[Morphologie-devel] bug in midicents-to-name

Kilian Sprotte kilian.sprotte at googlemail.com
Fri Jul 20 14:52:22 UTC 2007


Hi,

sorry I am late ....

> I found a bug into "mc->alpha" method, doing an infinite loop :
> this method has to call not itself, but "mc->alpha1" function.
>
> And the method "midicents-to-name" is ok when replacing "MORPH2" by
> "MORPH" at the very end of the expression (last line).

I fixed exactly those two things, its in revision 24.

> Also, I'm not sure the symbols we use for naming the notes are all
> clearly understable or rational (CN, CS-, CS--, ...) : should we find a
> better representation with other more common/understandable symbols ?
> I'm going to simplify the mess of numtochar functions : I have found ie
> ten different functions or methods to do quite the same job... it should
> be unified.

At a first glance, I like the symbols. :) But I will look at them more closey.
Just by playing around, I think I have found a small problem:

(morph:mc->alpha 6300 2) gives me
"ES4"

(I liked the idea of using the german system !!)

But then trying (morph:mc->alpha 7000 2)
I got "BF4" - and I think that's good.

So probably the above is a bug...
I will have a look now and write again, if I find and change anything....

Best,
  Kilian



More information about the Morphologie-devel mailing list