Skip to main content

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: foo_musicbrainz (Read 278386 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Re: foo_musicbrainz

Reply #425
@MordredKLB I realized that it is impossible to use it without instruktion

... "how to reconfigure the script to cut off the tail of the phrase? "
It's not a script. There's nothing you can change to fix this on your own. I'll have to push a new version to fix it.

Re: foo_musicbrainz

Reply #426
Can you make it easier to look up by TOC with a pregap? Perhaps look up a few common pregaps, 33, 31, 25.
Can you explain the issue a little bit more? I honestly never use the TOC lookup, but it seems to be working pretty well in my library.

Re: foo_musicbrainz

Reply #427
If I load tracks from a CD into Foobar, it is not aware of the pregap duration. This results in a different TOC and lookup on MusicBrainz may not find results ("No matching results were found."), unless a disc ID with 0 pregap has also been added to the database by someone. The IDs are not reliable anyway, and often occur on multiple editions. The best way is to see all slightly relaxed matches and make a choice. An alternative is to go through CueTools to get the right ID, but it takes more steps and manual tagging. Typical pregaps on old CDs.

I have version 0.4.6 from the Foobar repository. I don't know if there is a later update.

Re: foo_musicbrainz

Reply #428
If I load tracks from a CD into Foobar, it is not aware of the pregap duration.

If the tracks are ripped from CD to standalone tracks then of course pregap info is lost but fb2k is more than capable of reading pregap info from an actual CD in the drive or a cuesheet.



foo_musicbrainz has always been able to detect this and apply it correctly.

Re: foo_musicbrainz

Reply #429
snotlicker is right. The code to apply the pregap to the discid before submitting is still there and seems to be working correctly. "Relaxing" the pregap values essentially means trying with the original pregap (if one even exists) and then creating a disc id with several other pregaps. I'd need to check, but I assume that would require multiple musicbrainz calls, and doesn't seem like something you'd necessarily always want to do.

I don't have a disc drive in my computer anymore and all my media was ripped to mp3 long ago so I'm not sure I even have CUE files to do extensive testing of this with.

Re: foo_musicbrainz

Reply #430
Tagging existing files without CD or cue is exactly the use case I'm thinking of. I also don't maintain cue files with correct filenames that could be loaded into Foobar.

Re: foo_musicbrainz

Reply #431
Tagging existing files without CD or cue is exactly the use case I'm thinking of. I also don't maintain cue files with correct filenames that could be loaded into Foobar.
I guess it comes down to the user's intention. If you're looking up by TOC the component assumes you want that EXACT release. If you don't have a pregap, then the only exact match is probably a web release, but it's also probably more frustrating to get no results than an extra result that isn't a perfect match.

If we don't find a pregap at all, loose matching seems most beneficial, whereas if you have one changing it to return more results seems... counterintuitive. I'll have to think about this some more and see what's possible.

Re: foo_musicbrainz

Reply #432
Actually after doing a little more reading, I think this is kind of moot issue. Sorry, it's been a little while since I've looked at this stuff so my memory needed some refreshing.

Standard pregap for REDBOOK is 2 seconds. That's required. If there is no pregap foo_musicbrainz defaults to 150 (2 seconds * 75 samples/s). The pregap can be longer than 2 seconds, but this typically only happens in cases where there's a hidden pregap track. There is no chance that the "32" value you shared is the initial pregap in 44% of all albums (32 samples is not a valid initial pregap which must be at least 150, and 32 seconds is a really odd number for a pregap).

The numbers your link seems to be showing must be pregaps "between" tracks, which aren't typically found anymore. If the album was ripped correctly then it's my understanding that those pregap samples would already have been included in the sample. If that's the case then there's nothing to do. I don't want to play around adding random extra samples in the hopes we find a match. Also, given how this functionality is written, there's not a clean way to make multiple requests at this stage. It would require substantial refactoring of the code for what I'm assuming would be low/no benefit.

Unless I'm substantially misunderstanding the issue I don't think there's really anything to do here.

Re: foo_musicbrainz

Reply #433
The 31, 32 examples are additional sectors to be added to the default of 150 so you'd try with with 181, 182 etc. My screenshot of 2seconds above equates to 150 sectors on top of the default 150 so you'd start with 300 as per the detail shown here...

https://musicbrainz.org/cdtoc/cnzqAtpXYHa_LF9hj6j1AuYggz8-

Re: foo_musicbrainz

Reply #434
The 31, 32 examples are additional sectors to be added to the default of 150 so you'd try with with 181, 182 etc. My screenshot of 2seconds above equates to 150 sectors on top of the default 150 so you'd start with 300 as per the detail shown here...

https://musicbrainz.org/cdtoc/cnzqAtpXYHa_LF9hj6j1AuYggz8-
Ahhhh... I missed the pregap += in the code.

In the case I'll take a look and see what can be done, but if it's going to require a substantial refactor then I think it's probably not worth the effort and mucking up the code significantly.



Re: foo_musicbrainz

Reply #437
If the title of the album (which is very common) contains notes (CD1), (Limited Edition), (Instrumental) extension issues "No matching results were found". how to reconfigure the script to cut off the tail of the phrase?
I did this. It attempts to remove a number of common edition or descriptor strings (CD1, Limited Edition, Deluxe Edition, 10th Anniversary, etc.) from the ends of albums. It does a case insensitive search and then lops off everything after that in the album field. It also recognizes that it can guess wrong so I added a checkbox to quickly get the original value back:




Some album tag examples:
"Master of Puppets" -> No special text found so the check box will be hidden
"Master of Puppets Deluxe Edition" -> Will become "Master of Puppets"
"Master of Puppets [DELUXE EDITION]" -> "Master of Puppets"
"Master of Puppets {deluxe edition} Elektra Records 1986 Hetfield/Ulrich/Hammett/Burton" -> "Master of Puppets"
"Master of Puppets - 20th Anniversary" -> "Master of Puppets -" since I only strip [{( from the start of the search text
"Big in Japan (Japan Edition)" -> "Big in Japan"
"Big in Japan" -> No change
"Hellbilly Deluxe" -> No change

I tried to err on the side of caution, so usually "Edition" will need to be in the album title unless I'm reasonably confident that it wouldn't typically be part of a title. Test it out and if you've got some suggestions of other strings to search for let me know and I might add them.

Re: foo_musicbrainz

Reply #438
Myself I have thrown all those descriptors into a custom tag and stripped them from album in order to avoid the problem, but:
* There is "version" where you would put "edition"
* there are users who remove that word ("version"/"edition") because it is redundant
* there are "digipak" and "promo" and "limited edition bonus" and "mispress" and ...
* "CD1" and "CD 1" and "CD#1" and "CD #1" ...

So what about a checkbox to disregard everything from "("? That would technically be everything from "(" or " (" I guess.

Re: foo_musicbrainz

Reply #439
Myself I have thrown all those descriptors into a custom tag and stripped them from album in order to avoid the problem, but:
* There is "version" where you would put "edition"
* there are users who remove that word ("version"/"edition") because it is redundant
* there are "digipak" and "promo" and "limited edition bonus" and "mispress" and ...
* "CD1" and "CD 1" and "CD#1" and "CD #1" ...

So what about a checkbox to disregard everything from "("? That would technically be everything from "(" or " (" I guess.
So to take a step back, the whole point of this is to make searching for results easier and save the user having to manually remove text when from the album name when needed so as to provide a proper match. Regardless of whether the component automatically does it for them, or they have to manually make edits, the end result will be that whatever is currently in the ALBUM field will be overwritten, so it doesn't really matter if there are redundant words or if "version" instead of "edition". I want to filter out the most common use-cases and be done with it. I could create a phrase list 100 long, but at some point utility begins to diminish. I've also seen lots of examples of files where these types of edition naming schemes come in, so I've got some idea of what's reasonably likely to be encountered by the average user. I don't want to try and predict how user Porcus tags his stuff, unless Porcus is distributing files to a whole bunch of people who will then be running this component on them. Hopefully that makes sense.

I don't want to add a checkbox for "(" because it's not uncommon to be used in album names (I have 70 in my library). It might make more sense for "[" (7 in my library) or "{" (0!), but again I'm not trying to hit every corner case. Also every little checkbox adds extra complexity to a dialog that should ideally just be "hit okay" 90% of the time.

That said, "digipack", "promo", "mispress" are all great. "limited edition" is already picked up, but "bonus tracks" would be good too. I'll add some extra CD variations as well.

Re: foo_musicbrainz

Reply #440
Am I missing something obvious? I've just eagerly installed 0.4.7 b3 thinking that my issue with using the Original Release date would be addressed but I can't see any changes.

Re: foo_musicbrainz

Reply #441
Am I missing something obvious? I've just eagerly installed 0.4.7 b3 thinking that my issue with using the Original Release date would be addressed but I can't see any changes.
It's done, but not in beta 3. Planning to release 0.4.7 tonight if there aren't any other issues found.

Re: foo_musicbrainz

Reply #442
Oh, okay I'll wait a wee while!

Just a small remark re trimming edition text. I'm seeing increasing downloads having  [Shm-cd] appended to the album title. Any chance of trimming it out?

 

Re: foo_musicbrainz

Reply #443
Oh, okay I'll wait a wee while!

Just a small remark re trimming edition text. I'm seeing increasing downloads having  [Shm-cd] appended to the album title. Any chance of trimming it out?
Nevermind. Merry christmas. :)

Also, I added SACD, but should definitely do SHM-cd as well.

Re: foo_musicbrainz

Reply #444
Many thanks. Date handling now meets my needs - great.

But... [SACD] is trimmed but [SHM-CD] isn't.

Re: foo_musicbrainz

Reply #445
So to take a step back, the whole point of this is to make searching for results easier and save the user having to manually remove text when from the album name when needed so as to provide a proper match.[...]
I might have gotten it wrong: Is this about searching, i.e. making queries MusicBrainz, or is it about what is written back to files after a search proved fruitful?

Both 0.4.7.b3 and 0.4.7.b4 as posted here, cause immediate crashes, so I cannot test. However. 0.4.6 gives zero hits when searching for my Master of Puppets using artist "Metallica" and album "Master of puppets ("

Re: foo_musicbrainz

Reply #446
Many thanks. Date handling now meets my needs - great.

But... [SACD] is trimmed but [SHM-CD] isn't.
Yeah, it's not in the b4 build since I was already posting when I saw your response.

Re: foo_musicbrainz

Reply #447
I might have gotten it wrong: Is this about searching, i.e. making queries MusicBrainz, or is it about what is written back to files after a search proved fruitful?
It's just searching to get results back. Very rarely will musicbrainz ever have "Limited Edition" or whatever in the album titles.

Quote
Both 0.4.7.b3 and 0.4.7.b4 as posted here, cause immediate crashes, so I cannot test. However. 0.4.6 gives zero hits when searching for my Master of Puppets using artist "Metallica" and album "Master of puppets ("
That's not good. Did beta 1 or 2 (not sure I posted 2) cause any crashes? What version of foobar are you on? Also, when you say immediate crash do you mean on foobar startup or when performing some operation with the component?

Are you able to debug using the pdb file (beta4)?

Re: foo_musicbrainz

Reply #448
Just noticed that with latest beta, if I overtype the date field if there is no date given in any of the releases, it is ignored and a blank date is written to the tags on exit.

Re: foo_musicbrainz

Reply #449
Just noticed that with latest beta, if I overtype the date field if there is no date given in any of the releases, it is ignored and a blank date is written to the tags on exit.
Good catch. I'll fix that.