Photo and Data Recovery - Inconsistent Results

[iantham]

My 2.5' 160GB external HDD connected by USB suddenly changed the file system to RAW. I believe this is called a corrupted hard disk.

1. I tried the trial version of ZAR to recover two directories that stored JPG format photos with great success (almost 100%, 741 files or 5.16gb)

2. I then bought the full version ZAR 8.3 build 19 to recover/save the entire 160GB disk that probably had about 40GB of data stored on it (JPEG, Word, Excel, Mpeg, PDF, etc).

3. During the saving/recovery process (licensed version) I clicked on the root directory to save all the files (everything) at once, to another external HDD.

My questions are as follows:

1. Using the full licensed version, although, I am able to see the JPEG thumbnail (after recovery), and the 1Mb-3Mb of storage spaced used for each of these JPEG files I am unable to open it. It reads "Failed to load photo" on adobe photo shop (on about 70% of the photos).

Strangely, the same photos which I recovered during first run using the trial version (that I saved on a different file directory) works fine and opens without a problem.

However, these same files (about 60%), stored under a different file directory that were recovered using the licensed version, 2nd run, would not open. Why?

2. Does recovering different types of files eg JPEG, Word, Excel, etc in the same run (recovery process) limits its success?

3. Also, if my corrupted HDD (hard disk) was formated under FAT32, and I save the recovered files in a different HDD formated under NTFS, does this limit the degree of success?

4. Many of the word file recovered contained nothing (20-30 blank pages).

5. Generally the health of my corrupted hard disk appears good (no bad sectors).

[Alexey V. Gubin]

1. That seems strange. I suggest you try to locate an autosave from your first run. Use "Open save file" on the first screen to load it. ZAR stores the files in C:\Program Files\ZAR\Autosave\, and the file name will contain a timestamp. Two files are saved per run, and you probably need second earliest.

2. No.

3. No.

[iantham]

Thank you for the swift reply.

Just curious,

1. Will doing multiple runs on the corrupted hard disk yield different results?..... Will I be doing more damage to the files by doing more runs (assuming the only problem with the hard disk is corrupted data).

2. And will the results (% of recovered data) fall will each subsequent run I make? Currently it takes about 5-6hrs to scan thru the entire 160GB external hard disk. And 3-5 hrs to save the recovered data on my other hard disk linked to a netword.

Thanks,
Ian

[Alexey V. Gubin]

Theory is:
1. Additional runs cause no damage.
2. %% of results should be the same each run.

Practice is:

1. If there is some subtle hardware problem, with a memory, or with buses, or whatever, simply inducing load on a system may cause further damage to data. Rare but happens.

2. If you run ZAR twice with same settings against the same volume, the results CAN be different. Throughout the algorithms, randomization is sometimes used. Hence, if a math problem (which ZAR has several to solve) has two close solutions, an arbitrary one will be picked. The result is similar to that you observe, one run you get complete recovery, the other run you have some crap results. There are even a couple of records of such a behavior here on the forum.

Since the new build is along anyway, I've adjusted the math part slightly, so it should now give more stable results (at the expence of about ten minutes longer run time).

So, could you please download and try http://www.z-a-recovery.com/zar-83-022.exe