Re: [mb-users] Rare transcriptions

2012/10/4 Frederic Da Vitoria <[email protected]>

> 2012/10/4 symphonick <[email protected]>
>
>>
>> 2012/10/4 Frederic Da Vitoria <[email protected]>
>>
>>> 2012/10/3 symphonick <[email protected]>
>>>
>>>>  2012/10/3 Frederic Da Vitoria <[email protected]>
>>>>
>>>>> 2012/10/3 symphonick <[email protected]>
>>>>>
>>>>>> I had planned to do CSG works structure as my next RFC anyway, so
>>>>>> let's see if we can agree on something here.
>>>>>>
>>>>>> Different versions of a work - when should a new work be created?
>>>>>>
>>>>>> 2012/10/2 Nicolás Tamargo de Eguren <[email protected]>
>>>>>>
>>>>>>>
>>>>>>> On Wed, Oct 3, 2012 at 12:33 AM, Alex Mauer <[email protected]>wrote:
>>>>>>>
>>>>>>>> On 10/02/2012 03:05 PM, Frederic Da Vitoria wrote:
>>>>>>>> > I am always annoyed by this kind of edit: this is not a recording
>>>>>>>> of Clair
>>>>>>>> > de lune but of a transcription of it. IMO this edit is currently
>>>>>>>> valid,
>>>>>>>> > because there is no other way to enter it. I know we decided we
>>>>>>>> shouldn't
>>>>>>>> > create a new work for each transcription, and I agree that in
>>>>>>>> cases such as
>>>>>>>> > this one it would be overkill and it would clutter the database
>>>>>>>> with Works
>>>>>>>> > with only one Recording. But wouldn't adding a "transcription"
>>>>>>>> attribute to
>>>>>>>> > the Recording-Work AR be a good idea?
>>>>>>>>
>>>>>>>> I would say they should be separate works, and we should improve our
>>>>>>>> display of related works.
>>>>>>>>
>>>>>>>> I just ran into BWV 645 which is a transcription for organ of BWV
>>>>>>>> 140 so
>>>>>>>> obviously the traditional catalogs do this at least in some cases.
>>>>>>>>
>>>>>>>
>>>>>>> Personally, I'd agree with making all transcriptions separate works
>>>>>>> (and making works include instrumentation info). Of course, that gets messy
>>>>>>> sometimes when we don't know what a specific recording is, so we'd need
>>>>>>> some kind of catch-all work for those... :/
>>>>>>>
>>>>>>
>>>>>> "Wachet auf" is a good example. There's the original chorale by
>>>>>> Nicolai (1599), two versions by JS Bach, and lots of
>>>>>> transcriptions/arrangements that can be based on Bach's organ version or
>>>>>> the orchestral version - probably impossible to tell in many cases. Also
>>>>>> the same arrangement (say for violin & piano) can of course be performed on
>>>>>> many different instruments.
>>>>>>
>>>>>> How should this fit together?
>>>>>>
>>>>>> 1) Wachet auf, ruft uns die Stimme by Philipp Nicolai ("earliest
>>>>>> version"?)
>>>>>> 2) Zion hört die Wächter singen by JS Bach (Tenor Choral from BWV 140)
>>>>>> 3) Wachet auf, ruft uns die Stimme, BWV 645 by JS Bach (organ
>>>>>> transcription of 2)
>>>>>> 4) Wachet auf, ruft uns die Stimme by JS Bach (Coro from BWV 140)
>>>>>>
>>>>>> and 3 random arrangements:
>>>>>>
>>>>>> A1) Wachet auf by JS Bach (possibly based on 3)
>>>>>> http://www.amazon.com/Canadian-Brass-More-Greatest-Hits/dp/B000003EO3
>>>>>> A2) Wachet auf, ruft uns die Stimme by JS Bach (probably an
>>>>>> orchestral transcription of 2)
>>>>>> http://www.amazon.com/Sleepers-Awake-Cantata-No-140/dp/B001BEEC76
>>>>>> A3) Piano arrangment by Wilhelm Kempff
>>>>>> http://www.deutschegrammophon.com/cat/single?sort=newest_rec&COMP_ID=BACJS&PRODUCT_NR=4790314&ART_ID=KEMWI&flow_per_page=50&UNBUYABLE=1&per_page=50&presentation=flow(DG website has a title based on BWV140, a cover scan of another release
>>>>>> had "Organ prelude")
>>>>>>
>>>>>> Thoughts?
>>>>>>
>>>>>
>>>>> And we could add
>>>>> http://musicbrainz.org/recording/ba5543c9-30fb-4ed5-aa07-10161dc3e993, although it has one foot outside the scope of classical, we'll have to
>>>>> decide how to handle this type of situation too.
>>>>>
>>>>>
>>>> Yeah, it makes sense to have a separate work for this. But I wonder
>>>> what happens if more "jazz" versions exist of Wachet auf by other artists?
>>>> Would "normal" MB want to have only one work, but CSG wants one for every
>>>> arranger/artist?
>>>>
>>>>
>>>>> My first thought is that the idea of creating a Work for each
>>>>> transcription has the advantage of being simple. I like simplicity. It
>>>>> would probably mean removing Transcriber AR at least from Classical music.
>>>>>
>>>>> Some users feared that it would trigger the creation of too many Works
>>>>> which would be seldom used. Do we have a way to measure the current
>>>>> situation? Counting the Arranger ARs to Recordings?
>>>>>
>>>>>
>>>> That's likely. We will also need a way to show arranger in the search
>>>> results, or else we must repeat the arranger in the disambiguation field.
>>>>
>>>
>>> Is this is worse than the current situation?
>>>
>>
>> Sorry, I don't quite understand. What do you mean with "the current
>> situation"?
>>
>
> I only meant what we have now. Currently, how does arranger appear in
> search results.?
>

Not at all.


>
>
>
>>  & what should we do when we dont know what version is being used?
>>>> example: Handel - Messiah in original version & Mozart's arrangement.
>>>> Should there be a super-work?
>>>>
>>> Messiah (super-work)
>>>> -Messiah (original)
>>>> -Messiah (Mozart arr.)
>>>>
>>>
>>> Wouldn't creating super-works be better than current situation? Users
>>> will be aware that there is a choice to be made only if the database offers
>>> them a choice. If only Handel's Messiah was in the database, some users
>>> would probably use it in place of the others.
>>>
>>>
>>
>> Yeah, there's always a risk of that works get misidentified. Perhaps it's
>> better with a super-work that you're encouraged to use if unsure about with
>> version you have.
>>
>>
>>>
>>> & what to do if I know a work is a transcription, but not who wrote it?
>>>> Or when it could be a transcription, but it could also be the original
>>>> score with different instruments?
>>>>
>>>
>>> I think there should often (always?) be a Work for unspecified
>>> transcriptions, if there isn't we end up with the current situation with
>>> Myra Hess' Bach transcriptions which are probably a mess: users with a
>>> release with a piano transcription tend to use Hess' transcription because
>>> it is the only transcription currently in MB. This is probably often the
>>> correct choice, but still I'd prefer if we had a "unspecified transcriptor"
>>> transcription of those works so that users would have a choice and
>>> hopefully use "unspecified" transcription Work when they don't have any
>>> indication.
>>>
>>> I thought I'd prefer to use the super-work w. transcription ARs for
>> everything unspecified. But how would this work when there are subparts
>> involved? If we have
>>
>> Super-work: Bach cantata 147
>> *Work: Bach cantata 147
>> **Part of 147: Jesus bleibet meine Freude
>> ***Jesu, Joy of Man's Desiring (Myra Hess solo piano transcription)
>>
>> An unknown piano transcription can't be connected to the cantata
>> super-work. Maybe every work needs a super-work:
>>
>> Super-work: Bach cantata 147
>> *Work: Bach cantata 147
>> **Super-work: Part of 147: Jesus bleibet meine Freude
>> ***Work: Part of 147: Jesus bleibet meine Freude
>> ***Work: Jesu, Joy of Man's Desiring (Myra Hess solo piano transcription,
>> also connected to super-work?)
>>
>> Now a recording of "Jesu, Joy of Man's Desiring" with an unknown
>> arrangement links to the super-work with (recording) arranger [unknown]. Or
>> we could have a work for all unknown arrangements. Would it be a work for
>> every instrument? I suppose that could get messy, performers could be using
>> the same arrangement for different instruments, but we would split it into
>> several works based on a performance choice?
>>
>
> Good question. Maybe we could set a limit around here, this is close to
> what I believe is currently being used. But once again, how frequent are
> these? Wouldn't we be introducing unnecessary complexity (in the sense that
> the user would have to decide first if he needs to create a Work or not)
> for a few edits? I really would like numbers. Given the current complexity
> of creating a simple Work hierarchy into MB, would requiring a Work for
> each transcription be such a big deal?
>
>
I can't give you numbers, sorry. (Numbers on what?) If we introduce
super-works, the basic structure will always be like this:

Super-work
*Work
**Part foo super-work
***Part foo work
**Part bar super-work
***Part bar work

and so on.

-- 
> Frederic Da Vitoria
> (davitof)
>
> Membre de l'April - « promouvoir et défendre le logiciel libre » -
> http://www.april.org
>
>
> _______________________________________________
> MusicBrainz-users mailing list
> [email protected]
> http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-users
>



-- 

/symphonick

_______________________________________________
MusicBrainz-users mailing list
[email protected]
http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-users