In classical music, the composer and the performer (musician) are two different individuals, and a single piece has many different interpretations and recordings by different people - both the composer and the performer are what builds a unique recording.

In current osu! convention however, the Artist field is just taken by the individual artist (example: beatmapsets/1187093#mania/2474006). This is confusing, as the actual composer is hidden in the tags. A parallel could be drawn to modern songs where a composer writes songs for an artist, and then the artist name (usually singer) is used in the Artist field (or Composer feat. Artist) - however I believe the classical situation is different enough to warrant a specific indicator, as the artist and the composer are completely separated and independent, a performance is not a "cover" as there isn't an "original recording" or "original release".

Proposal:


Example: In the previously linked beatmap, the Artist field would be "Maurice Ravel (perf. Jean-Philippe Collard)".

This is not a common situation in osu! as there aren't many classical pieces being mapped due to their timing difficulty, but some examples do exist and I believe it would be good to include this standardization, as the current situation is not well handled by the existing Ranking Criteria.

EDIT after some discussion: Added "Various Performers" edge case.

Absolutely vouch for this. With music like classical pieces or anything that is meant to be interpreted by numerous different individuals, there's an absolute need to have a mechanism to 1) not mislead players about who composed the song and 2) give proper and equal credit to performer and composer, which is one of the core ideas of interpretation.

Sounds good.

While I still have my doubts on what would qualify for this convention as it is meant for classical music alone and a whole grey area might open up, I believe this is a proper standarization to properly credit the people who originally created a composition which, more often than not, will have a myriad of different interpretations.

So I vouch for the idea even if it might need to specify more stuff.

neat.

yeah

this applies to stuff like covers as well

PuigdemontLoL wrote:

While I still have my doubts on what would qualify for this convention as it is meant for classical music alone and a whole grey area might open up, I believe this is a proper standarization to properly credit the people who originally created a composition which, more often than not, will have a myriad of different interpretations.

So I vouch for the idea even if it might need to specify more stuff.

Speaking of specifying more stuff, what kind of "grey area" would this open up?

It was brought up in the mapping hub and even here with mint's words, but where do you draw the line between something being a mere performance and an entirely new work based on previous works? Even Midi performamces were brought up, and now covers as said by mint. What to do with those? Do we treat them under the same umbrella? Those are the questions to be answered to avoid the most confusion.

But do not misunderstand, I do fully back this idea due to how important should be to properly credit both the original composer and the respective performer.

PuigdemontLoL wrote:

It was brought up in the mapping hub and even here with mint's words, but where do you draw the line between something being a mere performance and an entirely new work based on previous works? Even Midi performamces were brought up, and now covers as said by mint. What to do with those? Do we treat them under the same umbrella? Those are the questions to be answered to avoid the most confusion.

But do not misunderstand, I do fully back this idea due to how important should be to properly credit both the original composer and the respective performer.

I don't think this is much of a contention point personally;

Covers and arranges typically involve some form of re-writing of a previously commercially released song. Even if they are performed the same as the original song, they are still different from a classical composition being performed. Definitely not a legal expert by any means, but I think they are also considered differently on a copyright standpoint.

Think of it as a playwright & actors in theatre; Is the Macbeth written 4 centuries ago the same as the one played nowadays? As far as I'm aware, yes, though I suppose there is some differences of interpretation from one troupe to another when it is performed on stage. The same goes for composed classical pieces of yore; What Bach or Beethoven wrote hundreds of years ago is still what is being played today, even if there can be some interpretation based on the musicians/conductors performing the piece.


I'm curious as to what you mean exactly by "MIDI performances"? Depending on what you have in mind, this could go either way.

I'm thinking about MIDI performances are, something that you play on digital audio workstations (DAWs) by simply putting the MIDI and interpreting how the notes are played, how loud each notes are, and how the tempo would be like.

Regardless, I'm agree with this proposal.

Definitely agree for the most part given how vast interpretation can be for classical pieces that sometimes don't get as much credit as they should.

why only classical and not all utaite type things

Nao Tomori wrote:

why only classical and not all utaite type things

That's because utaite songs are covers. If the instrument and the backing sound is literally the same, it's simply just a cover. This is not a case for live-performed songs (or rather, in this case, pieces), because it may be interpreted differently than what it would be by the composer.
See this reply instead for the better explanation: community/forums/posts/8496904

what if cover has new instrumental as well? those also happen

A case-by-case examination is needed if it's difficult to interpret.

Yes, covers are already covered in the current RC - the tagging is what the cover artist uses. I agree it can be misleading sometimes, but what should be used is what the artist intends to, as other parts of metadata. In case of classical music, it is convention that both the performer and the composer are prominently featured in every release - this makes metadata for it challenging not only in osu!, but in a lot of music platforms. This proposal is a solution for osu! to this common problem in music tagging.

I think covers are on a completely diferent spectrum than what this rule is intended to address. Something being a "cover" implies there was an original release by the original artist, whereas in classical music there simply isn't. This means there's no way the composer could ever get any Artist credit even across thousands of ranked performances of the same song, and what makes having a mechanism like this so important.

In that sense, we could be wording it something something like:

Proposed RC wrote:

If the song's original release isn't the recording by the original artist (such as classical music pieces), both Artist and Performer must be creddited with a `Artist (Perf. Performer)` notation in the Artist field.

+1

I'm down for some improvements to the handling of classical music

i am going to cover la campanella

I was writing something, but it looked dumb. Basically, I agree with this proposal since it helps clarifies who the composer is better than having them in the tags.

Sounds good

No issues with the initial proposal after discussing a few edge cases like having multiple performers to the point where we can't fill the artist field with every performer. I wasn't sure about what the metadata of MIDI recordings of classical music was going to be either, but I think that was clarified quickly; the MIDI creator becomes the performer.

That said:

Proposed RC wrote:

If the song's original release isn't the recording by the original artist (such as classical music pieces), both Artist and Performer must be creddited with a `Artist (Perf. Performer)` notation in the Artist field.

I wouldn't be in support of this, mostly because there are many composers who perform their own music and the original release often times include the composer themselves as the performer alongside many others.

Quick examples but obviously far from exhaustive:
https://rateyourmusic.com/release/album/terry-riley/in-c/
https://rateyourmusic.com/release/album/philip-glass/glassworks/
https://rateyourmusic.com/release/album/michael-tilson-thomas-ralph-grierson-steve-reich-roger-kellaway-tom-raney/three-dances-and-four-organs/

That said, there are also cases that I intuitively think using just the composer is correct if the performer and composer are the same (https://rateyourmusic.com/release/album/max-richter/the-blue-notebooks/ ), so maybe it's better to look at this on a case-by-case basis.

There are also cases where the composer/songwriters and the performers don't overlap in music either (I think pop music does this more often), and I don't think it's necessary to credit non-classical music a la Artist (perf. Performer) because the song was intended to be performed in the way that the songwriters intended, while classical pieces have some sort of improvisation. Song covers aren't part of the discussion because there's an existing song that the cover was based on in the first place.

My main issue with using "classical pieces" to word the rule is that it can have a bit unclear boundary as to when to stop using the marker or when to start doing so. Is there any other possible wording we can come up that makes it a bit more clear? (or at least a "case-by case basis" clause, as well as a clause that forbids song covers from using this marker as you say, which is necessary)

Do other services have a standard or generally consistent way of using this kind of performer-composer notation? We should probably try to use that if at all possible.

Ephemeral wrote:

Do other services have a standard or generally consistent way of using this kind of performer-composer notation? We should probably try to use that if at all possible.

There is no set standard that I know of for classical music. The most typical ones I know about:

1. Spotify uses "Composer, Performer(s)"
2. Youtube uses only "Performer(s)" and relies on the fact that most performers indicate the composer in the name and covers of the album which is easily visible.
3. RYM has a dedicated Composer field that is listed next to the Performer in the listings.

Since we lack both dedicated composer and album fields, I feel going with the more explicit marker like Artist (perf. Performer) is probably a good idea, but there may be some standardisation out there in the internet I'm not aware of (I'm not too fond of Spotify's implicit naming because it's pretty misleading at times when there's multiple performers and composers, when the composer isn't easily recognisable, etc.)

As quenlla said, this is a problem classical music metadata faces basically everywhere, spotify, youtube, etc. because of having a single artist field, which sometimes makes music hard to identify - you can be listening to something on Spotify and have no idea who the composer is, it's not well handled by modern music databases I'd say. This proposal is an attempt at circumventing this issue standardizing the way to put both performer and artist into the one field.

What about pieces where original composers are not known? (commonly happens in folk songs)

What about pieces where original composers are not known?

Could do the same as currently when the artist isn't known, "Unknown Artist" - or maybe could use "Anonymous" / "Unknown Composer". I think such cases are quite rare in osu! so it could be considered when it happens.

pwhk wrote:

What about pieces where original composers are not known? (commonly happens in folk songs)

This probably fits well enough under the existing "Unknown Artist" rule. Conversely it's also interesting to consider the opposite case ("Unknown Performer" could be a thing)

I somewhat disagree with this proposal. I actually quite liked it initially and I do agree with you on the notion that we should be crediting performers/composers somewhat equally instead of throwing the composer into the tags.

However, I disagree with this mostly on the basis that I don't think '(perf. Performer)' is an *official* indicator on how we should label performers metadata-wise, and only seems to band-aid this crappy situation. It's just rather odd, considering that I have never personally seen anyone do this to credit a performer like so. The things that were mentioned such as Spotify and YT are extremely notorious for their horrible mishaps with metadata, and RYM which is more or less equivalent to a fanmade wiki for music are reasons why I don't thinking using them in this scenario would be ideal, especially YT and Spotify where there's typically no standards for any sort of metadata on these platforms.

If anything, I think this should be implemented through the webpage interface, similarly to how we do genres and it would be like a fill in the black if applicable for this piece or smth. I would imagine this would be better since I think would work a lot better and it would be less intrusive and more official imo, whilst giving proper credit to the composer in question. And I know this is less ideal to work with and way more of a hassle, but since we're working with an indicator that's pretty much unofficial, I just can't help but think that this just isn't a full-proof way of working around with this.

If you have any more official indicators on sites that can be used as full-proof metadata for this specifically, then I would encourage you to link some here.

having a composer field in metadata box and making that visible on site would definitely be the best option but kinda doubt its happening anytime soon

This was brought up in NAT as well, I've been thinking about it since then

Generally:

People think it'd be beneficial to have some way to indicate it

However, perf. does not seem like the ideal way to do this to me.

- its meaning may be ambiguous
- when it does and does not apply can be hard to define, and would be hard to follow on the broader scale of all of osu!. It seems like a great idea for small set cases, but trust me, if it's not water tight when it does and doesn't apply it will be very confusing for others and other discussions.
- current allowances do allow for formatting with multiple artists already, but some points were brought up about it not being clear enough.


So instead, I think broadening the scope of an existing allowance would be better.

This one in particular:
For Remixes/Covers, the original artist may be used in the artist field, as long as the title field is modified to clearly show that the song is remixed.

I think could simply be changed to

For Remixes, Covers, or Performances, the original artist may be used in the artist field, as long as the title field is modified to clearly show that the song is not the original.

This would allow stuff like
Composer - Title (Musician's Performance)

Within already existing allowances, and without adding an extremely niche+specific guideline.

I agree that adding a specific rule for such a niche case may not be the best solution, when it could be interpreted in different ways and cause confusion - and maybe as you suggest modifying an existing allowance is better. Having the musician in the title field as you say seems a bit weird to me, but if it makes it fit better within the osu! ecosystem it may be a solution.

Musician in title field seems like a very inelegant solution IMO

is merge