Unsharp Mask

[Mike Kovacs]

I haven't had a need to use it for CMKY images but I just tried and indeed you are correct that it won't work on them. I tend to output to inkjet or contone photo printers, not colour separations.

[Henry Ambrose]

Pay particular attention to what paulr wrote above about the workflow approach.

The reason to not sharpen or USM (much) until you are ready for output is that you often don't know the particular printing method you will use or the size of reproduction. Different printing methods or devices will call for varying degrees of USM to look best. And since you probably don't know now what kind of printer you will own in 5 years or that a magazine will publish one of your photos next spring, you should save final sharpening for the last step when you know how and how large it is to be reproduced.

A little USM immediately after scanning will help to regain some of the sharpness that is lost through the scanner. This amount of sharpening is appropriate for your "archive" files to which you will turn in the future for printing. At that point you would prepare a copy of your original file for output, keeping your archive copy unaltered.

As Frank wrote about offset printing, you have to sharpen the heck out of the files to get good results for common printing resolutions. Files that look drastically over sharpened will look great in ink and paper at 150 line screen. those same files would probably look horrible on an Epson inkjet, so would choose very different sharpening for inkjet, offset, Lambda or Lightjet printing.

What Michael wrote about Imacon "secret" USM is one reason that Imacon scans look so good right out of the machine. It seems to be just enough to get the scan back to the level of sharpness of the original. If you are using a typical flatbed you will certainly benefit from a little USM right after scanning - but just a little bit.

[Brian Sims]

Thanks for the advice....this is good stuff. I'm taking Paul's advice and running some experiments. I'll share the results when I'm done.

I have to apologize to Michael Mutmansky--I attributed something to him that was incorrect. I found the advice in his article to be very intuitive. I was reading many pieces about USM at the time, and somewhere I read that a minimum amount of USM at scanning was advisable. That seemed to fit with Michael's strategy that your scan is like your first properly exposed silver print. I extrapolated from that to mean properly exposed AND focused. Still, I want to apologize for my error of attribution.

I am puzzled by the reference to Imacon's "secret." If Imacon was trying to hide something they could easily have not allowed a negative number in the USM field. The knowledge of using -120 to eliminate all USM on the Imacon is common knowledge. That's kind of like referring to the condenser lenses in an enlarger as creating a "secret" and unfair sharpening advantage. I think it just means that zero is a reference point for the minimal amount of USM the Imacon designers thought was required to offset the inevitable degradation of digitizing film.

Again, thanks to you all for taking the time to help me sort this out.