#musicbrainz

the songs from this are showing up as...

mediainfo |grep Perfomer > Performer : Tom Constanten (organ);"Ron ""Pigpen"" McKernan (organ)";Jerry Garcia (guitar);Bob Weir (guitar);Mickey Hart (drums (drum set));Bill Kreutzmann (drums (drum set));Phil Lesh (electric bass guitar);"Ron ""Pigpen"" McKernan (percussion)";"Ron ""Pigpen"" McKernan (harmonica)";Jerry Garcia (vocals);Phil Lesh (vocals);"Ron ""Pigpen"" McKernan (vocals)";Bob

Weir (vocals)

I see them individually in picard as e.g., Performer [electric bass guitar] | Phil Lesh

is this coming from musicbrainz?

yeah, i guess it is. ok, where is this coming from ON musicbrainz and why is this ONE gd cd like this?

s/cd/release/

<04elomatreb> that's just how Picard stores performer information

<04elomatreb> the data is sourced from the relationships at the bottom

i was wrong

it's not one album.

of 3362 flacs, there are 696 that aren't Grateful\ Dead$

most of those have this individual band thing (one album is the dead and dylan so that's not the same thing).

individual performer thign.

i don't suppose there's an easy way to get this to be more uniformly like the vast majority of the grateful dead recordings I have?

I guess you can delete those tags/have Picard not write them? If they’re on some albums but not others, and that bugs you

rock tag on black metal nags me to hell and back 😅

<iconoclast_hero> "i don't suppose there's an..." <- you want to look into the two variables for file naming when writing out: %albumartist% vs %artist%

if I need it on some albums like VA-stuff, I manually overwrite it after successful detection and afterwards write out/move the files

if you don't want especially the relationships: picard->options->metadata->untick

<08wileyfoxyx> yvanzo: could you please refresh test server?

mr44er, so you're talking about FILE NAMING, i.e., what is in the actual path of the file, correct? this is what end up being in the metadata.

yes. to hopefully make it clear, what I mean

detect album, wipe out unwanted stuff either manually afterwards, but before writing

...or with automagic if possible in the metadata options of picard

i hope you can see the pictures on irc-side. check same entries what you want to keep, different you want to kick out, then write 🙂

i can see some links.

i'll try to log into the matrix.

Yes, i manually removed some from the Fillmore release i mentioned origionally, but that's really stopgap and the problem ends up being on the cli with mediainfo

and the inconsistency.

because right there, i have like 1/6th of my dead albums i can't find by using "performer" anymore.

if it is coming from mb, and is inconsistent in mbdb, then picard's not going to be much help and there's no way I'm addressing that many files by hand.

if you get links, my matrix server should serve them as http

yes, thanks, i was able to see them.

i found a good example.

workingman's dead.

and to make the point further...

why is this sort of inconsistency acceptable in MBDB/

not only within the single band, but within a release group?

not every single release in a group is edited that deep

6fdb82d8-9aee-35ed-8092-8434866bab7f

I don't want any of my files being edited to that level.

I don't need to know pigpen has a harmonica credit.

anyway, it seems to me that i'm out of luck because both cases fall within the MB guidelines and once the performer is broken down to individual band members, there's no field to go back to.

i'd have to write something that refers the artist back to album artist on all albums that aren't flagged as multi artist.

which would only be a partial solution. i think.

do you have a link to a specific release?

theres muuuuch in the group

but I still think that picard-options-metadata-"use standardized instrument and vocal credits" <- untick, should do the job, but only before writing out

or maybe untick everything there

another way should be scripting that %artist% should always overwrite %performer%, regardless of the entry/value

iconoclast_hero: if you don't want Picard write the performer tags use a script: $unset(performer:*)

ok, thanks.

that's probably where i'm headed.

You'd need to delete the tags once from all files manually I think. But Picard won't write the performer anew then

I don't fully understand the problem with the performer credits though. Vast majority of players don't even show them. And there are still the artist and albumartist tags for what is considered an albums or tracks primary artist

outsidecontext (IRC): ah, nice. good to know that wildcard works, digged not too deep in fancy scripting

the players aren't his problem, grepping from mediainfo is.

```

$ mi Grateful\ Dead\ --\ Workingman\'s\ Dead*/*Uncle\ John*flac|grep -E ^Artist

...should be more correct, but I don't know

artist should be from title, albumartist from album

s/title/track/

I see

Yes, with $unset such wildcards work. With $delete I think unfortunately not

yeah, grep from mediainfo.

ultimately, i have a database of music and associated tags... and the way that i generally interact with it is through bash.

the tags are added via picard, but getting info back out is through bash.

it occurs to me that i could abstract all that data to a real (e.g., sqlite) database and query it that way, but i've written like a dozen successful sqlite queries...

Does anyone have a Bandcamp release with hidden tracks for me? I need one to test my parser :)

<02UltimateRiff> kellnerd: this one's got like 5 https://voodoopup.bandcamp.com/album/parasite

<02UltimateRiff> (good album too, lol)

Thank you very much UltimateRiff <3

Lol, I already found the first bug in my parser before even getting to the hidden tracks 😂

Apparently I have only been testing with releases which have physical packages available.

[14:00] <reosarevok> I would be in favour of pushing all languages with active translation into beta, but maybe there's a good reason we haven't

I'd do that for test at least. for beta sometimes, if something is oddly or badly trasnalted it could lead to someone adding data wonkily based on bad understadning

Hi