HiYa Pete and Everyone!

On Jan 12, 7:47 am, pdimage <[email protected]> wrote:
> On 8/1/09 22:29, "aussieshepsrock" <[email protected]> wrote:
>
>
>
> > HiYa Pete and Everyone,
> >    My intended Scanning Methodology - Seperate from my Media Storage
> > Options - is something like this. I've only done a 50 image or so
> > 'test' run to sort out file size and physical process considerations
> > at this point. Some of this is based on some comparative tests of
> > various 'scanner driver' options.
>
> > TIFF with internal compression OFF
> > Photograph Fronts:
> > 600 DPI Resolution
> > 24 BIT Color Depth
> > Digital ICE OFF - It's mucking much more than it's fixing.
> > Unsharp Mask (in scanner software) at the High Setting because it
> > appears to be a well behaved and subtle implementation in my testing
> > up to this point.
>
> > Photograph Backs:
> > 300 DPI Resolution
> > 8 Bit Grey Scale
> > Unsharp Mask set to High
>
> > All images receive Levels Adjustments Set Manually. The sliders for
> > each color channel are tweaked individually so the sliders are just
> > past the Highest and Lowest Point on the Histogram Display for Each
> > Channel - ie the darkest/dimmest value is changed from zero to 9 if
> > the scans histogram shows no info below 10. I am cautious about
> > overpowering a particular channels level adjustments and making an
> > image look 'wierd'. I believe this is called manually clipping the
> > highlights and shadows.  I can find very little 'standards or good
> > practices' info via google or yahoo searches. This is just how I've
> > learned to go about getting good scan results since my first encounter
> > with a grayscale only flatbed back in the early nineties!

Let me dive in. :-)

"Well you seem to have quite a job on so here's a few tips. "

It is going to be a bit of a slog. It's the most photo scans I've ever
done at once, although I have worked a couple times at jobs where high-
speed document scanning was a part of what I had to do. A rather
different beast that only in a narrow sense is the same as scanning
photo's. :-)

"The optical
> resolution of your scanner - say 600x600ppi for this purpose - is the limit
> for original capture - higher resolutions like 9600x9600ppi can only be
> provided by interpolation ... "

Your input is greatly appreciated, but I'm fully up on the Optical vs
Interpolated with Scanners. I have actually re-discovered the
knowledge that my Epson 4870 PHOTO Perfection scanner only does
Transparencies & Film at 4800 dpi! Document/Reflective scans top out
at a respectable 1200x1200 true optical resolution. If memory serves,
it's because a different lens and a narrower scan path is used for
film that gives the higher resolution, but don't quote me on it.

"Unsharp masking is better
> done selectively per image in Pshop if you have it as ramping the edges to
> provide a sharper image can produce artifacts."

You are quite right about the Unsharp Masking in Photoshop being an
incredibly better tool than the ones in scanning software itself.
However, when the autoexposure system isn't used in the epson driver
and it's harsh restoration and autopilot systems are avoided, the
Unsharp set up in the Epson Scan has a very light touch in the 1200
dpi scan tests I've done. As a matter of fact, it's about a quarter of
the strength my Photoshop Experience tells me that  would be necessary
to negatively effect an image's quality in any way. There is nothing
dramatic about the differences between ON or OFF, it's there,
measurable, but subtle.

">     Levels is a destructive process which affects the entire image
- if you
> move the black point or white point by 10% you are not only disposing of 25
> channel levels from each colour - you are creating 25 new ones for each
> colour as each channel must have 256 levels. I use the non destructive
> curves if at all possible and reserve level adjustment for very poor low key
> originals."

I only have personal experience to draw upon because authoritative
information about this has been difficult to find, but I have doubts
as to your statement's applicability to how I edit the levels and how
I carefully monitor my levels adjustments and their effect regarding
each level and how the levels act in conjunction to generate the whole
image. I'll try to find a better way to write how I edit levels, how I
approach them as a photo person, and what makes my methods seem to be
'non destructive' from my perspective as a photographer and someone
trying to be faithful to what is or isn't in a scan.

">     Highest resolution? I would say around the 200/300ppi mark
unless they
> are earmarked for substantial enlargement. The human eye can only resolve
> around 180 levels, b/w newspapers print photos at around 80 lines of dots
> per inch (the cheap paper limits the res) and we see them well as images.
> Glossy colour mags 133/150/175 lines of dots per inch and they look very
> acceptable even though the CMYK space is smaller than RGB. Computer monitors
> are limited by dot pitch and can only manage hardware res around 90ppi so
> any res above this is a software representation - tv's are worse with poorer
> dot pitch."

This information is extremely valid, but I have a sense my thoughts on
resolution and your's aren't entirely describing the same things. I
wish to reach into each print with the scanner and pull out the
maximum amount of detail I can so that someone like me somewhere in
the future has access to everything the print has to offer. That is
the Scanning Resolution. I'm settling on 1200dpi, but you are quite
correct that any visual image outputted at that extreme high density
wouldn't be accesible with the human eye when outputted, but that is
not an issue in this case because while my original photo might be 2x3
in and scanned at 1200 dpi (a lavish resolution setting in most
cases), the output of the file can be a 'wasteful' 2x3 in at 300pi or
a lavish 8x12 in at 300dpi because the Input DPI is being used to
generate the Output DPI. I have also found that excessively over
scanned images allow for surprisingly powerful damage removing methods
in Photoshop which leave the 'underlying' low resolution photo
relatively untouched.

">     My archive of high res images is stored at 360ppi, medium res
at 180ppi"

I think there is also the disconnect between the 'historical image
archive' I'm contemplating and the 'working' image archive you seem to
be describing because the ppi's are based upon the capabilities of
your output source. I'm most concerned with archiving the maximum
image capability of my source materials. The connection my project has
to output methodologies is indirect at best. It will be a resource of
source materials that on screen viewing, printing, or publishing, can
then be derived from.



">     Finally I would add the fact that re-resing is always possible
with a
> good image editor - a 200/300ppi digital image can be easily upped to
> 1200ppi without problems. The image editor is simply doing what the scanner
> does over and above it's optical resolution - interpolation - but probably
> doing it much better in the case of Pshop."

Accurate info, but not directly applicable to my methods and goals.
Yes, photoshop is the best image upscaler around and is quite usable
when wielded judiciously. I have most assuredly used it, especially
when making 8x10's from 4x6 originals I need bigger prints off of.
However, If I scan an image at 1200dpi and someone in the background
turns out to be important to someone years down the road, there will
be lots of pixels to fish out the best image that's possible. If I
scanned it at 300dpi, there is no way to interpolate the missing
900dpi of information, the result would just be a really big high
resolution file describing low resolution information. The intention
of this scanning archive is to stand in for missing originals should
they be lost or destroyed and to secondarily give a lot of people
access to the photos that sitting in storage somewhere would not give
them.

>
> Pete

I sincerely and dearly appreciate your input pete. I also would
appreciate whatever  input on Levels and such you might have or point
me to. At no point do I force harsh adjustments, cause color oddities
or saturation issues, and I definitely don't place a slider so it cuts
off any part of the actual values the scanner picked up. (except for
whites in dusts or scratches). Archive and Museum sources have noted
the dangers of excessive level and curves adjustments to the integrity
of the scan itself and to the actual usability of images edited in
such a manner when it comes to differently calibrated monitors,
various current or future OS's, and especially when outputting the
file to hardcopy. The resulting edits as applied to the image might
make it visually quite wonderful and compelling, but at the expense of
the integrity of the image file and it's 'true' linkage to the source
material and it's visual content.
--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed Low End Mac's G3-5 List, a 
group for those using G3, G4, and G5 desktop Macs - with a particular focus on 
Power Macs.
The list FAQ is at http://lowendmac.com/lists/g-list.shtml and our netiquette 
guide is at http://www.lowendmac.com/lists/netiquette.shtml
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to 
[email protected]
For more options, visit this group at 
http://groups.google.com/group/g3-5-list?hl=en
Low End Mac RSS feed at feed://lowendmac.com/feed.xml
-~----------~----~----~----~------~----~------~--~---

Reply via email to