How do I remove or disguise the blackish metallic "glow" that appears on some dark areas of old B/W photos? They almost appear to be blue.

Should I scan in B/W instead of RGB?

Example please?

Perhaps you're speaking of the iridescent tarnish that sometimes affects the silver (ie: dark areas have the most silver)?

I know exactly what she's talking about Doug and I think your explanation is probably correct. It does take on a bluish hazey look. I don't know how exactly to get rid of it and so far all I had was areas that were small spots easily cloned out. If you have a vast area to clean up cloning won't work. So I am eager too to find a cure for this desease.
DJ

I still can't get my images to upload here so this is the link to my photopoint album.

http://albums.photopoint.com/j/View?u=1269000&a=9769878&p=55580751

This is strictly a guess on my part. Would a color range selection work for making local adjustments to the affected areas? Maybe you could make a selection based on color range, save it as an alpha channel, duplicate that channel, gaussian blur a little, then load it as a selection to make corrections on. I don't know if it would work, but that would be my first attempt. I'm not on the same field as others here, so take my thoughts with a grain of salt.

Ed

Sounds like a form of deterioration known a mirroring, caused by oxidative-reductive deterioration of the silver. It is usually the most pronounced in dark areas and can be most clearly seen by varying the viewing angle. One way to reduce it is to scan in the photo in color mode and examine each channel in RGB, duplicate the best channel and do the same for CMYK mode and duplicate the L channel in LAB mode, then combine the seperate images into one and finish by using burn/dodge tools and clone tool, etc.. Thats a toughie to correct though. I've used the above technique to clean up a few photos showing mirroring, but the results range from great to dismal. Hope this helps you some, though. Good luck, Tom

there has been an EXHAUSTIVE thread on the adobe photoshop forum for several weeks concerning a restoration with the same problem. not sure if you have to be registered for the link to work, here it is (http://www.adobeforums.com/cgi-bin/webx?50@108.rT21avn8hIE^1@.ef41488) ; if it doesn't work, the name of the thread is mold and mildew stains and the author is Linda K.

here (www.graphicspalmbeach.com/photos5.html) is the link to the site where she has posted her work. she really has done a nice job, think it took significant time.

i have several photos with this problem and think i posted some of them over in the archives, if you get a hankering to do more.

i know i tried using magic wand and color range selection on the blue, and adjusting with selective color with some improvement but no cigar. tom's suggestion sounds promising.

Went back in my files and found a good example of mirroring. See attachment. Interesting link, Kathleen. It logged me in as you! Is that what they call Identity theft? Tom

Sjm
Yup, that's what I thought you were refering to. Nasty stuff hey? Looks like you might just have to try out a few experiments with what the guys mentioned and from some insite you get off Kathleen's link. If you discover the magic method of fixing this don't forget to let us all know.

Kathleen
Your link worked fine and I notice it says I was logged in as Kathleen Crimm. I haven't read it yet but I plan to look it over and the same with the other link. Some how she managed to clean it up quite well in her example. Thanks for the links, they may be a great help.

Tom
What did you do to fix your example? You said the technique you suggested can produce a range of results and this on looks pretty bad to begin with.
DJ

I just had another thought on this (very bad news!). If it seems to be too time consuming, there might be another thing to try. This is only a thought, but if you could get two pieces of clear glass (no bubbles) a little larger than the photo, you could sandwich the two pieces together with a little petroleum jelly smeared between the two pieces. Put the glass sandwich on the scanner, then put the photo on top of that. Scan the image. Here's my thinking, which may not be very good: The deterioration causes more light to be reflected from the silver, therefore making the photo to appear as it does. By using the pet. jelly between the glass, this might cause the scanner light to become more diffused, or scatter on the image, therefore bringing the overall tone of the scan closer to what you want. It would give a softer scan, and the contrast would probably be less. What do you think Tom, Doug, anyone? Is it worth a shot, or am I completely off base? Also I don't know if there would be a depth of field problem. This might be pretty bizzare thinking, but what the heck - so was the pet rock! :D

Ed

Hey, they laughed at Galaleo and Columbus. It's worth a try and maybe you could use it as a top layer over the original scan and only cover the bad areas via a layer mask. You may have solved a big problem. Who knows.
DJ

Dj, Wish I had a quick and concise answer but no such luck! What happens is that the silver in the image oxidizes to silver ions which in turn begin migrating from the original site of the base particle and turn into elemental silver (metallic) ion state and it again reacts to light and darkens, only not in its original location--sort of "bleeding" over as it were. Masking the original image and blurring,replacing,dodging the "bleed" area can help, but so far, no quick fix! The problem, then, is actually too much image data and in the wrong place. Perhaps someone reading this will get an idea on how to remedy this. I, not having many wits to begin with, am at the end of what I still have!! Ed's suggestion would be worth a try, at least it might cutdown on the gloss! The example I posted also shows a loss of data(the "bleaching" on the suit and at the bottom which was the baryta layer of the paper showing thru) due to storage method(trunk in old unheated and damp barn) which will eventually consume the entire photo. As I understand it, there is ,as of yet, no way to really halt this type of deterioration, slow it down, but not arrest it. I used the aforementioned method to attempt a restore but it was a pretty grim result. Just aquired a new color correction plug in and am going to try it out on this photo. Perhaps with some success for a change! Tom

It appears that I'm way over my head with this fix! I'm still at the coloring book level compared to the pros here.

I'll try the suggestions and report back with results.....good or bad!

Thanks again!

Kathleen:

I've seen it more than once that someone will post their own personal link to one of those photo album sites, which includes their personal ID. So when someone follows that link, the site thinks its you. You might want to edit that post to show a visitor link rather than your 'owner' link.

Everyone else:

Good thread. Back in the pre-digital days one way to lessen the effect of silver 'mirroring' was to use omni-directional lighting or even polarized lighting with a filter over the lens as well. The idea is the 'mirroring' effect is caused by light reflecting at differing angles of incidence, giving the effect. Limit the light available and/or the angle of that light, and there's less light reflecting at different angles.

Now how that could be replicated on a scanner I have no idea. But I suspect you'd see differing amounts of mirroring if you turned the image at different angles to the scan bar.

please don't take my cookies away.

Sjm, You might try this--Doug gets the credit if it works--- Open your image, duplicate the background layer. Next go to image>adjust>invert. In the layers pallete select {color} as the blending mode and adjust the opacity slider to around 50 - 60%. Next, make the background layer active and mask an area showing mirroring. Adjust with curves to get it looking almost OK, then using the airbrush tool,set at around 11% pressure and large tip size lightly paint over the masked area with what color you think needs to be there.I tried this and it worked pretty well on the example I posted. In that case I used black as the color to fix the suit, but for light colored areas, use the color picker (eyedropper) to select the right shade to fix the blemish. Thanks for starting this thread!! Thanks Doug for tossing out the polarization tip. Thats the key---the migrated ionized silver is reflecting at a different wavelength--not much different, but just enough to be compensated for by some "layer and mode" play. Thanks!! Tom

Oops, here's the before/after. Please--I spent all of 5 min on the after--main point of interest is the left part of gentlemans suit and floor. Tom

Tom,

That's a pretty impressive fix! This has been a good thread, and worthy of a bookmark. Thanks for all the info.

Ed

Good work guys with your problem solving efforts. Looks like we may have a good fix for this type of problem in the future.

Tom_Great job restoring but I just have to say one thing. I never thought to see a restoration by you where you actually took the sepia tone out. :lol:
DJ

Oops--thats what happens when I try to work past my bedtime!! Tom

I've also encountered this problem, with no good solution. Although I was told it's called "silvering out". I searched, but couldn't find any worthwhile remedy. I went ahead and treated it as I would any other lighting type defect.
I'll have to keep an eye on this thread, maybe someone will discover a good fix!

If the "old-fashioned" way of dealing with this problem was via a polarizing filter or some such, how about a high-quality copy-neg of the original rather than making a direct scan? Either scan the neg afterwards or have a print made from the CN and scan that, depending on your available equipment.

I did resort to using a copy stand. I would take about 4 photos, rotating 90 degress, and then put them on seperate layers. I would fiddle around until I got rid of the the majority "silvering". That's the only method that I've been happy with so far.

I have had some luck using tracing paper on the scanner between the scanner and photo. It may take a layer or 2 to cut down on the reflection of the silver halide that is showing up in your image. Scan it once without the tracing paper (or onion paper) and then scan it with the paper, with some work, you should be able to attain a very nice looking image.

I have used this several times in the past with great results. Wish I could post an image, but my clients have asked that they not be posted on the web.

Hope this helps.

Paul Rupp :wavey:

Paul:

How do you avoid the grain of the paper? I've actually been doing some fairly extensive tests in this area (for textured prints, not for mirroring) and I've yet to find a material that doesn't add more problems than it removes. I've tried papers, vellums, plastics of all sorts... you name it.

Doug,
Velum is a good material to use. I couldn't think of the name when I posted. I have a local Drafting supply house that I go to that has a nice selection of papers and velum.

Another thing you could do (and I have this at my disposial as well) is take a piece of clear plexiglass and frost it yourself with a very fine sandpaper (wet is best) you'll have to experiment with several coursenesses (is that a word??) I have 3 that I have made, one with 600 grit, one with 400 and one 220, all done wet sanding. You MUST get a uniform frosting.

As gar as grain, if you do a litle gaussian blur you should be able to get any grain out.

It is still better then trying to get the silver out!

HTH

Paul Rupp :wavey:

The best material I found so far is some .25mm styrene sheets I found at the local hobby/train shop. It looks opaque from a distance, but you can actually see your hand through it. I discarded it as a candidate because of, again, the grain. It even comes in a thinner grade which would probably be even better.

I'm still searching for the perfect material for my de-texturizing tests (there's another thread about that somewhere here).

Hello all

I just discovered this site and am really excited as I have been looking for such a collection of experts as you all are :)

I think that I can add to (help solve?) this question. Being more of a photographer than a retoucher, I have copied hundreds of old photos with this problem. We always called it "silvering". I have been unable to get rid of the effect with a scanner, however when copying with either a film or digital camera the solution is cross polorized light. What this means is that you place polorizing filters on the lights (all in the same plane), then place a polorizing filter on the camera lens, then just dial the reflection out, then make the exposure. This technique also adds to the apparent contrast of the copy image.

From some of the posts on here I am convinced that any amount of time it takes to set up and shoot the original with my camera is surely shorter than doing some of the P/S tricks that some are trying. It has been my experiance that many of the same techniques we used to use with film cameras seem to work just as well with digital cameras.

Scanners are somewhat quicker if you have to set up a copy stand and lights everytime you want to use it, but they also seem to add a lot more "dirt and specklie stuff" to the images.

Anyway, just my 2 cents worth, if you have questions or comments, please pass them on.

Mike

Welcome aboard Mike. The "Silvering" or "mirroring" is a pretty common problem with the developing out papers and even is seen on the printing out type as well. Sadly, the oxidative-reductive cycle which produces it is irreversable, but being able to copy the original and eliminate or greatly reduce it is sure better than loosing the image for ever. Thanks for the description of using cross-polarization. If you are interested see the attachment for a close up look at mirroring(silvering-both terms appear in literature about this problem). This shows a section of a photo on developing out paper. the better looking area was shot with the light angle fairly high, the one showing the mirroring is the same area with the light at a low angle. Tom

OK this may make me look real stupid here and maybe I'm way off the mark, but if polarized lenses reduce the glare produced by how light reflects off surfaces and it seems to work for camera lenses. Why couldn't a polarized overlay placed between a textured photo or even the old photos damaged by silvering, reduce that effect? Doug mentioned looking for materials to do this, why not try a polarized overlay? I did a search and found a place that makes such a product but I don't know what the prices of such a thing would be. You can check it out here (http://www.precisionfilters.com/). It seems like it may be worth a try. It's possible it could affect the rest of the photo scan detail too but if you're experimenting with things why not?
DJ

I have been gathering some information on a type of polarizing filter which is used to "tint" the windows on private jets. From what I understand so far, this unit consists of a remotely controlled polarization sheet of material, controlled by a small touch pad . Sounds interesting, especially if the size of the filter itsself would be such that use on the platten a flatbed scanner would be possible. The ability to control the polarization remotely would eliminate the need to shift the position of the photo being scanned and allow for fairly rapid corrections. When I get a reply back from the company making these , if it looks like a useful tool, I'll post what info I have, Right now everything is a little hazy, fact wise. Tom