Tranny scanning help please........?

Question

im using SilverFast ver 6.4 software and a 7200dpi dedicated scanner.

Can you please give me your "procedure"; as in the order you do what, dust/scratch reduction etc.

Any feedback would be greatly apprieciated, im up on res/input/output etc, I would just like some proceedures you find work well..........seamless and Dr many thanks for you help so far......if you want to email me anything please do.
\
i scan to tiff, at 48bit hdr, gamma 2.0,

any other info ask and i will edit it in


Thanks, to all

Answer 1

Antoni,

Good to hear from you again-I haven't heard from you in a while. Love the new avatar.



I have one of those radioactive camel's hair brushes that I keep right by my scanner. I use it to dust off every piece of film that goes into the scanner on both sides right before I actually put it in the scanner. You'd be surprised at how far this goes with eliminating dust.



After scanning, I go into Photoshop and look over the scan at about 50% to check for dust, etc. Anything that I find gets touched up with the spot healing brush.



I then look at the scan as a whole, and look back to the original slide sitting on the light box next to my computer. If they look pretty much the same color wise, I sharpen it up a little bit(I like 150% with a 3.0 pixel radius), and call it a day.



If they don't look the same, I typically go to adjustments-match color-neutralize and see what happens. If I'm still not happy, I go in and play with the levels and curves until I am. This can be a long process-I'll sometimes spend an hour or two playing with the colors in a scan to get them to where I think they should be.



I don't, as a general rule, do anything more than this, as post-exposure manipulation isn't my thing.



By the way, Dr. Sam, using the in-scanner dust reduction doesn't reduce resolution. Basically, what Digital ICE does is scan the film with a broadband IR lamp a second time. Big chunks of dust are opaque to opaque to IR, while the dye clouds in film are transparent or translucent to IR. Dust lights up on the IR scan sort of light bones(or teeth :)) do on an X-ray, making the scan software able to recognise the dust and blur it out.



This is also why you can't use ICE with traditional black and white films, or with Kodachrome. They both contain elemental silver, which is also opaque to IR.

Answer 2

Personally, I don't use the scanner for dust and scratch reduction. Think about it. What this does it to let you choose a dimension beyond which you do not care about the detail. You still have this option in Photoshop, so why let the scanner destroy your image quality for you?



I prefer to eliminate the worst offenders (like 3-5 pixels) "by hand" - one-by-one - with the spot healing brush set to 7-8-9 pixels, if possible, or the clone tool if it is not possible to use the spot healing brush. YES, it is tedious, but it lets you keep the high resolution your scanner is capable of giving you.



If you are doing a 20x30 print, you will be glad you did it this way. If you are only doing a 5x7 print, you can get rid of the worst (big) offenders, resize, and use Photoshop "Dust and Scratches" and play with it to find a balance between getting rid of the junk and keeping the good stuff.



See these and see if I put any gems of wisdom in the comments.



http://www.flickr.com/photos/samfeinstein/sets/72157601316601148/



~~~~~~~~~~

ADDITIONAL

~~~~~~~~~~



Ben, I guess I don't mean a loss of resolution, but rather a loss of detail. I know that the dust reduction looks for sharp contrast at the small pixel level and blends things in. The problem is, you lose eyelashes, for instance, and other things with fine, fine detail that are only 1,2,3 pixels themselves.



I do this in Photoshop (Elements) as you can always undo it if it goes to far. If you do it in the scan itself, you can't bring back the detail without doing a new scan.



I don't use nearly that much sharpening, by the way. I usually use 50% at 1 pixel radius, 1 threshold level. Sometimes I go as far as 100% at 1.5 pixels, but that's usually it, unless I am trying to save (someone else's) image. (haha)



~~~~~

MORE

~~~~~



Ben, I'm going to try that dust bit with the scanner next time. I didn't realize how it worked. Thanks.

Answer 3

I can't help you with the scanning, Antoni, except to recommend that you keep everything as clean as you can.



In postprocessing:



1. I take care of any contrast, color and saturation adjustments first.



2. Then I examine the image for those areas that are blurred or out of focus.



3. I make a selection of all of the areas that I want to have sharp detail, and save it as a channel.



4. Using that selection, I make a layer mask choosing hide selection.





5. In Photoshop I'll hit it with the dust and scratches filter, adjusting the radius to take care of most of it. Since these are blurred areas, loss of detail is not noticeable or significant.



6. Still working on the blurred areas, I usually use the patch tool to clean up areas the filter didn't take care of. This works faster and better for me than the other tools, but I use them also.



7. I may, or may not, apply more blurring to the edited areas to smooth out any transitions there may be.



8. I delete that layer mask, use the original save channel to create a new selection and invert that.



9. I create a new layer mask, again choosing hide selection and clean up problems with the detailed area. Use a variety of tools here depending on the problem.



10. Since it only really make sense to apply sharpening to detailed areas, I would do it at this point. The mask will prevent sharpening from effecting the rest of the image. This is a big benefit. Sharpening can bring out or create problems in the blurred areas. Having them masked out prevents this.



That's the basics I use.



Vance