In Stiller Nacht

[Daniel Wolf, TD 441.7:]
>After the renaissance some form of adaptive JI seems only to become
>current again with the emergence of chamber music without continuo,
>especially the string quartet and quintet (the latter being Mozart's
>prefered genre). Unaccompanied vocal music will wait even longer to
>re-emerge, and indeed becomes especially problematic to tune (try, for
>example, Brahm's _In stiller Nacht_).

I did, thanks to Mark Nowitzky, TD 447.4, and I took that "good ol'
equal-tempered MIDI format" and ran it through the wringer, and...
it sure sounds vivid! Daniel, the subtleties of tuning choices that
you discuss in TD 442.5 aren't understood by the program, so (and
perhaps for other reasons as well) you may find the sound far from
ideal, but I'd be interested in your reaction (luckily, the piece is
short!).

As an alternative to the piano, I also used a string voice and a
french horn voice, each in separate files. All this in one 15K zip:

http://www.idcomm.com/personal/jadl/bra-stil.zip

JdL

Sorry to be so slow! This one took some digging to answer...

[Mark Nowitzky, TD 450.15:]
>Looks like a fun basement project you got going there...

Yes, it keeps me busy, and the incoming flack keeps me from being bored
(I guess, if I dish it out, I've gotta take it!).

>I liked the french horn files the best!

Kyool; I was shooting blind in choosing voices. Most synths, even good
ones, make a horrible mess of the "human" voices, so I skipped'em.

>I've got some feedback/questions/comments on the tunings you chose for
>the Brahms piece:

>1) I viewed your 5-Limit Horn MIDI file (bra-stil-hrnz5.mid) under an
>editor, and saw a lot more pitch wheel values than I expected to see.
>How come the tonic (key) of the piece, Eb, drifts around so much? I
>understand the need for more remote tones to drift, such as F or C in
>this piece. But why does your Adaptive JI algorithm move ALL the notes
>on EVERY chord change?

Good question! The answer is that I first decide how to tune each of
the pitch12's (the pitch, C ... B, with octave information removed)
relative to 12-tET, and then decide how best to "center" the resulting
chord. That centering motion is often strongly influenced by notes
(pitch12's) that are continuously sounding and are instructed to try not
to change pitch much.

More specifically, the piece is in the key of Eb, so, why not hold Eb
fixed? Part of the answer is that my program doesn't "know" that the
piece is in Eb (though it does "know" that Eb chords begin and end it!),
so doesn't know which pitch12 to hold fixed. But even if that were
known, to hold the Eb fixed would mean that other pitch12's would have
to move more, which can mean more total "pain" according to the way I
weight things.

The question of ideal centering of chords is a very difficult one, and
I'm becoming more convinced, with help from feedback on this list, that
my methods don't do a very good job some of the time. More on that in a
later post!

>2) The second chord in the piece is Ab Cb Eb F (an Ab minor triad with
>an added sixth). Here's how you tuned it (in bra-stil-hrnz5.mid), and
>how I would tune it:
>
> your choice my choice
> ----------- -----------
> NNN* ratio NNN* ratio
> ---- ----- ---- -----
> 4F 10/9 4F 10/9
> 5Eb 1/1 5Eb 1/1
> 5Cb 64/81 6Cb 4/5
> 5Ab 2/3 5Ab 2/3
>
> *where NNN = Nowitzkian Note Names (per
><http://www.pacificnet.net/~nowitzky/justint/nnn.htm>)
>
>Why do you choose the more Pythagorean tuning for the Cb, instead of
>opting for the simpler 4/5 ratio? Is the presence of the 4F = 10/9
>affecting your choice? Or would you also tune a vanilla minor triad
>that way too?

Well, I'll be @#$%^& if that isn't the same chord as in the Bach piece
that I analyzed in TD 405.2, in a different key of course. Here, as
there, my program picks an unexpected chord: in this case, it calls
it a Db 9 with missing root! Therefore, the Cb is tuned to 8/9 of Db
simply because that's the ratio I put in my tuning file (I'm taking your
pick, as I recall, for when there really is a dom7 or dom9). As you
point out, this makes it a Pythagorean third from Eb.

In 7-limit tuning, the Db 9 (with root missing) sounds really lovely,
IMHO, but in 5-limit it's probably flat out a bad choice. In the Bach,
Paul E recommends, as you do, tuning the minor triad properly, and I
agree (in 5-limit only!).

>3) Around the fifth measure of the piece, there's a Bbdom7 chord (Bb
>Ab D F). Here's how you tuned it in your 7-Limit Horn sample
>(bra-stil-hrnz7.mid):
>
> note NNN ratio
> ---- --- -----
> F 5F 9/8
> D 4D 15/8
> Ab n/a 21/16
> Bb 5Bb 3/2
>
>I suppose that if I ever used septimal harmony (7-Limit tunings), maybe
>I'd tune it the same way you did.

And, as you may know, my habits and my ears are tuned to septimal! (but
I've gotta get better at satisfying you 5-limit guys...).

>I was surprised as I listened and compared this to the 5-Limit tuning.
> The 7-Limit version sounded more "fluttery". (By "fluttery", I mean
>I heard rapid beats; it felt dissonant.) I would've thought that the
>7-Limit tuning would be more smooth.

Yeah, I'm not sure why you experienced that!

>So once again I question my own theory about dominant seventh chords
><http://www.pacificnet.net/~nowitzky/justint/dom7.htm>. Pat Missin's
>harmonica samples, discussed at the bottom of that web page, had just
>the opposite effect.

Do you get the same sensation in the other voices? I'm mystified; my
own experience with 7's is that they're smooth, oh so smooth. In their
own uniquely twangy kind of way, of course...

JdL