How restore faded photo

With old photo colors deteriorate with different speed. Blue gets typically lost first. So I think one way is to get the individual color channels sort of on a similar level

Please advise the steps. Thanks

Regards

I used only one one module, rgb curves, and only the red channel. You can also lift blue a little.

I am sure other people have more complex and better ways.

Used G’MIC plugin for GIMP, I played with Auto Balance and Color Balance in the Colors section. Preffered the Auto Balance results Then adjusted the contrast and lowered the brightness some. Check your distro if using linux their repositories may have G’MIC available as a GIMP plugin. If not check here;

I sometimes find with GIMP auto levels that while the result is great for faded colours the contrast may be too high. In that case I increase the black output levels and possibly decrease the white output levels to reduce the harsh contrast. I have attached an example of auto input levels in GIMP working on one of my faded images. Obviously would need further tweaks but it gives the idea of how successful it can be for restoring faded colors with a simple click.

BTW, I have also worked with Darktables negadoctor which is a very handy tool if scanning or photographing negatives.

image

Used G’MIC plugin for GIMP, I played with Auto Balance and Color Balance in the Colors section. Preffered the Auto Balance results Then adjusted the contrast and lowered the brightness some. Check your distro if using linux their repositories may have G’MIC available as a GIMP plugin. If not check here;

Thanks for your advice.

I can’t find ā€œWhite Balanceā€ under Colors. Only ā€œAuto Balanceā€ is there. Pls refer to upload screenshot

Regards

My advise would be to let this photo faded as it is. In my opinion the picture doesn’t improve after edit. Maybe restoration is possible with artificial recolorisation. But then you have to know the colors of the wall, the shirt, the floor, the armchair and the skin.

Thanks for your advice.

Whether you meant Negadoctor, the plguin of Darktable ?

How to Convert Film Negatives With Free Program - Darktable Negadoctor Tutorial 2020

I retain the film negatives of all my hard-copy photos. They are in good condition and well packed. If I can’t restore the faded photos tinted heavily with yellowish color I’ll scan their film negatives. I have done some tests before, scanning the film negatives on Samsung Galaxy S22 Ultra Mobile phone, using Samsung Galaxy Tablet as lightbox. Afterwards converted the digital negatives to digital positive on GIMP with ā€œInvertā€. It did a marvelous job.

I have >1,000 hard-copy photos to be converted to digital positive photos. It is not an easy job for me. Besides I have to adding description on all digital positive photos, such as date and place of their captures, name of persons on the photos etc.

I’m not going to print the digital positive photos but will create slideshows on them, adding background musics, subtitles, narrations, etc. I have long experience in creating slideshows running OpenShort, ShortCut etc.

On Internet search I found following YouTube video;

Darktable Newb :: Part 1 :: Getting Started :: Concepts and Basics

darktable NEWB :: Part 2 :: First Edits

darktable NEWB :: Part 3 :: More Modules, Masking and Denoise

They will be helpful to me because Darktable is completely new to me

Regards

You will find a white/black levels adjustment in the colors menus of GIMP. I’ve got some decent results with just playing with all the plugins and options available in GIMP. There will be limits of the restoration that can be done, camera and optics, and the film of the era, plus the aging that has occurred to the photos.

My try using rdarktable. On an impulse i took a different route, and inverted it in gimp and then treated it in rdarktable using negadoctor. Don’t ask me where ths idea came from :sweat_smile: but I’m quite pleased with result. Still a bit too green maybe in the skin? Should the shirt be blue, or? Hard to know what the colors actually where originally.
Nice pratice though, thanks, would have been fun to try it on a full-res scan.

d534ed7c93d68be2f76ed260295fc91e23987198_2_690x898-inverted

a sligthly different interpretation:
d534ed7c93d68be2f76ed260295fc91e23987198_2_690x898-inverted_01

third and final version, I’ll stop now :laughing:

for some reason I didn’t notice the dark shadows, so I lifted those a bit…
d534ed7c93d68be2f76ed260295fc91e23987198_2_690x898-inverted_02

Hi AdmFubar,

This is solely for practice in order to learn.

On GIMP
Colors → Auto → Equalize

then

Again on GIMP
Colors → Brightness - Contrast

Now the red wall removed but the image is not so sharp. Pls refer to image upload

Regards

GMIC-Color-AutoBalance + GIMP LAB
G`MIC => Color=> Auto Balance + GIMP LAB

Negadoctor is a module in DT and not a plugin. I have an Epson V850 scanner and the software is crap for negatives. It clips the highlights really badly. So I tend to scan the negatives as a positive which gives the orange inverted color output as a 16 bit Tiff file. I then use negadoctor to invert it.

I then use GIMP for most spot and scratch removal as it is much quicker than DT for extensive restoration tasks.

GIMP
The wall is in white color according to my recollection …
photo-sample60a GIMP

This photograph inspired me to create my GIMP plug-in Colour Cast Reduction1…
It was based on an idea proposed by Krikor and coded by Tim Tran in the GimpLearn forum.

As noted by Terry and satimis, the GIMP auto-levels or auto-equalize can be helpful but produce excessive contrast. GIMP Auto/Stretch Contrast can be very useful when the image is excessively dark, light, or lacks contrast. Unfortunately it is not available in Gimp Python, so I had to devise a way of achieving a similar result.
The plug-in creates a contrast layer with an opacity of 10% to compensate for the common loss of contrast in faded photographs, but a better adjustment can be made with the base layer.

col-cast-reduction1.zip (3.1 KB)

As with any automatic means of colour-cast reduction, it uses the assumption that the image should average to neutral. Try it on a sunset or a candlelit scene and the results are likely to be very peculiar!

The result with this image with no further adjustment.

ccr1