Re: Streamlining my thoughts on the clarinet woodwind-diagrams

[Wim van Dommelen]

On 7 Feb 2013, at 18:46 , Joseph Rushton Wakeling wrote:

> On 02/07/2013 02:26 PM, Wim van Dommelen wrote:
> > clarinet-family --> clarinet (what we have now, but without the  
> > "hole")
>
> "hole" .... ?  I guess you mean the extra touchpiece on the LH 1st  
> finger?
Internally it is coded as "hole", "h" :-)

> Ideally you'd like to have some backwards compatibility, so I  
> suggest keeping "clarinet" for the base stencil.
Yes, of course.

> >     - clarinet --> clarinet-with-left-low-gis (obviously adds the  
> > left-hand
> > low-gis key)
>
> Call it clarinet-lh-gis.
Mmmh, that (lh . (gis)) is already taken for the upper key, using the  
same, not completly describing name will again confuse others.
> >         - clarinet-with-left-low-gis --> bass-clarinet (adds the  
> > "hole" and
> > right low-ees)
>
> I'd use clarinet-full-boehm as the name for the clarinet with low  
> ees, and make bass-clarinet an alias for that (or else, bass- 
> clarinet is the full-boehm instrument plus the "hole").
The regular (soprano-)clarinet doesn't have a low-ees, so that order  
will not work.

> >             - bass-clarinet --> low-bass-clarinet (adds left low-d,  
> > right
> > low-d, left low-d, thumb low-c and thumb low-cis)
>
> I'd call it bass-clarinet-low-c.
Compatibility in the name?

> >                 - low-bass-clarinet --> low-bass-clarinet-BC- 
> > Prestige (adds
> > thumb low-d)
> >
> >                 - low-bass-clarinet --> low-bass-clarinet-S- 
> > Privilege (adds
> > thumb low-ees)
>
> I wouldn't bother with the "low" here. :-)
Yeah, but they (BC and S) have similar models running only to the low  
ees, same series, same name, matching the regular stencil, not the  
low. So which confusion is "better" explained.

> > The "clarinet-with-low-gis", "bass-clarinet" and "low-bass- 
> > clarinet" will then
> > be "intermediate" stencils, but otherwise complete and callable  
> > from the
> > outside, which is fine for example for writing down a specific  
> > thing in the high
> > registers or a special effect in a general fashion. The generic  
> > stencil for the
> > bass-clarinet" will suit the Buffet Crampon and Selmer current top- 
> > models to
> > low-ees and will do for specific notations in the high register for  
> > all of them.
> > All the key variations are in the low range as far as I know.
>
> Question: how do you translate the bass clarinet diagrams into key- 
> name diagrams?
I don't understand the complete question. Is that something other than  
the result you get with (the already present) \override #'(graphical .  
#f) as stated in the reference manual?

> > More variations can be done, e.g. one could add: "clarinet-with- 
> > left-low-gis -->
> > clarinet-BC-Tosca" specifying the special F correction key for the  
> > right little
> > finger. etc.
>
> Yes, but that seems to be taking it a little far.  It's difficult to  
> imagine a situation where a composer would want to specify such a  
> key, because its existence and practical effect cannot be relied upon.
Isn't a chicken-and-egg combination? I spoke to my teacher today and  
she got the question of a composer "tell me what is possible?" as  
opposed to someone writing down something and the player having to  
find out how to do it. As the key corrects the lowest f (of a soprano- 
clarinet), it will probably not used/prescribed. But if some effect,  
whatever can be done with it, someone will use it sometime.

But it is a little (much) too far, I agee.

> > If really needed it should be possible to insert an older or newer
> > system also. The problem is that all the really usable variations  
> > are typically
> > brand/model variations, I'm struggling with the naming of it.
>
> This may be handled better by documentation describing how to create  
> your own custom diagram, rather than trying to support many.
That is not that easy, it is not ly-code, but Scheme code and tougher  
(more and more stacked upon another) as I expected. The basic files  
involved are together already 3200 lines (some are comments, but 90%  
is code :-).

Regards,
Wim.