Enhance Faded Handwriting

[richardsullivan]

I found some older posts with members' generous attempts to decipher faded handwriting. I used their techniques as well as my own to try and reveal what is written here.
This is a 1860 cabinet card with ghost writing. I worked with a 2400dpi scan but RP only allows a 100k version to be uploaded. An enhanced portion I did looks like perhaps, or maybe not, "Schraeder" is written above the printed <Main St.>
I found that by viewing it as I slid the sliders in Levels, Channels and various special effects filters such as Emboss that each offered some different view of what this could be. The dark splotch above and to the left of "456" looks to me like a machine-printed <'35>.
It looks as there are two lines of handwriting that stretch across the card. The card is very small and the writing tiny.
If any members possess forensic handwriting skills or can advise me as to a technique I would be very grateful. This is a family photo. I can email the 2400dpi scan if anyone would like it.

[Tony W]

Richard, I suggest that you should upload it to a free file hosting site such as Flickr or Photobucket and then post the link here to the full size image. This way you will possibly get several opinions rather than just the one

I cannot see anything like your description on this image manipulating it in all sorts of ways - except for ghostly head above main street, my imagination taking over

[richardsullivan]

Thanks Tony:

I'm not sure how Photobucket works, as far as downlaoding. Can the large file version be downloaded using this code?

http://i766.photobucket.com/albums/x...SamLake002.jpg

I will attach an enhanced portion to show the handwriting.

Thanks for jumping in.

[John Wheeler]

Hi Richard
I agree with Tony about posting the image. I have had to do very similar things as yourself for family images since I have many images from about 1840. Some handwriting was very faded.

It is usually helpful to do 16 depth per channel RGB scans instead of 8 bits scans and save them as TIFF. The TIFF makes a big difference because some of the forensic techniques do not work with the JPEG compression even at the highest quality setting. In JPEG the chroma components are compressed considerably higher than the luminosity components.

I would be more than happy to give it a shot yet I would get the best results with 16 bit TIFF scans. Let me know.

Here are some alternate techniques to consider.

1) Shine a white/blue LED flashlight on the surface above the writing. Tungsten filament lights do not have near enough blue content and that could help.

2) Using high intensity LED flashlight, shine the light almost parallel to the card (i.e. from the side). I have had even subtle impressions in the card stand out high enough to read.
This is a higher tech and non destructive version of rubbing a lead pencil across the surface.

3) Use the above technique yet illuminate through the back of the card. Sometimes the fibers in the card contain some compression and becomes more visible when lighting through the back of the card.

Hope that helps.

I have not tried this as of yet but have heard reports that under some circumstances when illuminating and viewing under #1 or #2 if you use polarized lighting and polarized viewing that it sometimes helps increase the differentiation/contrast


All the above techniques supposedly work better with monochromatic light (i.e laser light where you can tune the color) yet that is very expensive and is risky for one's eyes. I have only taken the steps to do this with white/bluish LED lighting.

[richardsullivan]

John, I tried the flashlight technique and it does reveal additional writing or printing I didn't see before, but not enough to read. I will rescan as per your instructions and see if that reveals additional information and will report back. Thanks!

[0lBaldy]

You may want to try "The Color Deconvolution plug-in" from 4N6site.com

Your eyes must be much younger than mine.... I can't see anything readable...

[Tony W]

Richard like OlBaldy still cannot see anything on the larger image. If the rescan does not help and as you were able to discern a difference using flashlight it may prove worthwhile to try copying using a DSLR moving your lightsource to an acute angle

[DWThomp]

Originally Posted by richardsullivan

I'm not sure how Photobucket works, as far as downlaoding. Can the large file version be downloaded using this code?

Mediafire.com is a good, free site to share files. You can upload any type file to share.

[richardsullivan]

No, they're not good eyes, believe me. But I have the advantage of the high dpi scan, plus moving the sliders allows me to see the effects increase and decrease, back and forth, and that animation brings out detail.
Thanks for the plug-in tip. I am excited to try that.
Thanks to everybody for their input as well. This site is indispensable.

Originally Posted by 0lBaldy

You may want to try "The Color Deconvolution plug-in" from 4N6site.com

Your eyes must be much younger than mine.... I can't see anything readable...