Why is LR adding a pink edge/highlight on imported files?

GREAT!  Knew that someone would be able to identify the problem and offer a solution. 

Bob - Does LR offer the same ability to change the profile on export or otherwise?  Would save a step in cleaning up the scans I've already completed.

Per - I will check the scanner and see if I can change the imbedded profile there. 

I appreciate everyone's help and suggestions.

Here's a link to the original scan which I hope will download for you.  Its very large so Google adds a virus warning but it was clean when uploaded.  GCNP as scanned.tif - Google Drive

My scanner software has three settings for dynamic range - full, auto, and custom.  Playing with the settings I've found that in most cases with very bright areas "auto" will always give the pink edge upon import.  "Full" works most of the time unless there are extreme differences in light and dark areas.  "Custom" works much the same as "full".  Any other of the auto adjust features have to be off to avoid the banding.  So to me it appears to be a blow-out issue that Lightroom must be adding to highlight those blown out areas but is somehow permanent instead of just a warning highlight.

Thanks,

I've tried a variety of color profiles (including sRGB and Adobe 1998), none of which significantly changed the over exposure/saturation or affected the pink banding.  I also turned off the GPU, which had no affect.  I'm using a Windows based Microtek Scanmaker i800, at 4800 dpi and multiple scan lines.  Scans are saved, in TIF format, to a common disc and then imported into LR on my iMac.  If any of the "auto" or "custom" options are turned on in the Windows software, I tend to see more pink lines in the original scans but it doesn't happen in a consistent manner.  The pink banding is always limited to the brightest edges and doesn't appear in dark or underexposed scans.  I was usually shooting Kodak Ektachrome or Kodachrome slide film and the banding appears in scans of both films.  The over saturation appears frequently but can usually be adjusted out.

I could be wrong, but I think this looks like the kind of "pseudo-solarization" you can sometimes see when one channel hard clips just before the others. In some special circumstances you can even see this with raw files, if the clipping occurs in the sensor.

Did the OP try to rescan without clipping the highlights?

And I just have to ask: Why this extreme oversaturation?

Screen shot of the develop module as imported.  As you can see there are extensive areas with the red highlighting which can be handled by reducing highlights or exposure.  The pink banding appears along the edges of the blowout areas but doesn't seem to be affected by any adjustments and the pink outline remains after these basic (exposure/highlights) adjustments.  I also checked the "Add During Import" settings to ensure that nothing was being done to the file during importing.  The only addition/adjustment was addition of a copyright to the metadata. 

As to screen calibration, I use an iMac and MacBook Pro, both with Apple calibrations, performed recently.  This problem occurs on both and has for a long time (I'm only now reaching out for help with it).  The pink banding also carries over to prints so it appears to me to be independent of monitor or calibration.

Thanks,

David

Sorry about that. This is the file after importing into LR.

The following is the original scan.

Have you done any edits to these files in Lightroom?

Would it be possible for you to share one of the original files with us, like it came from the scanner - preferably the one with the white shirt. You can put it on Dropbox, and post a link to it here. If the original is a jpg, you might be able to post it here, the same way you posted the images above.