I have lots of songs for which I don’t have other album tracks.
Running those through Picard is a very bad experience, since Picard will always try to match them to an album.
Especially with ‘hit’ songs that could be released on best-of compilations, or on box-sets, you can grow a nice beard before all tracks from a large folder are matched.
The same goes for many classical compositions.
Most classical works were never intended to be on what we nowadays call ‘an album’.
Is it possible to tell Picard not to look for albums, but only to look for the data of these single tracks?
As far as I can tell, Picard does not have that option. However, Philipp Wolfer wrote a plugin called “Load as non-album track”, which strips all album tags and changes the value of the “Album” field to “[non-album track]” (you can change this string in Options → Metadata). But it only works on tracks that are already matched.
There is a MusicBrainz-compatible tagger called beets which can do what you’re asking. The option is called singleton mode. beets is a command-line programme. Since you’re concerned about how long tagging takes, you might like beets since it has an “autotag” mode, and it can resume if interrupted.
Back in Picard, I wonder if matching would speed up noticeably if you set the “Preferred release types” in Options → Metadata → Preferred Releases to (mostly) ignore any types that don’t seem to apply to your case.
Why?
The source is called MusicBrainz, not AlbumBrainz
There are pop song that were never originally on an album, and there are classical compositions in their own right that may have a duration of some 15 minutes.
Why would Picard need to be forced to look for albums for those, and then download all tracks from some album it just guessed the song came from, and that you have no interest in?
Why?
The source is called MusicBrainz, not AlbumBrainz
Picard is an album-oriented tagger—the first one ever, I believe. Before Picard, MusicBrainz first tagger was not album-oriented. I’m sure there has been some discussion about whether to retain a non-album mode, but I don’t know what the rationale was.
I stand (sort of) corrected.
While I do believe there is some value in my request, I just found out that Picard was not responsible for me growing a beard.
It was the wikidata-genre plugin.
Without it, there is no delay to speak of, and I would probably never have brought this up.
It’s only with this plugin that everything slows down to a snail-like pace.
I’ll try to contact the developer to see if something might be done about it.
If you want to manually edit something, that is up to you.
But when a search is done to identify a file…
Imagine you own a Porsche 911. You walk into a parking lot, are you going to take any Porshe 911, or do you want your Porsche 911.
There are “meta” differences between a file that comes from an album or box or radio edit or compilation or etc. The same as there are differences between your car and mine. That’s why a “search and identify” is going to find the original source.
Then we have ascertained an intrinsic difference between Picard and MusicBrainz:
If I am correct, the boys and girls of MusicBrainz live by the adagium ‘better no data than incorrect data’.
But in cases like these, Picard will be doing the opposite.
Wouldn’t that be a valid argument for adding the option to Picard to treat tracks as non-album tracks?
Although I never tried, this could be configured in Picards settings Preferred Releases, as I understand.
Move the slider for single to maximum (->right) and the slider for album to minimum (<-left) and Picard will match to singles instead of albums.
I just tried that with 20 tracks.
With both runs Picard matched them all with exactly the same albums.
All of them being (even a bit obscure) compilation albums.
So it doesn’t seem to do a job in even trying to find the original albums where the songs might have been released on originally. So what’s the use in that really?
But to be clear, the reason I started this thread was because of extreme delays occurring when matching singles.
Now that I know Picard was not to blame for that in the first place, this discussion has become pretty academic.
I will just unset album, albumsort, albumartist, MusicBrainz release ID’s, etc. for when I want to match single songs.
That’s possibly because that’s where the tracks came from/what they actually are (for instance the length or AcoustID doesn’t match the original album recording, in which case Picard should never try to match it to the wrong version, no matter what your sliders are set to).
Honestly though, if you only want the most basic of tags (just artist + track name) with no care for specific details, or matching to the specific correct recording, that’s simply not what Picard + MB aim for. A lot of other taggers cater to that need.
No, I don’t believe that’s true.
If I take single popular songs from an original album that I have, and then after removing all tags and then scanning it separately, Picard will usually say the song came from ‘The world’s greatest…’ , ‘Die besten 100…’, Hits from the 80’s’, etc. etc.
But it will not mention the actual original album it came from.
You misunderstand. It’s not about me not caring for detail.
Probably the contrary.
It’s about getting incorrect information from Picard.
But as I said, it’s now clear to me how this works, and it’s not a big bother.
When scanning a folder with separate tracks, I’ll just need to remember to disable Picard from writing all album related information.
Thanks for trying it without the tags on an album song!
That’s a real shame. MB has a lot of bad data/acoustID’s submitted to the wrong songs/recordings that need to be merged (I’m guessing that’s the issue), but that really requires elbow grease from users at the end of the day.
Sorry, I misunderstood, I thought you didn’t want the release information, so it wouldn’t matter which release it matches to.
)