Audio mode - Data compare, Exact & Similar

[therube]

Audio mode - Data compare, Exact & Similar.


I'll note that Data Compare is not an indicator of the validity of an audio file,
or at least not the "completeness", necessarily, of an audio file,
only that the "data" therein is exact or similar.


OK, this is really weird - & I'm looking at .mp3, now... (as the same did not hold true for .wav).


Take a .mp3, copy it (so it at that point is exactly the same), then truncate it, deleting the final (physical) half of the file.
Do the same, but this time, truncate the first half of the file. And for good measure, truncate the open & closing 1/3's of
the file.

So you have original, truncated top, truncated bottom, & also truncated top & bottom.

Now, make copies of all of those.

Note that all versions play, truncated or not.
Note that all versions display "media information"; bitrate, length...
Now, obviously, any truncated version can only play what is actually there, but they will play.


Now, in the DC realm, DC finds all of those associated files, orig & respective copy, or various truncated versions & their respective copies, as Exact & Similar data compares.


What I found odd, perhaps, was with the Similar quick match (15 second) mode, it did not find the original file to compare with the file where the bottom half had been truncated. (Maybe given the dissimilarity of lengths, it was not considered for comparison? Where had it been that only the first 15 seconds were taken into consideration, that part would & does compare exactly - up to the point of truncation.)


Anyhow, one should not take an Audio mode Data or Similar comparison match as meaning that a (at least .mp3) file is "valid" or "complete" (in the sense that a user might expect).

(On these, purposely truncated files, that do play, a utility like MediaInfo may note "Conformance errors" with the file.)

[DigitalVolcano]

From what I can see, this is normal behaviour. MP3 files can remain playable and report valid-looking metadata even when they’ve been truncated, so Duplicate Cleaner will still match the parts that exist. This is why it’s important to also use the ‘match length’ criteria if you want to avoid matching incomplete or truncated files.