Displaying featured artists as "feat. [Name]"

[Spymaster]

If your tracks contained the musicbrainz_releasetrackid tag, or even musicbrainz_albumid, I could possibly produce a Mp3tag update script to add the artist tags to your collection, but that wouldn't help on the Asset front.

I had a very quick look at MediaMonkey and it doesn't appear to support many of the MusicBrainz tags; out of interest, how did you get 'Artist1 feat. Artist2' into your tags as it seems to only support the more traditional 'Artist1; Artist2'?

I don't use musicbrainz at all really, other than as lookups while using dbpoweramp.

I enter "Artist1 feat. Artist2", literally in that format, into either the "Album Artist" or "Artist" field, upload it to my NAS, and Asset splits it easily into two separate artists for its database.

[Spymaster]

Wouldn't the easiest solution be for Asset to present its database for searching and then present the "Artist" field for accurate display? Wouldn't that be the best of both words, be mega easy and require no effort at the user's end?

[simbun]

Wouldn't the easiest solution be for Asset to present its database for searching and then present the "Artist" field for accurate display? Wouldn't that be the best of both words, be mega easy and require no effort at the user's end?

Not sure what you mean. Could you provide an example using an artist tag containing 3 artists?

[Spymaster]

Not sure what you mean. Could you provide an example using an artist tag containing 3 artists?

Sure, so the artist could be tagged as "Robbie Williams feat. Kylie Minogue; Tom Jones".

Asset would take that away and store each artist separately in its database as "Robbie Williams", "Kylie Minogue", "Tom Jones".

If you played the track now, based on what Asset sends back, it would display the artist as "Robbie Williams; Kylie Minogue; Tom Jones".

But if Asset pulled back the artist string attached to the file, rather than the artists as held in the database, it would display "Robbie Williams feat. Kylie Minogue; Tom Jones"

Similar to what Streammagic appears to be doing of its own accord.

[vilsen]

I think what you're asking can be accomplished in Asset already, without touching your existing ARTIST tags. You could use DISPLAYARTIST (used by MusicBee) and a pseudo tag (e.g. ARTIZT) like this:

1. Add DISPLAYARTIST to the tracks that need the "feat." treatment, like this:

ARTIST=Peter Gabriel
ARTIST=Kate Bush
DISPLAYARTIST=Peter Gabriel feat. Kate Bush

2. Add the pseudo tag "artizt" to the index list in Asset. Don't tag your tracks with this pseudo tag, because it will be indexed based on ARTIST (see below).

3. In the rename list, add this:

preformat:=artizt:=%artist%
display:=Artizt:=Artist
preformat:=artist:=[IFVALUE]displayartist,[tag]displayartist[],[multitag]artist[][]

4. When configuring the browse tree, use "Artizt" instead of "Artist". You will still see "Artist" when you browse (because of the rename list).

This will make "Peter Gabriel feat. Kate Bush" display whilst playing, but you'll still able to browse both artists separately.

[simbun]

2. Add the pseudo tag "artizt" to the index list in Asset. Don't tag your tracks with this pseudo tag, because it will be indexed based on ARTIST (see below).

I had considered something like this but doesn't it break search i.e. the results will be a listing of the collaborations, not the unique artist names?

[Spymaster]

I think what you're asking can be accomplished in Asset already, without touching your existing ARTIST tags. You could use DISPLAYARTIST (used by MusicBee) and a pseudo tag (e.g. ARTIZT) like this:

1. Add DISPLAYARTIST to the tracks that need the "feat." treatment, like this:

ARTIST=Peter Gabriel
ARTIST=Kate Bush
DISPLAYARTIST=Peter Gabriel feat. Kate Bush

2. Add the pseudo tag "artizt" to the index list in Asset. Don't tag your tracks with this pseudo tag, because it will be indexed based on ARTIST (see below).

3. In the rename list, add this:

preformat:=artizt:=%artist%
display:=Artizt:=Artist
preformat:=artist:=[IFVALUE]displayartist,[tag]displayartist[],[multitag]artist[][]

4. When configuring the browse tree, use "Artizt" instead of "Artist". You will still see "Artist" when you browse (because of the rename list).

This will make "Peter Gabriel feat. Kate Bush" display whilst playing, but you'll still able to browse both artists separately.

But that would still require a massive amount of retagging at the user's end. Every single track with a featured artist would need to be changed. That's thousands of tracks in my case.

Isn't it possible for Asset to simply read the artist from the file itself for display purposes? That data is right there already.

[schmidj]

But Asset, and most any decent library system doesn't read anything from the file(s) itself when you are searching. It indexes everything that it is programmed to be able to search for into a database when you first add the tracks or when you tell it to redo the library. Then when you search, it searches the index files in the database. Think about your old fashioned library with the card catalog at the entrance Drawers of 3X5 cards by let us say author. You could quickly look at the cards and find out where the book you want is. But without the catalog you'd have to start searching through every book to find the one you want, a slow process. If Asset had to search the files themselves, instead of a database searching of a large collection would be very slow.

[Spymaster]

But Asset, and most any decent library system doesn't read anything from the file(s) itself when you are searching. It indexes everything that it is programmed to be able to search for into a database when you first add the tracks or when you tell it to redo the library. Then when you search, it searches the index files in the database. Think about your old fashioned library with the card catalog at the entrance Drawers of 3X5 cards by let us say author. You could quickly look at the cards and find out where the book you want is. But without the catalog you'd have to start searching through every book to find the one you want, a slow process. If Asset had to search the files themselves, instead of a database searching of a large collection would be very slow.

But I'm not talking about searching, I'm talking about displaying.

I totally get why Asset would search its databased index of artists. What I'm suggesting is that after the search is complete, that it then pulls in the artist details from the file for display purposes while the track is playing.

[simbun]

Sure, so the artist could be tagged as "Robbie Williams feat. Kylie Minogue; Tom Jones".

Is that the approach you're using? Is that a MediaMonkey thing? The standard approach is to store the names multi-valued or as credited on the release; for example, "Robbie Williams; Kylie Minogue; Tom Jones" vs "Robbie Williams feat. Kylie Minogue & Tom Jones".

With your approach - and some additional development - it would be possible to present an accurate (one value per artist) index/search and present the original string in track listings, though how many users would benefit from that I don't know.

But that would still require a massive amount of retagging at the user's end. Every single track with a featured artist would need to be changed. That's thousands of tracks in my case.

If you start out with a multi-valued tag and simply replace the first delimiter with ' feat. ' then it probably wouldn't be that difficult to automate, though with 300,000 tracks it's still not trivial.

[simbun]

What I'm suggesting is that after the search is complete, that it then pulls in the artist details from the file for display purposes while the track is playing.

Wouldn't you also want it on track listings; for example, album view?

[Spymaster]

Is that the approach you're using? Is that a MediaMonkey thing? The standard approach is to store the names multi-valued or as credited on the release; for example, "Robbie Williams; Kylie Minogue; Tom Jones" vs "Robbie Williams feat. Kylie Minogue & Tom Jones".

With your approach - and some additional development - it would be possible to present an accurate (one value per artist) index/search and present the original string in track listings, though how many users would benefit from that I don't know.

If you start out with a multi-valued tag and simply replace the first delimiter with ' feat. ' then it probably wouldn't be that difficult to automate, though with 300,000 tracks it's still not trivial.

It's not a MediaMonkey thing - it's simply the only way that works if you want to show the relationship status AND have it work for Asset.

If I tag as "Robbie Williams; Kylie Minogue; Tom Jones" then the "feat." relationship is lost. If I tag as "Robbie Williams feat. Kylie Minogue & Tom Jones" then Asset with read "Kylie Minogue & Tom Jones" as a single artist. The way I do it is the only way to include the "feat." relationshop AND keep the artists separate.

in terms of how many would benefit, it might eliminate the issue of different players interpreting artists differently (e.g. many apps only read the first artist and ignore the rest) because the field from the file itself would always be constant and reliable.

[Spymaster]

Wouldn't you also want it on track listings; for example, album view?

That would be even better!

[vilsen]

I had considered something like this but doesn't it break search i.e. the results will be a listing of the collaborations, not the unique artist names?

Isn't it possible to search for a string in a field/value? I'm not able to test this at the moment, but I would be surprised if searching is limited to exact values.

[vilsen]

Sure, so the artist could be tagged as "Robbie Williams feat. Kylie Minogue; Tom Jones".

But that would still require a massive amount of retagging at the user's end.

It's your tags that are unusual. You're asking that Asset should interpret your non-standard ARTIST values. IMHO I don't see why Asset should do that. In general, if the server/player would fix user problems or other tag oddities it would make the code unnecessarily complex. In these cases it's much better that the user fixes the issues with the help of (advanced) tagging programs.