, your pixels are capable of resolving more detail than the
9 micron pixels in the ST7. So, not only do you have more pixels to
capture an image, each pixel is tinier. That's exactly like using a
finer grain film.

I agree, you can digitally enlarge an ST10 image more than an ST7 or ST8
image before you notice noise. If you think I argued otherwise, you
misunderstood my point.

I am not deprecating CCD images in any way. I have been using a digital
SLR for a couple of years for nature photography, landscapes, and street
photography. I loved my D30. I adore my D60. I even use an STV and a
three-color wheel for imaging upon occasion. So, don't read my comments
as a knock on CCD imaging. That's not the case.

What I am saying is that it is still common for magazine publishers to
refuse digital images AND while that is partly a prejudice based on
outdated perceptions, there is still some merit in that position. If you
want photo quality images and you want headroom for cropping, digital
SLRs are just now getting to that point for 8x10, 11x14, and comparable
size images.

3.2 million pixels needs considerable digital enlargement to go beyond
4x6 and print at near photo quality. When digital images are pushed that
far, the noise and loss of sharpness is noticeable, and you don't need a
loupe to see it.

As to you comments about film v. CCD, that's a red herring I'll decline
to chase.

Cheers,

Mitch

-----Original Message-----
From: owner-astro-photo@seds.org [mailto:owner-astro-photo@seds.org] On
Behalf Of Robert Gendler
Sent: Sunday, October 27, 2002 6:56 PM
To: astro-photo@seds.org
Subject: Re: [APML] Matt's S&T Cover

Mitch,
Look at the November 2001 S&T cover and you will see
a CCD image (mine, NGC 6888), taken with an ST10. The
resolution is fine and the image fills the entire cover. Matter of fact
I would like to see a comparable film image of the same object at the
same size. I've said this before but it doesn't seem to register. Its
not
the number of pixels in an image but the actual resolution captured
which determines how it will tolerate enlargement. I think Matts
statement
reflects more his negative experience with CCD which was due to
technical
issues.
Rob Gendler
Email: robgendler@att.net
Web site: http://www.robgendlerastropics.com/
----- Original Message -----
From: "Glenn E. Mitchell II" <mitchell@dot-com-group.com>
To: <astro-photo@seds.org>
Sent: Sunday, October 27, 2002 6:32 PM
Subject: RE: [APML] Matt's S&T Cover

> S&T is far from unique in their preference for film compared to
digital
> images. Especially for a full-size cover shot.
>
> Lots of magazines will not even consider digital images. I run into
this
> with nature photography all the time. Some of it is prejudice, which
is
> beginning to erode. Part of it is a desire for images that can be
> cropped, still screen at 300 or 240 dpi, and be larger than a 4"x5".
>
> For a cover shot, you need something like a 300 or 240 screen at
8"x10".
> The typical CCD image doesn't approach 300 resolution much beyond
about
> 2"x3". Especially if the image was binned.
>
> S&T has been willing to display CCD images inside the magazine, with
> appropriate sizes of a few square inches. It looks like they use 180
> screen for those images, which provides reasonable quality.
>
> Mitch
>
> Glenn E. Mitchell II, Ph.D. (aka Mitch)
> 10" f/10 Meade LX-200 GPS (Big Blue) s/n 01088
> Supercharged at ASO by Dr. Clay Sherrod, # 2314-2010
> N 30° 30' 49.2" / W 084° 20' 13.8"
>
>
>