History:Performance Restructuring Proposal

From MusicBrainz Wiki
Revision as of 21:59, 9 December 2005 by BrianG (talk | contribs) ((Imported from MoinMoin))
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigationJump to search

Proposal

Tree Structure

performed {additional} {guest} {instrument} (as {role}) on
performed {additional} {guest} {vocal} vocal (as {role}) on
 performed {additional} {guest} background {vocal} (as {role}) on
 performed {additional} {guest} choir {vocal} (as {role}) on
 performed {additional} {guest} speech (as {role}) on
  {additional} {guest} narrated
  {additional} {guest} read
orchestra performance... (untouched)

Attributes

"Vocal" is selected from a drop down menu as one of the following:

 [select vocal]  (selectable if no additional vocal type is needed)
 Classical tones  (non-selectable group-item)
  Alto
  Baritone
  Bass
  Contra-tenor
  Mezzo-soprano
  Soprano
  Tenor
 Recitative
  Rap
 Some other items such as African chants

As you can see here, the vocal tones can all be just a vocal, a background vocal or a choir vocal. If this is correct for all of them I don't know. Perhaps the different ARs can have different drop down lists (choir would not make much sense for rap for example).

The attribute "Role" is set by a text field. For narration and reading this is not appropriate since both reader and narrator are outstanding persons who are just defined as having no role on a story.

Difference between reader and narrator:

A reader reads an audiobook to us (he also reads the spoken parts of characters). A narrator tells us the story but he does not speak the spoken parts of characters in the story. Either they are spoken by others (in radio plays - then "performed speech" is used for them) or he tells us what they said (re-narrate).

About the attributes "additional" and "guest" I'm not sure at the moment.

Examples

Jim Dale read Harry Potter and the Sorcerer's Stone (feat. narrator: Jim Dale) (disc 1) ("feat. narrator" is actually wrong here! It should be "feat. reader" or something)

Sting narrated Peter and the Wolf (feat. narrator Sting)

SomeName performed Oboe as The Duck on Peter and the Wolf (feat. narrator Sting)

Wallu Walpio performed speech on Introduction by Wallu Walpio

Kent Broadhurst performed speech as The Hypnotherapist on Regression

Discussion

Now rip my proposal into pieces. :)



OK, here we go:

Why don't you add a sub-type "lead" to vocal performance? This way the main type would really mean: "Some vocals, but i don't know how exactly".

Also, it seems to me, some of these sub-types could also be represented by an attribute like {vocal-role} which would then be either lead, background, or choir. I think the main point you show, is that vocals should be categorized along two dimensions: {vocal-role}, meaning how the vocalist relates to the rest of the band (lead, background,...), and {kind-of-voice}, meaning the vocal "instrument" or style the vocalist uses (alto, rap,...). I would prefer having these as two independent attributes to a single AdvancedRelationshipType.

Finally, note that your {role} attribute is a neat idea, but currently unimplemented. There cannot be a text-field to AdvancedRelationships. --DonRedman

  • Yes about the lead vocal you seem right. And the two dimensions is exactly what I wanted. But I think this way it's better. It's an easier step from the current implementation to this one. I don't know if it's possible to have two drop down lists in AR. Also I think it would be easier two have separate drop down lists for vocal/background/choir this way. But well, I'm not the developer. And about the {role}: Yes I know it's not possible at the moment. My proposal can be implementet completly without this but it is just an idea for future developments which the developers might consider some day. --Shepard
    • So, I propose:
performed {additional} {guest} {instrument} on
performed {additional} {guest} {vocal-role} {kind-of-vocal} vocal on
 performed {additional} {guest} speech on
  {additional} {guest} narrated
  {additional} {guest} read
orchestra performance... 
  • With attributes
 {vocal-role}
  lead
  background
  choir
 {kind-of-vocal}
  Alto
  Baritone
  Bass
  Contra-tenor
  Mezzo-soprano
  Soprano
  Tenor
  Rap
  • It somehow feels strange that speech is a sub-type to vocal (which thematically it is), but does not share the same attributes, but that is the best I can come up with right now. Also note, that IIRC there cannot be unselectable elements within attribute lists (or can they if the description is left blank as in AR Types? I'd have to ask Robert).--DonRedman
    • If not then we still have no way to say "I only know he performed vocals but I don't know what kind of" which is one big point I wanted to correct. Also "Recitative" is not unselectable in my proposal. --Shepard


ok, as discussed on the mailing list when this has come up, i think rap is a really bad role to include. it is not a concrete style, like the classical ones (which refer to a specific vocal range), and a lot of rap type singing strays between the 'real' singing, and rapping, if not a combination of the two. it's just a horribly grey area and there's no real need to make the distinction. i mean we don't define any other type of contemporary singing - eg 'growling' for death metal, 'whining' for radiohead, etc :) (all of which i find equally silly btw!). everything contemporary should be kept as 'lead'. --Gecks

  • I agree that it is perhaps not correct to group vocal styles together with vocal ranges/tones. Nonetheless I would like to have this AR to be a bit more descriptive. I mean we have tons of instruments now but only a few vocal styles. It was proposed somewhere to use SurvivalOfTheFittest on the instruments to see which we need and which we can combine and whatever. I think somehow they are also a grey area but we do implement them!
    • Well an instrument is different, because you know categorically what instrument is used (invariably from the liner notes) whereas a singing 'style' is almost completely subjective these days. I don't consider the two areas similar :) --Gecks--
      • You run issues without the "rap" option with groups that have multiple vocalists where one's role is singing while the other's role is rapping. THC!! has three vocalists - a female singer, a male singer, and a female rapper. ハレンチ☆パンチ also has 3 - one does lead singing, second does rap and some backup singing, and the third does just chorus singing. m-flo has Verbal as rapper and various guest vocalists for the actual singing. CRIMSON has a lead singer/samisen player and a rapper/harp player for their two vocalists. With all of them its divided into the one(s) doing the lead singing and the other doing the rapping; putting all of them under "lead vocals" simply feels inaccurate. Rapping in general might not be as clearly defined as the different roles in classical singing but with groups that have both singers and rappers it's very clear in their specific cases. It'd be more useful to the people with that music to have that info in the database than to simply lump all the vocalists together as having performed lead vocals. --DJKC
    And one thing: what do you think about recitative? Keeping it in my proposal and perhaps building it in with a separate link type or something? I'm not quite sure if I used a good translation for what I wanted to say.. in german we have "Sprechgesang" - directly translated this would be "speech-singing". My dictionary said recitative but when I translate that back to german it says either "Sprechgesang" or "Rezitativ" (which is not what I want). Compare http://en.wikipedia.org/wiki/Recitative and http://en.wikipedia.org/wiki/Sprechgesang. --Shepard

I don't know exactly how to propose a structure, but there should be a way to mark relationships as "ArtistY performed {additional}(and/or){guest} vocal on SongX" without having to select kind-of-vocal or vocal-roll. --GURT


Original author: Shepard