Atelier Démocratique « Non-occidental key signatures »
 Welcome, Guest.
 You can read all messages, but to be able to post,
 please Login or Register.
Nov 24th, 2017, 7:03am 
   Atelier Démocratique
   New views & staff types

   Non-occidental key signatures
« Previous topic | Next topic »
Pages: 1 2 3  Reply | Notify of replies | Send Topic | Print
   Author  Topic: Non-occidental key signatures  (Read 9412 times)
Question: Do you support the proposal below?

Yes     11 (78.5%)
No     3 (21.4%)

Total votes: 14

Please read this before voting

     

Jean-Armand Moroni
Board Senior Member
****






   


Gender: male
Posts: 1438
Re: Non-occidental key signatures  
« Reply #30 on: Aug 20th, 2009, 12:24am »
Quote | Modify

As a matter of fact, Scala is already dealt with in this request:
http://www.myriad-online.com/cgi-bin/workshop/YaBB.pl?board=file;action= display;num=1131530337
 (Jzeitlin spotted this)
 
Interestingly enough, in the discussion quoted, the first reply deals with "Hey, you should also implement scales with a number of notes per octave different from 12".
 
It seems that there are at least 4 distinct matters, in increasing complexity (as far as I can guess):
 
- Support of non-tempered scales with 12 notes per octave, without any change during the piece, with notes having the same name from an octave to the next, and with an octave being the 2:1 ratio. This is already possible (correct me if I am wrong), but an interface with Scala files would be appreciated.
 
- Support of non-tempered scales, with 12 notes per octave, but with an octave being slightly different from 2:1 ratio. This is frequent in tuning concert pianos, so it is an important case.
 
- Support of scales with a number of notes per octave different from 12. This would use non-occidental signs (e.g. turkish commas), which would be dealt with like # and b instead of as an effect on each note.
 
- Support of scales with notes having a different name from an octave to the next one, the name replication interval being different from an octave. I fear this would have impacts everywhere in the software, unless a trick is possible. Such as combining the two above features: "the next note with the same name is not in a 2:1 ratio" and "there are more than 12 notes between two notes that bear the same name".
 
In every case above, I assume that, from the software point of view, there are still only 7 different note names, but with more alterations than just # and b. We could also assume that each note name has its own line or space. This means that when you raise 3.5 lines on the staff, you get the note with the same name, even if it is a note with a 3:1 ratio.
offline
jzeitlin
Board Newbie
*





   


Posts: 8
Re: Non-occidental key signatures  
« Reply #31 on: Aug 20th, 2009, 4:04am »
Quote | Modify

on Aug 20th, 2009, 12:24am, Jean-Armand Moroni wrote:
In every case above, I assume that, from the software point of view, there are still only 7 different note names, but with more alterations than just # and b. We could also assume that each note name has its own line or space. This means that when you raise 3.5 lines on the staff, you get the note with the same name, even if it is a note with a 3:1 ratio.

 
No, don't make this assumption - my favorite recent example, the Bohlen-Pierce scale, actually has nine distinct note names in a thirteen-"semitone" spread on a framing interval of 3:1 - the "most standard" option for naming notes in B-P uses the letters A-J, skipping I - a B-P scale in C-lambda would be called as C-D-E-F-G-H-J-A-B-C'.  The four accidentals would be C#/Db, F#/Gb, H#/Jb, and A#/Bb (and yes, they're enharmonic).  Kees van Prooijen has defined some B-P modes which only use 7 note names, but those also have the possible complication of sharp and flat accidentals between a pair of notes not necessarily being enharmonic, such as G# and Ab in his so-called "Major" mode.
 
(Side note/question: Is there some way of getting a translation of comments not in one's native language?  I have no French at all, and feel that I might well be "missing" something in not being able to follow the threads that are in French - and I might expect that anyone who does not have any English would feel the same way about threads in English.)
offline
Jean-Armand Moroni
Board Senior Member
****






   


Gender: male
Posts: 1438
Re: Non-occidental key signatures  
« Reply #32 on: Aug 20th, 2009, 11:17pm »
Quote | Modify

If there are more than seven note names, it could be a stronger modification to the software than anything else I listed. Establishing the specifications would be especially tricky, I bet.
 
As for automatic translation, well well well... a nice counter-example has been posted today in this thread:
http://www.myriad-online.com/cgi-bin/bbs/YaBB.pl?board=news;action=displ ay;num=1250778556
 
where "Perhaps just a softer and less dense organ mix to feature the horn melody will work for you."
 
is automatically translated into "Peut-être juste un doux et moins dense à l'organe de mélange caractéristique de la corne mélodie travaillera pour vous.",
 
which means "Perhaps just a soft and less dense in the mixing device specific to the (cattle) horn melody will labour for you."
 
In fact all automatic translations I have seen in this forum are utterly comic.
offline
jzeitlin
Board Newbie
*





   


Posts: 8
Re: Non-occidental key signatures  
« Reply #33 on: Aug 21st, 2009, 2:18am »
Quote | Modify

on Aug 20th, 2009, 11:17pm, Jean-Armand Moroni wrote:
If there are more than seven note names, it could be a stronger modification to the software than anything else I listed. Establishing the specifications would be especially tricky, I bet.

 
I would consider that likely.  Nevertheless, I would also consider it a worthwhile modification to be made to the program, since - if done correctly - it would vastly increase both the general capability and the flexibility of the program.
 
I thought that part of the problem might have been with supporting this sort of functionality for MIDI export, but a look at the specification - or, more accurately, several summarizations or interpretations of various sections thereof - shows that there ARE provisions in the specification for setting the MIDI Note Number:Frequency relationship, both in bulk (all 127 notes in one data dump) and individually, and outside of the 'pitch bend' mechanism that appears to be the only current option in HA for such adjustment. Whether this would require HA to know about specific MIDI devices, as opposed to it being a universal command, was not clear in what I read.
 
on Aug 20th, 2009, 11:17pm, Jean-Armand Moroni wrote:
As for automatic translation, well well well... a nice counter-example has been posted today in this thread:
http://www.myriad-online.com/cgi-bin/bbs/YaBB.pl?board=news;action=displ ay;num=1250778556
 
where "Perhaps just a softer and less dense organ mix to feature the horn melody will work for you."
 
is automatically translated into "Peut-être juste un doux et moins dense à l'organe de mélange caractéristique de la corne mélodie travaillera pour vous.",
 
which means "Perhaps just a soft and less dense in the mixing device specific to the (cattle) horn melody will labour for you."
 
In fact all automatic translations I have seen in this forum are utterly comic.

 
Automatic translation quite often is ... interesting.
 
I perhaps overstated the case when I said I have no French; I can generally pick out cognates to English words, and have a vague idea of the grammar, from (a) having studied another Romance language (Spanish) and (b) taken courses called "English through Latin and Greek" and "History of the English Language", all back in the Paleolithic era of my education.  So, between knowing the context of the discussion, and being aware of the sorts of errors that automatic translation makes, that translation, odd as it is, would, together with a French original, probably be sufficient for me to get a general idea of what was intended.  The problem is finding a source for those automatic translations.
 
I often wonder whether the problems are greater in going FROM English than TO English, because English has words imported from so many other languages, and many synonyms and homophones - perhaps a far larger sample of each than most other languages.  That, however, would be a discussion for another forum entirely.
offline
Jean-Armand Moroni
Board Senior Member
****






   


Gender: male
Posts: 1438
Re: Non-occidental key signatures  
« Reply #34 on: Aug 26th, 2009, 10:21pm »
Quote | Modify

on Aug 21st, 2009, 2:18am, jzeitlin wrote:

I thought that part of the problem might have been with supporting this sort of functionality for MIDI export, but a look at the specification - or, more accurately, several summarizations or interpretations of various sections thereof - shows that there ARE provisions in the specification for setting the MIDI Note Number:Frequency relationship.

 
In this case you would still have 12 notes per octave. Of course one could greatly modify the base frequency of each note, so that you put more notes per octave. But:
- The number of octaves you have available would be lowered, as there is a fixed number of MIDI notes, 127. With a 53 notes per octave scale, this would make only 2 octaves and a half.
- There could be some limitation in the frequency modification that is allowed; either in the MIDI specification or in the instruments that implement it. To give an example, if you implement a 53 notes per octave scale, you would have to rise some notes many octaves up. The frequency modification could be made (in the instrument) by using the same sample and modifying its tempo in order to modify the frequency. Such a solution would be convenient for a frequency modification of less than a half-tone, but would produce dreadful results if applied to large modifications
offline
jzeitlin
Board Newbie
*





   


Posts: 8
Re: Non-occidental key signatures  
« Reply #35 on: Aug 27th, 2009, 2:57am »
Quote | Modify

on Aug 26th, 2009, 10:21pm, Jean-Armand Moroni wrote:
In this case you would still have 12 notes per octave. Of course one could greatly modify the base frequency of each note, so that you put more notes per octave. But:
- The number of octaves you have available would be lowered, as there is a fixed number of MIDI notes, 127.

 
Yes; this is inherent in the MIDI specification.  My point was that currently, the HA capability for retuning is documented as using the MIDI 'pitch bend' mechanism, which explicitly leads to problems if one wishes to place two notes whose adjustments are different onto the same staff, to be played simultaneously.  The retuning mechanism that I discussed in my previous post avoids that particular problem, though at the possible cost of reducing the range.
 
on Aug 26th, 2009, 10:21pm, Jean-Armand Moroni wrote:
With a 53 notes per octave scale, this would make only 2 octaves and a half.
- There could be some limitation in the frequency modification that is allowed; either in the MIDI specification or in the instruments that implement it. To give an example, if you implement a 53 notes per octave scale, you would have to rise some notes many octaves up. The frequency modification could be made (in the instrument) by using the same sample and modifying its tempo in order to modify the frequency. Such a solution would be convenient for a frequency modification of less than a half-tone, but would produce dreadful results if applied to large modifications

 
The specification did not appear to limit how much a particular tone could be adjusted using the mechanism I discussed; I did not look into any limitations on the pitch bend mechanism. I believe that MIDI does have an absolute range on frequency, but that would depend strictly on the precision of the data and how it is interpreted - I would have to go back and look to be sure of the exact numbers.
 
In practice, however, I would tend to believe that the 88 notes of the piano (seven-plus octaves) represent the widest range commonly used for instrumental music, and that vocal music by the best-trained vocalists is probably only about four octaves in the approximate center of that range.  The average person with limited singing experience would probably be lucky to achieve much more than two octaves.  Given that, I would expect that the 127 notes of the MIDI specification will in general prove sufficient for most purposes, even when tuned to a "high-density" set of frequencies (i.e., much more than quarter-tone, or 24 steps per octave).
 
* Note: I am using the term 'octave' to refer to a doubling of frequency, not to refer to any particular musical structure or note count.
 
A potentially more difficult question arises when it comes to notation, and composition, using the current HA models: At present, HA only allows two types of staff - the "conventional" Western five-line staff, or the Gregorian four-line staff. Both assume the standard twelve-note-per-octave scale when interpreting what note to play.  For tunings that use framing intervals of other than 2:1, or for tunings that use a number of notes within the framing interval other than 12, this could cause difficulty.  One notation that I have seen, that I do not like for composition, but which works, is apparently called 'piano roll', and it is used by the Tonescape program, and also as an explanatory notation by Elaine Walker in her B-P research (PDF file!). Support for this notation model may be necessary if the retuning capability is accepted; to maintain the ease of use that HA currently enjoys, I would recommend that some method of matching notes on the 'piano roll' with positions on a user-defined staff (which the user would use for composition if the piano roll was deemed inadequate) be added as well.
 
This request, if done properly, is certainly worth the five-star rating that Olivier gave it when he opened this thread, and possibly more.  Nevertheless, I still believe it should be done.
offline
Pages: 1 2 3  Reply | Notify of replies | Send Topic | Print

« Previous topic | Next topic »

« Atelier Démocratique » Powered by YaBB 1 Gold - SP 1.1!
YaBB © 2000-2002,
Xnull. All Rights Reserved.

Top of page
Legal information Last update:  (c) Myriad