[mb-users] Re: MusicBrainz-users Digest, Vol 50, Issue 27

[Brian Schweitzer]

>
> Brian, do you actually KNOW it is impossible, or do you only believe it
> can
> not happen? AFAIK, the PUIDs could be calculated by taking samples at
> regular intervals (say at 5 seconds intervals), which could miss the
> bleeps
> sometimes. (I did not say they are calculated this way, just that I never
> saw any indication this was not the case). Here
> http://blog.musicbrainz.org/archives/2006/03/finger_fingerpr.html , it is
> explicitly said: "It should have a lot fewer duplicates and collisions
> than
> TRM." "less" at least means collisions can happen :-(
>



Well, as for exactly how unique "unique" is, the only way we could really
know that for certain is to know better how puids are generated - and that's
the MusicIP Secret Sauce.

I can, however, point to a recent example I'm working on, where the
distinction between the two versions is even less audibly significant than
the clean/explicit distinction here.

I'm still waiting on MusicIP to generate most of the PUIDs (why???  I
submitted analysis at least a week ago and still no PUID found for 10
tracks!).  However, 2 of the tracks have already been assigned puids.

Now, this is vinyl sourced, but specifically was silence-trimmed to be
identical in length, start time, etc, to the CD releases.  Thus, given
MusicIP's claim (which seems to have been proven true by various people at
MB who've ripped tapes and LPs and had PUID matches found to CD-generated
PUIDs)
 that the source doesn't make a difference, if the two tracks are audibly
identical:

Normal case:
http://musicbrainz.org/release/cc8a35fc-ef70-40b5-b410-c1b5a029a866.html
Special LP case (different master mix):
http://musicbrainz.org/release/499345df-cce3-4a07-8bd3-90249a6bbbde.html

Also, in terms of sample rate, totally opinion here, but I would be shocked
if the sample gap was as large as 5 seconds...  I'd tend to guess it would
be closer to 500 ms (or less).

Brian
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.musicbrainz.org/pipermail/musicbrainz-users/attachments/20071028/9d50e7fa/attachment.htm