Conflate contours?

[krhes]

Dear All,

A new puzzle I’m trying to work out is attached as patch and image:

vertical to horizontal.omp (33.2 KB)
As you can see, what I would like is to create a new contour from the intersection of two curves (if one were to hear the top-left chord-seq and transcribe it, one would hear a clear direction for the top line, and bottom line at the least). How might one formalise (and solve) this problem? If I’ve not made the problem clear I’m happy to try again!
As ever, many thanks in advance for any help/suggestions,
Krhes

[haddad]

Dear Krhes,

The simplest solution in your case, is to deal with the chord-seq results, ie not to go into geometrical calculations which of course would be ideal, but would require more subtle algorithms.
So to make it easy, we will use merge on your first two chord-seq, and filter the highest pitch like so:

(the patch):
Patch 5.omp (24.3 KB)

Hope this will satisfy your needs for this problem.

Best
K

[krhes]

Dear Karim,

As ever, thank you. Changing list-max- to list min in the omloop I can also work out descending lines (which is very helpful). I wish there was a quicker way than conflating 2 lines at a time for a descending or ascending line, but that’s beyond me. I can see the string
((4200) (4200) (4300) (4400) (4700) (4900) (4300) (3400) (2200) (800))
consisting of the lowest values from these four curves:
(4200 4200 4300 4400 4700 5100 5600 6300 7200 8400)
(5000 5000 5100 5300 5600 6100 6700 7600 8800 10200)
(5600 5600 5700 5900 6300 6800 7600 8600 10000 11600)
(6000 6000 5900 5700 5400 4900 4300 3400 2200 800)
And can also do that with the other lines/voices. Some sort of band-filter to bound a value range?
I’m after something like this:

But as I said, that’s beyond me.
Time to rinse my eyes I think!
Anyhow, thank you again.
Best wishes,
Krhes

[haddad]

Hi again,

The simplest thing to do in this case (ordered by register) is the following:

The patch:
Patch 6.omp (20.8 KB)

best
K

[krhes]

Dear Karim,

I sit here marvelling at the elegance and simplicity of your solution (as well as cursing my own inadequacy!). Thank you once again.
Best,
Krhes.