Not Your Everyday Banana

by Bart Arondson

submit your photo


Hall of Fame
View past winners from this year

Please participate in Meta
and help us grow.

Take the 2-minute tour ×
Photography Stack Exchange is a question and answer site for professional, enthusiast and amateur photographers. It's 100% free, no registration required.

I've lost my Lightroom catalog during OS migration. I have couple of JPEGs developed from RAWs and would like to put few more JPEGs from other RAWs in similar mood. Is there a way to reverse engineer development settings having RAWs and JPEGs developed from them?

share|improve this question
add comment

3 Answers 3

There's no magical approach here. You'll have to look at each one. However, you have an advantage since you did the work yourself: you can, hopefully, look and be reminded of what you did.

Other than that, the only way is to be familiar with the camera, software, and RAW development in general, and then apply that expertise to each image.

share|improve this answer
add comment

I have noticed that in the combination of CameraRAW (via Bridge) and Photoshop some (!!) settings are written as metadata to the JPEG - at least that is what some of my Flickr images imply.

However there is a huge but with this: It is only a few settings - and some values are missing completely, e.g. noise reduction.

share|improve this answer
add comment

If all else fails do it the hard way.

I assume you have only done raw conversions and nothing more. If you look at it from a theoretical point of view then you have a high dimensional input (raw file), a high dimensional output (jpg) and you need the function that maps input to output. You are lucky because you have the function but need some parameters.

The dcraw and ufraw are two raw converters that can be used from the command line, I don't think that Lightroom let's you do this.

What you would need:

  • A distance function between two jpeg images. The sum of |pixel-differences| might work.
  • A search function. A Genetic Algorithm would be easy but probably not the best solution. Directed search is better. Maybe a greedy algorithm?

Anyway, this could be done with a bit of time and a quick computer.

share|improve this answer
    
+1 for the genetic algorithm idea. might work! –  Agos Dec 14 '12 at 9:35
    
High dimentional input is not a must. Heuristic algorithms taking samples from small places of images should work :) Not sure how it will work with cropping and rotations though... –  Random Jan 8 '13 at 10:10
    
Good idea of using only small patches of the image! "Not sure how it will work with cropping and rotations though... ", I clearly stated that only raw conversion was done. –  Unapiedra Jan 8 '13 at 14:53
add comment

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.