Scanning colour slides using VueScan and a flatbed Epson 1650 Photo

I have been using VueScan for a few months and I am very happy with

it because the quality of the scans I get is far better than what I

was able to get using the Epson TWAIN. <br>However, you may see in

the thread of comments on one of my recent photos (<A

HREF="http://www.photo.net/photo/1253909"

target="blank">Imperial Conerstone</A>)that the sharpness of my

posted files is still questioned. In that thread my present settings

are critized on the basis of what I mention under the technical

details of all my posted scanned slides: <br>QUOTE<br>

Scanned at 1650 dpi using VueScan software, at 48 bits depth, 2 scans

passes, to a 8 Mb TIFF file<br>UNQUOTE. <br>These are the settings I

use in VueScan: <br><u>DEVICE</u> <br>Media Type: Image <br>Bits pp:

48 bits RGB <br>Scan resolution: 1600 dpi (maximum optical resolution

of my scanner) <br>Auto focus: Scan <br>Number of passes: 2

<br><u>COLOR</u> <br>White Balance, <br>Gamma, <br>Device RGB <br>and

ICC Profile <br><u>FILES</u><br>TIFF file type: 24 bit RGB <br>TIFF

compression on <p>Would anyone advice me on what to do in order to

get sharper scans? I know, already, afast solution: selling the

flatbed scanner in e-bay and buying the new Minolta Dimage Dual Scan

III<br>:-)

In my experience with flatbed scanners, it is extremely difficult to

get them to register properly for multiple passes (too much banging

going on). That would be my guess, everything else being equal. I

agree that slides would benefit from multiple passes. I just haven't

had good luck scanning slides on flatbeds and being an amateur use

C-41 color negative materials. I would also scan without TIFF

compression if you can but I still think subtle mis-registration of

multiple passes is the most likely reason. My tutorial is getting

outdated but you may enjoy it:

http://homepage.mac.com/onelucent/VS/vsm.html GOOD

PS: Great image-making that cornerstone appear as a globe. GOOD

LUCK>>>

Thanks Steve.

I take your advice of not using the TIFF compression any more. Do you think that the output to a 48 bits TIFF file would be better than to a 24 bits one?

<u>SETTINGS and WORKFLOW:</u><br>These are the settings which have been agreed upon after a discussion in groups.google.com/comp.periphs.scanner and contributions by Steve Murray:

<p><u>DEVICE</u>

<brMedia Type: Image

<br> Bits pp: 48 bits RGB

<br> Scan resolution: 1600 dpi

<br Auto focus: Scan

<br> Number of passes: 1 (to avoid possible misalignments if performing more tha 1 scan pass)

<p><u>COLOR</u>

<br> White Balance,

<br> Gamma,

<br> Device RGB

<br> and ICC Profile

<p><u>FILES</u>

<br> TIFF file type: 48 bit RGB

<br> TIFF compression on <p><u>WORKFLOW:</u><br>After importing scanned file into Photoshop: Very useful to perform sharpening in two stages. <br>Apply a

light USM sharpening to the full resolution scan images to clean up the apparent defocusing caused by aliasing during the scan process, <br>Parameters like 70%, 1.1 pixels, 2 threshold work well.

<br>Levels, <br>Curves, <br>Basic color corrections with Selective Colours, <br>Cropping, <br>Selective masking and

localized Curves/Levels/color correction editing<br>Resize for pinting or screen display output<br>Clone off dust and scratches<br>

Finish up with USM tuning for the resulting final images with the reccomendation not to be too timid when

using UM as the last step on a soft jpg. Typically start at: Amount:

500% , Radius 0.2, Threshold 0. If it looks too extreme you can back

off some on the Amount.