data recovery

We’re currently working on recovery of audio files from a damaged hard drive. Audio files have the following string in the header “…AudicityBlockFile112B”. Files are short - 1,060,956 bytes each. The files for which we have filenames when put together in the order based on filenames create meaningful audio stream. Unfortunately there are a few thousand files recovered physically, w/o filenames. Question: Do these files contain metadata which would help us to establish the file sequence?
We’ll greatly appreciate your help, either through the forum, or via direct email at atgessen -{at}- yahoo / dot / com.

I’m not sure of this, but I think there’s no metadata in the files that will help put them in order… You could do that by the file creation timestamp, but I’m guessing you don’t have that available either…

What audacity version and which operating system was running on the computer that crashed?

The most relevant information regarding crash recovery is on this wiki page: http://wiki.audacityteam.org/wiki/CrashRecovery

Before trying to do anything make a backup copy of all the files. If you have any way of knowing the file creation time save that information too, it can help putting the files in order.

AFAIK there is no metadata in the .AU files to help.
If you were using Audacity 1.2.x, the files that are created during recording are in numerical order. If using Audacity 1.3.x they are not, and the only thing left to help you is the time stamp, (which I guess is no longer available?)

Steve, if there’s a feature request to add metadata to the .au files so that it will allow to reorder them in cases such as this, please add a +1 vote for me.

Thanks to everybody who responded with the info. Even though the answers were negative, at least it saved us from trying to find non-existing metadata. May I suggest to include something like, for example, batch ID and a timestamp in the .au files in the future releases?
pushkin

I don’t know how much that would impact on performance, or whether it is even possible given that during editing, new data blocks may be written for any part of any channel of any track, but it should not be necessary. Each file should have a unique name, which should be listed in both the AUP file and the backup project file. If the project goes sufficiently wrong for both of these files to be corrupt I have doubts as to how likely it is that there would be data worth saving.

Both the stability of Audacity, and automatic crash recovery have improved considerably and the developers continue to work to improve them further. If the addition of metadata will improve the ability to recover data from a crash I would expect that they have already investigated it. However, if anyone wants to suggest it as an enhancement to Audacity, feel free to propose it in the Adding Features section of the forum.