20:28 PM
Leftmost
The offset part is especially cool.
2013-10-08 28140, 2013
20:29 PM
kepstin-work
2013-10-08 28149, 2013
20:29 PM
kepstin-work wonders how those two got the same acoustid :/
2013-10-08 28147, 2013
20:30 PM
hawke_1
kepstin-work: maybe the really really close parts compensate for the not-so-close parts
2013-10-08 28107, 2013
20:31 PM
hawke_1
i.e. the error rate is averaged over the whole file and manages to be low enough
2013-10-08 28127, 2013
20:31 PM
kepstin-work
yeah, I suppose so.
2013-10-08 28130, 2013
20:31 PM
hawke_1
kepstin-work: That one of the problematic karaoke tracks? ;-)
2013-10-08 28150, 2013
20:31 PM
kepstin-work
yeah
2013-10-08 28101, 2013
20:32 PM
kepstin-work
well, that's my horrible different vocalist case, I think
2013-10-08 28102, 2013
20:32 PM
Leftmost
I don't see what's wrong with it, but I'm not very good at reading acoustids.
2013-10-08 28118, 2013
20:32 PM
kepstin-work doesn't actually know which fingerprints correspond to which tracks
2013-10-08 28153, 2013
20:32 PM
hawke_1
Leftmost: different vocalists to the same backing audio, I guess.
2013-10-08 28112, 2013
20:34 PM
hawke_1
This one is really weird.
2013-10-08 28115, 2013
20:34 PM
hawke_1
2013-10-08 28134, 2013
20:34 PM
hawke_1
Move up to -12, then -24, then -36…
2013-10-08 28118, 2013
20:35 PM
Leftmost
-48 too!
2013-10-08 28154, 2013
20:35 PM
hawke_1
and -60
2013-10-08 28158, 2013
20:35 PM
hawke_1
yeah, it’s consistent.
2013-10-08 28119, 2013
20:36 PM
kepstin-work
whoa.
2013-10-08 28129, 2013
20:36 PM
kepstin-work
how does that even
2013-10-08 28103, 2013
20:39 PM
kepstin-work
i wonder if it's a software bug somewhere
2013-10-08 28153, 2013
20:39 PM
hawke_1
dunno. I had another one where it was clearly stretched between the two. You could see the dark area shifting as you adjusted the offset.
2013-10-08 28158, 2013
20:39 PM
drsaunde joined the channel
2013-10-08 28107, 2013
20:40 PM
hawke_1
But nothing like that 'perfect match, shift, perfect match, shift' thing.
2013-10-08 28118, 2013
20:40 PM
kepstin-work
i'd expect stretched to be fairly common
2013-10-08 28132, 2013
20:40 PM
kepstin-work
particularly for older stuff that went through an A2D conversion
2013-10-08 28128, 2013
20:41 PM
kepstin-work kind of wishes that acoustid submissions would link the fingerprints to musicbrainz track ids :)
2013-10-08 28159, 2013
20:41 PM
hawke_1
2013-10-08 28130, 2013
20:43 PM
kepstin-work
yeah, that's consistent with a resample at slightly different rate or slightly different playback speed on analogue equipment
2013-10-08 28141, 2013
20:44 PM
hawke_1
Another one with that other weird offset, this time at 33 intervals (move up, start at -2)
2013-10-08 28145, 2013
20:44 PM
hawke_1
2013-10-08 28127, 2013
20:46 PM
kepstin-work
the only way to explain that is that chunks of the audio have been deleted/cut out from one of the files
2013-10-08 28142, 2013
20:46 PM
kepstin-work
or a bug in the audio decoder in fpcalc :/
2013-10-08 28101, 2013
20:49 PM
Freso
"acoustid submissions would link the fingerprints to musicbrainz track ids" +1
2013-10-08 28105, 2013
20:49 PM
Freso
And with that, bed time.
2013-10-08 28111, 2013
20:49 PM
hawke_1
make that +2
2013-10-08 28144, 2013
20:49 PM
luks
repetitive parts in a song are not that unusual
2013-10-08 28106, 2013
20:58 PM
hawke_1
luks: It’s not just repetition, or we’d see larger and overlapping chunks wouldn’t we?
2013-10-08 28125, 2013
20:58 PM
kepstin-work
in this, we're seeing different chunks of the song matching at different offsets
2013-10-08 28154, 2013
20:58 PM
hawke_1
right.
2013-10-08 28109, 2013
20:59 PM
kepstin-work
i.e. 0-10 matches 2-12, then 10-20 matches 28-38 or so
2013-10-08 28136, 2013
20:59 PM
kepstin-work
so one files looks similar to the other file, but with regularly spaced chunks of missing data
2013-10-08 28115, 2013
21:02 PM
kepstin-work
hmm. the 'differences' view in the fingerprint comparison is equivalent to doing an xor operation between the two, isn't it?
2013-10-08 28147, 2013
21:02 PM
hawke_1
I think it’s an AND
2013-10-08 28120, 2013
21:03 PM
kepstin-work
no, an AND would make it white when both sides are the same :)
2013-10-08 28142, 2013
21:03 PM
hawke_1
depends if black is 1 or 0
2013-10-08 28111, 2013
21:04 PM
kepstin-work
if you want 0,0->0 0,1->1 1,0->1 1,1->0 then you want xor :)
2013-10-08 28140, 2013
21:04 PM
kepstin-work
and and or both miss a case.
2013-10-08 28155, 2013
21:04 PM
hawke_1
Ah, so they do.
2013-10-08 28121, 2013
21:05 PM
kepstin-work
what do you know, fpDiff[i] = fp1[i + offset2] ^ fp2[i + offset1];
2013-10-08 28124, 2013
21:05 PM
kepstin-work
xor it is :)
2013-10-08 28132, 2013
21:06 PM
hawke_1
:-D
2013-10-08 28101, 2013
21:08 PM
kepstin-work wonders if rendering the fingerprint 32bits wide is meaningful, or if it's just an arbitrary choice
2013-10-08 28154, 2013
21:08 PM
hawke_1
it is.
2013-10-08 28159, 2013
21:08 PM
hawke_1
2013-10-08 28103, 2013
21:09 PM
luks
hawke_1: I think these kind of matches are cause by e.g. radio edits
2013-10-08 28121, 2013
21:09 PM
luks
if you compare original edit and radio mix, all of them would match at some point
2013-10-08 28157, 2013
21:09 PM
kepstin-work
luks: that doesn't create a repeating pattern of different chunks matching at different offsets.
2013-10-08 28117, 2013
21:10 PM
hawke_1
it would also be weird to have a radio edit of a cat stevens song. ;-)
2013-10-08 28137, 2013
21:10 PM
kepstin-work
why would you remove the first 2 seconds, then the 8-10 seconds, then the 18-20 seconds, etc?
2013-10-08 28155, 2013
21:10 PM
luks
I think it does, but if I'm not even able to type, I might be able to think right :)
2013-10-08 28120, 2013
21:11 PM
luks
kepstin-work: well, that's what I would call audio editing
2013-10-08 28141, 2013
21:11 PM
luks
if you need to create a short version under 3 minutes
2013-10-08 28145, 2013
21:11 PM
CallerNo6
for 70s stuff, wasn't there often a radio or "single" edit to keep the song within 2-3 minutes?
2013-10-08 28101, 2013
21:13 PM
kepstin-work
right, but they'd usually do stuff like remove an entire verse for that, not remove e.g. 2 seconds of ever 10 in a repeating pattern :)
2013-10-08 28117, 2013
21:13 PM
luks
you never know
2013-10-08 28106, 2013
21:14 PM
luks
I used to debug a case when the person who sent me two versions and asked why chromaprint thinks they are different couldn't even tell they are actually different
2013-10-08 28121, 2013
21:14 PM
luks
even I couldn't tell it by listening, I had to play them side by side in audacity
2013-10-08 28154, 2013
21:22 PM
CallerNo6
... and the difference was?
2013-10-08 28112, 2013
21:23 PM
CallerNo6 guesses "different satanic backward-masked messages"
2013-10-08 28110, 2013
21:27 PM
hawke_1
lol
2013-10-08 28159, 2013
21:45 PM
hawke_1
2013-10-08 28142, 2013
21:46 PM
kepstin-work
so strange.
2013-10-08 28118, 2013
21:47 PM
hawke_1
Fingerprint only submitted once.
2013-10-08 28139, 2013
21:47 PM
kepstin-work
maybe they ripped it from ascratched cd :)
2013-10-08 28126, 2013
21:48 PM
hawke_1
Hmm, would that do that?
2013-10-08 28139, 2013
21:48 PM
hawke_1
Seems strangely consistent.
2013-10-08 28120, 2013
21:49 PM
hawke_1
2013-10-08 28154, 2013
21:49 PM
hawke_1
luks: Do you store the release MBID with the submission, by chance?
2013-10-08 28140, 2013
22:02 PM
hawke_1
It appears that it’s not “Classics, Volume 24”
2013-10-08 28134, 2013
22:23 PM
CallerNo6
I understand that not everybody who scans to pdf will take the time to make the page numbers agree. But if you're the Getty and the document your scanning is a about a schema for art metadata? I hold you people to a different standard.
2013-10-08 28157, 2013
22:23 PM
CallerNo6
grr :s/your/you're
2013-10-08 28148, 2013
22:28 PM
misterswag joined the channel
2013-10-08 28147, 2013
22:30 PM
kepstin-laptop joined the channel
2013-10-08 28123, 2013
22:32 PM
hawke_1
What’s a Getty?
2013-10-08 28146, 2013
22:40 PM
Clint
i assume the reference is to the museum
2013-10-08 28148, 2013
22:40 PM
CallerNo6
2013-10-08 28152, 2013
22:40 PM
CallerNo6
but there must be others
2013-10-08 28151, 2013
22:41 PM
CallerNo6
haha. -1 to hawke's rfc
2013-10-08 28102, 2013
22:42 PM
CallerNo6
mb-style is on a roll
2013-10-08 28139, 2013
22:44 PM
hawke_1
-1 to that. Silly. Hopefully she won’t veto it.
2013-10-08 28148, 2013
22:44 PM
hawke_1
god forbid we improve things incrementally.
2013-10-08 28152, 2013
23:18 PM
kepstin_ joined the channel
2013-10-08 28110, 2013
23:21 PM
Lotheric joined the channel
2013-10-08 28148, 2013
23:26 PM
Lotheric joined the channel
2013-10-08 28147, 2013
23:35 PM
storrgie joined the channel
2013-10-08 28159, 2013
23:44 PM
reosarevok joined the channel
2013-10-08 28100, 2013
23:49 PM
JonnyJD joined the channel
