Another NEW barbershop JI recording!

Posted below is an e-mail I sent to some barbershop tuning people, including
Dr. Jim Richards, physicist and author of "The Physics Of Barbershop Singing"
and Dave Mohr, a Just Intonation Network member and physicist by trade.

Anyway, rather than retype everything, I think it's all said below.

If you have trouble with line-wrapping readability, click the little
"unwrap lines" link at the top of the posting (if you are viewing on
the website).

Enjoy! Check the score too.

Sincerely,
Aaron

-----

Greetings BHS tuners,

I've completed a second recording!

See www.darktownsaints.com/shine

There you will find a part tape, an analyzed score, and two bonus mixes of the Polecat -
"Shine On Me"

My new mixed quartet (me and my fiance and two other more beginner barbershoppers)
are learning it so I made a tape. This piece is also one of tuning interest as arranged.

Back in 2003, I analyzed the whole polecat book, but there were some things I was
confused about and some uncertain issues. Dave Mohr, you brought this one specifically
to my attention.

Anyway, the tenor-lead minor 3rd on LIGHT-HOUSE, goes from being the top two notes of
a major chord (as 5:6 ratio) to being the top two notes of a seventh chord (a 6:7 ratio if
tuned as a barbershop seventh). Most people (myself included) agree that it sounds weird
and awkward to hear the lead sharp and the tenor flat as they collapse in on each other.
BUT, and here's the revelation: it is barely heard at all when in the full chord context! The
tenor becomes this background part and the note is less of a scale-note identity and more
of a chord-note identity. The ear hears the note having its purpose from the chord. The
lead almost sounds the same, within the chord context. I think the average listener would
not identify any pitch shift in the lead, within the quartet sound.

Dave originally argued that most likely the bass and bari would tune to an unchanging
lead, and the tenor would also stay, and we'd have a 5:9 ratio between bass and tenor,
giving a 20:25:30:36 chord. While this is possible (and I suppose it'd be interesting to
actually make all these comparisons, which I COULD get around to doing), I feel it is best
as I've done it. I've preserved the root motion and relative harmonic motion, and I've also
got all barbershop sevenths and no compromises in the harmonic content.

What I did was record the bass and then the lead. I was decently in tune, but honestly, I'm
no gold-medal singer, even if I have my moments (though I think I have potential with
some hard work...). And the chance my best moment shows up when the recording is
going is very unlikely, especially since I'm doing all the recording and can't just focus on
singing. So much as my takes were not bad, I had to chop up some takes to create a
decent performance. I also had some wavering and a few off pitches that were simply due
to lack of vocal support. So I then centered the bass and lead on the analyzed notes, but
did not adjust the wavering or vibrato. I then sang tenor and bari along with that. I found
that quite a decent portion of the tenor and bari notes that I sang were off by less than 2
or 3 cents, some were right on (meaning Melodyne analyzed that the center of all the
wavering and vibrato was at a certain pitch). Still, I had some support issues.

So the file shine_on_me_natvib.mp3 is the first complete step. It is a relatively unedited
recording. All the parts were moved to center on the correct pitch, but all my unsupported
wavering and vibrato is left in.

Then I edited to a mild amount of vibrato, did a couple other edits, darkened the timbre by
lowering the formants a little until it sounded as natural as I could get it. The result is the
normal part tape, the full mix being the Q file.

Finally, I made a version with NO vibrato, just for comparison. This one sounds less
natural, but is definitely appealing in a way.

Overall, I've learned especially that "naturalness" of the sound is as or more dependant on
natural levels of vibrato and timbre and tones over the course of notes than on pitch. I'm
really looking forward to what I'll find with the Realtime track(s) now.

Any questions, comments, feedback are welcome. More to come eventually.

In Harmony,
Aaron

Aaron Wolf wrote:
>
> > Anyway, the tenor-lead minor 3rd on LIGHT-HOUSE, goes from being the top two notes of > a major chord (as 5:6 ratio) to being the top two notes of a seventh chord (a 6:7 ratio if > tuned as a barbershop seventh). Most people (myself included) agree that it sounds weird > and awkward to hear the lead sharp and the tenor flat as they collapse in on each other. > BUT, and here's the revelation: it is barely heard at all when in the full chord context! It is obviously masked by the parallel chromatic descent in another part which is a barbershop clichee. I strongly suspect (but don't have any way of checking) that the chromatic and teh comma descent always go together.

And it reminds me of my habit, developed while playing alone, of using little inflections to imply harmonic motion and which has endeared me to so many fellow musicians. (Of course I have no idea whether I did that correctly, but neither have my fellow musicians :O).)

klaus

Great work, Aaron!!!

I took Jim Richards' class at Harmony College, in '96 I
believe.

Thanks for this version of one of my favorite polecats.

-Carl

Very Cool!

It was great to get the detailed track info, too!

I definitely like the novib version. Not better, necessarily, but
very interesting.

Thanks, Aaron!

-P

I forgot to add (though it may be noticed without pointing it out):

In response to Gene's post about JI not sounding like actual quartets,
I think my natvib version of Shine On Me does sound more like a
normal quartet. In other words, I think the simple wavering and
vibrato are enough to have less lock and sound more like a typical
quartet. I don't in any way think that's good though. And I think
it shows that there's no reason to doubt that a normal quartet is
still aiming for JI (if they are good enough to know what to aim for
at all).

-Aaron

--- In tuning@yahoogroups.com, "Aaron Wolf" <backfromthesilo@...> wrote:

> In other words, I think the simple wavering and
> vibrato are enough to have less lock and sound more like a typical
> quartet. I don't in any way think that's good though.

I agree that there's no reason sounding like a typical quartet, even a
typical really good quartet, is necessarily the goal. However I will
observe that maybe locking less often would tend to do that--locking
on long sustained chords especially, and in particular the final chord.