jack6800 wrote: 05 May 2020, 17:37
As a OLD IT engineer, all my knowledge about defragmentation is Speed Up performance of disk.
With you being an old IT engineer you are certainly aware of something called "backup"?
I know you get me many favors already by response and guide me direction, but would you kind enough to get me some more steps/help?
Will do as your question applies to everybody who is failing with Photorec.
Because those OLD JPG are my golden memory with ex-GF and my wife will not tolerant if I go find pay-help. However, I could pay you by paypal in advance as a data recover tutorial so that I could save both of my marriage and memory.
Sorry, I do not offer professional recovery services. You are probably able to pay a professional company by Paypal as well to save both marriage and memory.
Here is a simplified representation of that recovery issue when you loose metadata. Metadata is descriptive data as file name, length, different dates, attributes, whatever. Data is the content of your file(s).
Imagine you have a disk drive or an SD card. You are taking three pictures. Storage is mostly cluster-oriented. A cluster is one ore more physical sectors.
Look at screen 1. The first and second cluster is used by directory information. The first file uses clusters 3,4 and 5. The second file uses 6 and 7. The third files uses the cluster 8,9 and 10.
Now imagine that your directory information in cluser 1 and 2 gets inaccesible or overwritten.
Loss of metadata!
Photorec will scan your drive searching for fingerprints. Many files have specific fingerprints at their beginning.
So Photorec will find the JPEG fingerprint in cluster 3. It will simply put cluster 3 and the following clusters into a recovered file. Maybe the header contains length information about the file. In this case Photorec will know how many clusters the file needs. Otherwise it should stop once it finds a new header which is located in cluster 6. Unfortunately there is no guarantee that header type information is inside a file by chance...
Anyway, Photorec considers cluster 6 as the beginning of a new file. As cluster 8 contains header information as well, Photorec will start a new file here. It could be that the file starting in cluster 8 will become very lengthy - don't know.
It depends on additional available information.
Now we go to screen 2: You delete the second file located in cluster 6 and 7 and take another picture (screen 3) that will be saved in the clusters 6, 7, 11, 12 and 13.
Now your metadata gets destroyed (screen 4). Picture 1 in the cluster 3,4 and 5 gets recovered perfectly. File 2 gets recovered partly with cluster 6 and 7. Seeing the fingerprint in cluster 8 Photorec might stop here. But if it continues it has a problem to select the correct following clusters.
The third file located in cluster 8, 9 and 10 can be recovered perfectly but without any available length information Photorec might attach cluster 11 and following clusters as well.
I hope you got the message. Photorec is a fingerprinting specialist. There might be other software around that is able to evaluate remaining pieces of metadata, try it out. Maybe a profesional recovery has programmed some genius algorithm, I don't know.
Now you might have an idea why you can't replace a backup.
Edit 5.5.2020 21:58 CET: Labeled the "pictures" as screen and changed the text accordingly because it's confusing to mention pictures to describe the process and talking about storing pictures on the storage simultaneously.

- cluster mess.png (16.88 KiB) Viewed 12720 times