GENERAL DISCUSSION - Level at which the proportion applies

[martha-thomae]

Regarding mensurations & proportions, we have divided the issue into two parts:

1. How to encode the difference between the sign that appears on the page and the actual meaning of that sign? This is discussed here.
2. How to encode at which level the proportion is acting (i.e., at the level of the semibreve, breve, etc.)

Here we are discussing the second part of the issue. So far, there have been two suggestions about how to encode the level at which a proportion is acting:

• Proposal A: Encode something similar to @level that has the values semibreve, breve, longa and so on.
  @annplaksin since this was your proposal, would you mind giving the exact details? The name Anna used is @affected.
• Proposal B: Encode @tempus.num and @tempus.numbase when the proportion affects the relation of breve-semibreve, or @prolatio.num and @prolatio.numbase when the proportion affects the relation of semibreve-minim, and so on.
  @craigsapp would you mind summarizing the details of this proposal?

---

Tagging some people into this discussion, please feel free to include anyone else:
@lpugin, @DILewis, @giulianodibacco, @annplaksin, @craigsapp

[martha-thomae]

One thing I can see is that in Proposal A we have one new attribute with pre-existing values, while in Proposal B we have to add many attributes. However, Proposal B can provide a more accurate encoding if different ratios apply at different note levels.

My question is: Is this possible? Can different ratios apply at different note levels given a particular mensuration/proportion sign? (@annplaksin you might be the best to answer this question)

[martha-thomae]

In that case, I think we could go with proposal A to avoid adding several attributes, but just one with pre-existing values (namely, "semibrevis", "brevis", and "longa").

[lpugin]

Agree. And @level looks like a good name to me.

[annplaksin]

Just adding a bit more on use and processing:

In a case of a C| and 3 sesquialtera we can process it like this:

	C|	proportion	3
Mx	4		4
Lg	2		2
B	1		1
Sb	0.5	x * 2/3	0.3333
Mi	0.25		0.16....

@level gives you the note level that will be affected, you can multiply the old value with @num \ @numbase and retrieve the new relative note value and change all the smaller ones according a map of the note level relations...

For implicit changes of mensuration -- in this case changing from tempus imperfectum to tempus perfectum -- we still need to add @tempus="3"

[DILewis]

So to be clear, This table is for <mensur level="semibrevis".... If it were <mensur level="brevis", the only difference would be the value for the 3rd row?
This seems intuitive to me.

[annplaksin]

Jep... the @level attribute tells you on which note level you should apply @num / @numbase

[martha-thomae]

See discussion above: @lpugin, @annplaksin, @craigsapp :)