Using fpcalc and the acoustID API it returns the correct acoustID.
2012-08-19 23257, 2012
hawke_
Although the duration from fpcalc doesn’t match the duration from Picard. Maybe that’s why?
2012-08-19 23205, 2012
hawke_
But I thought Picard used fpcalc to do the calculation…
2012-08-19 23238, 2012
kepstin
picard determines the duration itself, from the tag/stream parser in mutagen.
2012-08-19 23230, 2012
hawke_
Wait, so it uses its own duration for lookup but fpcalc’s duration for submit?
2012-08-19 23232, 2012
hawke_
That’s asinine
2012-08-19 23246, 2012
kepstin
well, i don't know which it uses for submit
2012-08-19 23250, 2012
kepstin
but the one displayed inside picard it determines itself (it shows it even if you don't run fpcalc, of course)
2012-08-19 23224, 2012
hawke_
It uses the fpcalc one. :-/
2012-08-19 23212, 2012
brianfreud still thinks Picard should write TLEN :P
2012-08-19 23221, 2012
brianfreud
I don't know why luks was opposed
2012-08-19 23204, 2012
hawke_
TLEN?
2012-08-19 23235, 2012
hawke_
That seems kinda bad, to me.
2012-08-19 23246, 2012
hawke_
I mean, the duration should be calculable from the file without that
2012-08-19 23222, 2012
kepstin
the only case where duration can't be easily calculated from the file are vbr mp3 files that are missing the vbr header.
2012-08-19 23239, 2012
kepstin
need to decode those to find out how long they are :(
2012-08-19 23216, 2012
hawke_
Yep.
2012-08-19 23240, 2012
hawke_
But even that isn’t that hard
2012-08-19 23239, 2012
brianfreud
kepstin: I wanted TLEN, as well as the duration from the server, in metadata.
2012-08-19 23201, 2012
jcazevedo joined the channel
2012-08-19 23235, 2012
brianfreud
Otherwise, it's essentially impossible, within Picard, to detect if a file's duration is +/- n seconds of the MB track its associated with.
2012-08-19 23216, 2012
hawke_
Look at the “Duration” original/new value?
2012-08-19 23240, 2012
brianfreud
Back when we used PUIDs, I'd also hacked Picard to add those, so I could save files < 10sec as if they had PUIDs
2012-08-19 23243, 2012
hawke_
Doesn’t the duration value do exactly that, even if it doesn’t save it to the file as TLEN?
2012-08-19 23249, 2012
brianfreud
I *think* "Length" now has the server's duration value; haven't tested to see which value that holds (actual file, pre-existing TLEN value, server value, etc)
2012-08-19 23248, 2012
hawke_
The original value hold the actual file duration, the new file hold the server value
2012-08-19 23205, 2012
hawke_
(where “actual” means picard-calculated)
2012-08-19 23226, 2012
culinko
ermm, Heaven/Eearth capitalized?
2012-08-19 23232, 2012
culinko
Earth*
2012-08-19 23201, 2012
brianfreud
Assuming that's correct, wouldn't that mean that that for a file you've saved with Picard, that value now just == server duration, with no way to access the file's duration?
2012-08-19 23222, 2012
brianfreud
Assume server = 1:50, and file = 1:46.
2012-08-19 23250, 2012
hawke_
No, it keeps the original duration, doesn’t save it to the file or update the original value in the UI when you save.
2012-08-19 23223, 2012
hawke_
TLEN just seems like a terrible idea in general, just begging to be out of sync.
2012-08-19 23236, 2012
brianfreud tests
2012-08-19 23241, 2012
brianfreud
ok, yes, that's correct.
2012-08-19 23249, 2012
brianfreud
TLEN has it's purposes though
2012-08-19 23249, 2012
hawke_
Hmm…one buggy calculator or corrupted mp3 and you can’t trust it.
2012-08-19 23257, 2012
hawke_
And if you can’t trust it, you might as well not have it
2012-08-19 23215, 2012
hawke_
(Since that means you need to recalculate it anyway.
2012-08-19 23207, 2012
brianfreud
Hmmm, looks like the forums don't have "search" anymore.
So what are the uses of the TLEN header, that aren’t nullified by a buggy encoder/tagger?
2012-08-19 23250, 2012
brianfreud
iirc, the various reasons to have TLEN set included: some DJ mp3 software requiring it (else not even loading the files), Windows Media Player requiring it for VBR seeking to work properly, Nautilus and Windows Explorer both using it to correctly display vbr mp3 lengths in list-view mode, and per the ID3 spec, a TLEN must be present for any ASPI (audio seek point index) tag to be valid/used.
Isn’t that what the Xing header is for? (calculating the VBR duration)
2012-08-19 23213, 2012
kepstin
ah, so id3v2.4 added a new 'standard' set of tags to replace the defacto standard xing tag?
2012-08-19 23214, 2012
brianfreud
Xing's is a separate header from ID3
2012-08-19 23230, 2012
hawke_
I realize that, but isn’t Xing’s the one that everyone/thing uses anyway?
2012-08-19 23255, 2012
brianfreud
depends on your encode
2012-08-19 23259, 2012
brianfreud
*encoder
2012-08-19 23241, 2012
kepstin
well, lame adds a xing tag, and doesn't write a tlen or aspi id3v2.4 tag. And it's probably the most popular mp3 encoder...
2012-08-19 23245, 2012
kepstin
well, maybe not
2012-08-19 23256, 2012
brianfreud
LAME and Fraunhofer both don't write Xing headers. Only Xing encoders do. LAME writes LAME headers (which are an extension to Xing headers), and Fraunhofer writes VBR headers. Only the LAME header is actually open spec.
2012-08-19 23257, 2012
kepstin
the one include with windows is probably the most popular
2012-08-19 23205, 2012
brianfreud
LAME does write TLEN
2012-08-19 23210, 2012
hawke_
Does the MP3 spec not have any standard way to handle this?
2012-08-19 23220, 2012
kepstin
hawke_: the mp3 spec doesn't specify vbr.
2012-08-19 23226, 2012
hawke_
ah
2012-08-19 23239, 2012
brianfreud
I think lame needs --add-id3v2 before it'll do TLEN (among others)
2012-08-19 23252, 2012
kepstin
yeah, I don't tell lame to write any id3 tags
2012-08-19 23221, 2012
brianfreud
yeah, the mp3 spec doesn't specify cbr or vbr; it only says that a frame exists, and is bitrate X. CBR just does each frame at the same X, VBR does each one adaptively.
2012-08-19 23250, 2012
brianfreud goes to eat
2012-08-19 23228, 2012
kepstin
the replaygain tags that lame writes in it's extended-xing tag are mildly amusing - as far as I know, no other program supports them :)
2012-08-19 23252, 2012
kepstin
hmm. if you disable development mode on the server, it stops serving up the static resources, eh.
2012-08-19 23257, 2012
kepstin re-enables that for now.
2012-08-19 23208, 2012
kepstin
hmm. I guess it's the PLACKUP_DEBUG setting that does that.
2012-08-19 23233, 2012
kepstin
er, no, something' just gone weird on me :)
2012-08-19 23249, 2012
demosdemon
yeah it's plackup_debug that serves the statics
2012-08-19 23208, 2012
hawke_ joined the channel
2012-08-19 23218, 2012
demosdemon
the server's not supposed to server static files, usually nginx or apache take care of that since they are more efficient at it
2012-08-19 23248, 2012
kepstin
guess I can put nginx on my musicbrainz vm just to server static files, I suppose.
2012-08-19 23208, 2012
demosdemon
when PLACKUP_DEBUG is set, it runs a mini file server
2012-08-19 23223, 2012
demosdemon
it being catalyst, nothing to do with mb-server
2012-08-19 23228, 2012
kepstin
(right now I have my frontend proxy talking fcgi directly to plackup on the backend server)
2012-08-19 23254, 2012
demosdemon
usually you'd have a location /static block for those files
2012-08-19 23213, 2012
kepstin
demosdemon: yeah, but to do that, I need a web server on the same machine as the files are on :)
2012-08-19 23231, 2012
demosdemon
or move the static to a separate server
2012-08-19 23243, 2012
demosdemon
the one with nginx already running
2012-08-19 23246, 2012
kepstin
nah, I want to keep it all together for now.
2012-08-19 23210, 2012
demosdemon
normally I'd have my statics rewrite to a cdn like s3
2012-08-19 23230, 2012
kepstin
(technically it's really all "the same machine", the musicbrainz stuff is inside a vm on a private internal subnet)
2012-08-19 23212, 2012
demosdemon
you could even have nginx rewrite to another server that already has them (like test.mb)
2012-08-19 23225, 2012
kepstin
lol, that would be amusing.
2012-08-19 23226, 2012
demosdemon
if you don't plan on changing them
2012-08-19 23230, 2012
kepstin
might cause issues if they get out of sync
2012-08-19 23246, 2012
kepstin
but I already have them downloaded, might as well serve them up.
2012-08-19 23218, 2012
CallerNo6 joined the channel
2012-08-19 23226, 2012
DocPlatypus joined the channel
2012-08-19 23204, 2012
STalKer-X joined the channel
2012-08-19 23254, 2012
uptown joined the channel
2012-08-19 23209, 2012
JoeLlama joined the channel
2012-08-19 23218, 2012
uptown has left the channel
2012-08-19 23244, 2012
Kryztof joined the channel
2012-08-19 23250, 2012
derwin
man there are so many electronic labels, big ones, I'd love to clean up :/
2012-08-19 23248, 2012
yoasif joined the channel
2012-08-19 23223, 2012
voiceinsideyou joined the channel
2012-08-19 23214, 2012
CallerNo6 will get around to working on Lo�q some day.
2012-08-19 23230, 2012
jmvanel joined the channel
2012-08-19 23235, 2012
adhawkins-away joined the channel
2012-08-19 23202, 2012
reosarevok joined the channel
2012-08-19 23206, 2012
Vorpal joined the channel
2012-08-19 23212, 2012
jcazevedo joined the channel
2012-08-19 23238, 2012
adhawkins-away joined the channel
2012-08-19 23252, 2012
mb-chat-logger joined the channel
2012-08-19 23207, 2012
dankine joined the channel
2012-08-19 23232, 2012
jcazevedo joined the channel
2012-08-19 23255, 2012
ehrgeiz joined the channel
2012-08-19 23227, 2012
reosarevok joined the channel
2012-08-19 23250, 2012
reosarevok
Is there anyone who strangely doesn't have Portal 2 yet and wants a 75% discount coupon thingy?
2012-08-19 23220, 2012
reosarevok
Apparently they've given one to everyone who owns the game...
2012-08-19 23225, 2012
ehrgeiz
is it worth it?
2012-08-19 23219, 2012
reosarevok
Well, dunno, it's Portal
2012-08-19 23221, 2012
reosarevok
For less than 4€, I'd say yes, but then, I liked the first one
2012-08-19 23212, 2012
ehrgeiz
i'd take a look, yes
2012-08-19 23217, 2012
reosarevok
What's your steam user?
2012-08-19 23243, 2012
reosarevok
well, mine is my nick, feel free to add me and stuff
2012-08-19 23258, 2012
ehrgeiz
so i guess that coupon won't work via ps3store :P
2012-08-19 23223, 2012
reosarevok joined the channel
2012-08-19 23233, 2012
dankine joined the channel
2012-08-19 23214, 2012
JoeLlama joined the channel
2012-08-19 23226, 2012
derr84 joined the channel
2012-08-19 23256, 2012
jacobbrett joined the channel
2012-08-19 23254, 2012
yindesu joined the channel
2012-08-19 23231, 2012
yindesu
Is the Label documentation page still correct? i.e. we are interested in the Imprint, not the distributor/manufacturer? (unless said distributor/manufacturer also has their Imprint on the spine?)
