Newbie here. My wife inherited 411 Ektachrome or Kodachrome family photos taken when her children were young - around 60 years ago. They were scanned and made available in digital form with no editing or correction. They all have the familiar red shift which indicates the blue and green dyes have significantly faded. Most of them have the same general appearance which means once a reasonable correction factor is applied they could be processed in batch.
Contrary to most photo editors which have rather simple features RawTherapee seems to be a Swiss Army Knife of editors. It is very comprehensive and very complex and does some amazing things. Since the photo correction I want to do is (to me) a common task for anyone possessing old, faded photos I thought there might be a āpresetā or āauto adjustā feature or at least a tutorial somewhere that could lead a newbie through the process. Your help will be greatly appreciated.
Welcome to the forum. If you have one of the images scanned, one you donāt mind posting, it would work better to see what sort of cast youāre facingā¦
What is the file format of the scanned images? If they were digitized as raw images with a DSLR or mirrorless camera, you can make significant corrections. With jpegs youāll find it had to make sufficient correction. RawTherapee does work on jpegs, but I doubt if youāll be able to make enough correction to correct the color casts.
If all else fails, just convert them to B&W. But to correct the color casts in RT, try these methods.
Find an area on the file that should be neutral grey. Click on it with the white balance eyedropper tool. Youāll probably need to click on different spots to get the best result.
In the RT color tab, try adjusting the sliders in the white balance tool.
Also in the color tab, try different settings with the color toning tool. Also try the RGB curves.
Staying in the color tab, go to the HSV Equalizer, and try reducing the S (saturation) of the color cast.
If these slides are anything like those I work with, the Kodachromes might have held their color well, but Ektachromes, Anscochromes, Agfachrome and others will probably have shifted and faded considerably.
Hi guys! Thanks for the quick reply.
I donāt recall but I think the slides were individually scanned with a device nearly 20 years ago. They are jpeg running 250 to 500 kb and roughly 1600 x 1900 pixels.
I am not doing archival quality restoration so the primary goal is to restore color balance. Anything beyond that - brightness/contrast/gamma and so on would be a bonus. I will upload a sample photo.
Iāve done some color restoration myself and itās always tricky to get things back into shape. The good thing is that your precious photos are scanned, but whoever did it was not really worried about the highest fidelity. The most limiting factor for you will be the fact that the images are jpegs. If you still have the originals, you might consider getting them digitally archived properly.
In any case, I would simply play with the RGB curves first and then use the Color Toning module to finetune and create multiple adjustment ālayersā.
Iām not that familiar with Rawtherapee, but I had a goā¦ I just used the Red compont of RGB curves, and also added a little saturation in the Exposure module. (not shown in the screen shot)
I tried to go a bit further, but didnāt have much success. I agree that it would be easier with a higher quality file, but also agree that itās great that these (probably precious) old slides have actually been digitized, before they have a chance to get lost or even more faded!
I do most of my photo restoration in gimp. Try auto levels adjustment to restore faded colours as it usually does a great job. I would not use darktable or rawtherapee for restoration work
I found a thread here Faded old Ektachromes - filter to repair?: Retouching Forum: Digital Photography Review which seems relevant. I havenāt tried following any suggestions yet, but GIMP (another opensource application like Rawtherapee) can do most stuff that Photoshop does, so it might be worth a try!
I just saw that Terry suggested GIMP too!
Is this in the normal levels tool in GIMP? I tried āauto input levelsā in that tool, but it didnāt do much. However, Colors>Auto>Equalize followed by a quick levels adjustment got me this:
The low level of data in the file is really showing though.
Iām off to bed!
Edid: Just had another go in GIMP, this time using levels (manually) on the separate RGB channels, followed by Hue & Saturation and also fiddled with color temperature.
And one moreā¦ IMO this is the best Iāve managed. Equalize, then Color temperature, then a quick midtones adjustment in Levels. All in GIMP still, it seems suited to this kind of job.
Wow! Thanks for all the suggestions. The altered photos you guys did look great. But as a newbie I have to go baby steps. This is a very capable editor. I tried editing several photos and was able to improve things. However, now I want to revert to my originals and start again. But I find a .pp3 file with the current values. How do I revert back to the originals? Read the Manual maybe?! It looks like there is lots of reading to be done here.
Yes the normal levels tool in GIMP āusuallyā does a great job, but not with this image. This image has defeated me and GIMP. If the original slide was available a new scan using the color fade options in the scanning software might help. I also wonder what programs like RawTherapee and Darktable would do if the slide was copied using a camera with RAW file format. This is one tough image to try and fix. I had a quick go but I am far from satisfied or proud of my effort here.
Glad to be of any help! Now, to revert to original, again I donāt use RT that much, but I think if you delete the .pp3 file things should return to original.
Of course, it never hurts to read the manual Iām going to have a look myself.
Well, I havenāt managed to read anything yet, but I canāt let your responses here go unacknowledged. āSteven 123sgā you have posted some very promising versions of my photo. I decided, for now, to try GIMP from an earlier suggestion. It looks like these photos are marginal for restoration having very little blue and green left to work with. Here is a photo featuring four generations and what GIMP did using Colors/Auto/Equalize.
Perhaps I will learn some fine tuning to partially replicate what Equalize button is doing. BTW, the dark hair beauty is my wife of 30 years. The photos 60 years old.
Thanks for the feedback! I agree that fine tuning the Equalize function would be niceā¦ Iām not entirely how or if it possible to do quite the same thing manually. Iām really not that well up in this stuff though. Great photo too!
This is a quote from the manual:
8.16. Equalize
The Equalize command automatically adjusts the brightness of colors across the active layer so that the histogram for the Value channel is as nearly flat as possible, that is, so that each possible brightness value appears at about the same number of pixels as every other value. You can see this in the histograms in the example below, in that pixel colors which occur frequently in the image are stretched further apart than pixel colors which occur only rarely. The results of this command can vary quite a bit. Sometimes āEqualizeā works very well to enhance the contrast in an image, bringing out details which were hard to see before. Other times, the results look very bad. It is a very powerful operation and it is worth trying to see if it will improve your image. It works on layers from RGB and Grayscale images. If the image is Indexed, the menu entry is insensitive and grayed out.
One thing that I found worth trying was duplicating the layer, applying the Equalize function to the top layer, then reducing the opacity. This effectively allows the strength of the effect to be varied.
I can give more details on doing this if it helps! Hope you donāt mind me posting another version - Iāll happily remove it again if you like
Thereafter I did some layer-work with your original and the stress-image above and I used a chroma noise filter (GIMP>GāMIC>Testing>Iain Fergusson>MS Patch Chroma) with this result: