"well-temperaments sound alike"

Johnny asserted here, way back on Jan/8/07 (and I'm just now catching up with archives, sorry):

> Now,
> all well-temperaments sound more alike to modern day ears than any
> one well-temperament sounds in comparison to ET. Due to the
> circumstances surrounding the publishing of a popular treatise on
> the organ and a companion for tuning it, unlike anything of its
> kind, at least until Neidhardt in 1706, the musical basis for
> choosing a well temperament goes to Werckmeister III. There is
> nothing else justifiable, in my opinion.

Just wait a doggone minute, here, if the minute hasn't already passed. What "modern day ears" universally, or even severally, would affirm that all of the following "well temperaments" sound even enough like one another to be confusable?

- Neidhardt's 1732 "5th-circle #11" (my illustration of it is here:
http://www-personal.umich.edu/~bpl/larips/00NeidhardtFifthCircle11.jpg
and download the complete 1732 set from:
http://harpsichords.pbwiki.com/f/Neidhardt_1732_Charts.pdf
http://harpsichords.pbwiki.com/Tuning
...the one that has E-G# very obviously wider than Ab-C, but also has Bb-D wider than F#-A#, and Eb-G wider than B-D#!)

- Neidhardt's 1732 "5th-circle #5" from that same set...the one that has C-E wider than either of E-G# (narrowest) and Ab-C (middle); C-E is wider than its counterpart in equal!

- Neidhardt's 1732 "3rd-circle #1, Village" from that same set...whose C-E is only size 2, and whose F#-A# and Ab-C are 10 each, but Db-F and B-D# are only 9 each...therefore crossing against Dr Lindley's usual premise that Db-F has gotta be the biggest...even though it one was of Neidhardt's own favorites (that time) enough to have given it his "Dorf" recommendation.

- Mine ( http://www.larips.com ).

- Kirnberger 2 (the *published* one in the 18th century, and widely cited by his encyclopedist compatriots thereafter...the one with D-A-E tempered 1/2 comma each).

- Werckmeister (the common "3" one that Johnny and some others apparently enjoy/prefer, while to me it's well-nigh unlistenable in music that goes beyond two flats or sharps....)

These various "well temperaments" have their major-third size peaks in different places, and different balance around the circle, and yadda yadda yadda...enough such that at least *this* pair of "modern day ears" would never confuse them, given a good 15-minute test with a nicely assorted batch of intervals, in flowing musical context. Let's say by playing Bach's four duetti 802-805, for such a test, with their naked two-part texture to present these differences....

Brad Lehman
http://www.larips.com

Brad,

Hi there...good to hear from you. Why not give us some MIDI files or
otherwise of these Duetti. I personally am big on putting these ideas
to the test, especially after seeing others who are so convinced of
their differences get them all wrong!

I'd personally someday like to test your ears by seeing if you could
identify your reconstructed tuning from among the usual suspects. I'd
tend to be more of a believer than a skeptic, which is firmly what I
am at this point.

Granted, like you point out, many WTs are very similar in character in
certain keys---the true test of their differences would be to hear all
of them over a wide range of keys, like you suggest. The most critical
things would happen in keys with 4 or more accidentals, I think.

In the meantime, I'd love to hear your response to listening to
Herman's post:

/tuning/topicId_69073.html#69073

-A.

--- In tuning@yahoogroups.com, Brad Lehman <bpl@...> wrote:
>
> Johnny asserted here, way back on Jan/8/07 (and I'm just now
catching up
> with archives, sorry):
>
> > Now,
> > all well-temperaments sound more alike to modern day ears than any
> > one well-temperament sounds in comparison to ET. Due to the
> > circumstances surrounding the publishing of a popular treatise on
> > the organ and a companion for tuning it, unlike anything of its
> > kind, at least until Neidhardt in 1706, the musical basis for
> > choosing a well temperament goes to Werckmeister III. There is
> > nothing else justifiable, in my opinion.
>
> Just wait a doggone minute, here, if the minute hasn't already passed.
> What "modern day ears" universally, or even severally, would affirm
that
> all of the following "well temperaments" sound even enough like one
> another to be confusable?
>
> - Neidhardt's 1732 "5th-circle #11" (my illustration of it is here:
> http://www-personal.umich.edu/~bpl/larips/00NeidhardtFifthCircle11.jpg
> and download the complete 1732 set from:
> http://harpsichords.pbwiki.com/f/Neidhardt_1732_Charts.pdf
> http://harpsichords.pbwiki.com/Tuning
> ...the one that has E-G# very obviously wider than Ab-C, but also has
> Bb-D wider than F#-A#, and Eb-G wider than B-D#!)
>
> - Neidhardt's 1732 "5th-circle #5" from that same set...the one that
has
> C-E wider than either of E-G# (narrowest) and Ab-C (middle); C-E is
> wider than its counterpart in equal!
>
> - Neidhardt's 1732 "3rd-circle #1, Village" from that same set...whose
> C-E is only size 2, and whose F#-A# and Ab-C are 10 each, but Db-F and
> B-D# are only 9 each...therefore crossing against Dr Lindley's usual
> premise that Db-F has gotta be the biggest...even though it one was of
> Neidhardt's own favorites (that time) enough to have given it his
"Dorf"
> recommendation.
>
> - Mine ( http://www.larips.com ).
>
> - Kirnberger 2 (the *published* one in the 18th century, and widely
> cited by his encyclopedist compatriots thereafter...the one with D-A-E
> tempered 1/2 comma each).
>
> - Werckmeister (the common "3" one that Johnny and some others
> apparently enjoy/prefer, while to me it's well-nigh unlistenable in
> music that goes beyond two flats or sharps....)
>
>
> These various "well temperaments" have their major-third size peaks in
> different places, and different balance around the circle, and yadda
> yadda yadda...enough such that at least *this* pair of "modern day
ears"
> would never confuse them, given a good 15-minute test with a nicely
> assorted batch of intervals, in flowing musical context. Let's say by
> playing Bach's four duetti 802-805, for such a test, with their naked
> two-part texture to present these differences....
>
>
> Brad Lehman
> http://www.larips.com
>

>Brad,
>Hi there...good to hear from you. Why not give us some MIDI files or
>otherwise of these Duetti. I personally am big on putting these ideas
>to the test, especially after seeing others who are so convinced of
>their differences get them all wrong!

I'm curmudgeonly enough, and experienced enough with testing these things on harpsichords, to insist that the testing is next-to-worthless unless it takes place on instruments that Bach actually knew: harpsichord and organ. Modern piano tone (whether real or synthesized) doesn't have anywhere near the upper-partial content where the conflicts really stand out.

And MIDI fake harpsichord or organ tone just doesn't sound convincing enough...nor do fake organ tones do what real organ pipes do, which is to "draw" slightly better into tune on sustained pitches.

It's not merely a *listening* experiment. It's a *playing* experiment of playing the pieces with one's head right there next to the soundboard (or the ranks of pipes), at various tempos, plus noticing how the tensions in the tuning affect one's timing and phrasing as a performer.

As for "giving" anything, what more can I do than I've already done, which is to record these Duetti on organ:
http://www-personal.umich.edu/~bpl/larips/cd1002.html
and make the samples of that album freely available too; click through "disc 2" there.

Brad Lehman
http://www.larips.com

On 1/17/07, Brad Lehman <bpl@umich.edu> wrote:
> And MIDI fake harpsichord or organ tone just doesn't sound convincing
> enough...nor do fake organ tones do what real organ pipes do, which is
> to "draw" slightly better into tune on sustained pitches.

Quibble: A MIDI file doesn't contain any timbre information; that all
depends on the synth used to render it, which can vary over the whole
range of quality.

The organ pipe thing is an excellent point though, the pipes are
coupled together so they perturb each other and can even lock in
phase, if the interval is close enough to just to begin with. I wonder
if anyone's ever tried to mimic it with software?

Keenan

Hey,

Fair enough, but Brad--all else being equal, are you saying you
couldn't recognize your tuning among others if they were on
synthesizer (or sampler or physically modelled instrument)?

I'm get suspect of all the large claims people make about various WTs
when the first thing that happens when I propose an actual _hardcore_
listening test is that people start back-pedalling and making excuses
as to why they wouldn't be able to hear such-and-such an effect. It's
like you can *feel* people getting nervous about making a mistake
publicly!

As for upper partials, like I said, we don't have to go realistic--a
sawtooth timbre on a synth is great for upper partials, and they're
perfectly harmonic like on a harpsichord. I dont see why you wouldn't
take the challenge---one certainly hears the differences between
temperaments, all else being equal. The question is, could we identify
our favorite from among them, randomly? I too prefer acoustic music,
but c'mon, let's not hide our bold claims behind hot air about
synthesized tones.

I'm willing to bet the farm that anyone who took the time to actually
tune a harpsichord or organ to Brad's satisfaction (or Johnny's or
whoever's) in the comparison set I proposed would still experience the
same general result---if they are confused by synthesized tones, they
will be also confused by acoustic ones.

Also, the reverse I think would be true--if it's easy to recognize
synthetically, it would be easy acoustically (assuming of course that
amazingly accurate acoustic tuning was done).

I'll never forget a basic experiment I did with my wife. She's a
popcorn fanatic, and always complained when I bought generic Jewel
popcorn. She wanted Orville Redenbacker all the time, only Orville. So
one day, without her expecting, I cooked up two bowls, one Jewel, one
Orville, and gave her a simple blind test, and you know what? She
prefered the Jewel! So now she says---always Jewel popcorn!

My father-in-law did the same with my mother-in-law. My mother-in-law
claimed Starbucks from the store was better than brewing Starbucks at
home with their beans. She was wrong, because she couldn't tell the
difference. We concluded that the *irrational experience* of going to
Starbucks, and having a paper disposable cup was comforting and
rewarding for her, so she preferred it!

Get my drift?

-Aaron.

--- In tuning@yahoogroups.com, Brad Lehman <bpl@...> wrote:
>
> >Brad,
> >Hi there...good to hear from you. Why not give us some MIDI files or
> >otherwise of these Duetti. I personally am big on putting these ideas
> >to the test, especially after seeing others who are so convinced of
> >their differences get them all wrong!
>
> I'm curmudgeonly enough, and experienced enough with testing these
> things on harpsichords, to insist that the testing is next-to-worthless
> unless it takes place on instruments that Bach actually knew:
> harpsichord and organ. Modern piano tone (whether real or synthesized)
> doesn't have anywhere near the upper-partial content where the
conflicts
> really stand out.
>
> And MIDI fake harpsichord or organ tone just doesn't sound convincing
> enough...nor do fake organ tones do what real organ pipes do, which is
> to "draw" slightly better into tune on sustained pitches.
>
> It's not merely a *listening* experiment. It's a *playing* experiment
> of playing the pieces with one's head right there next to the
soundboard
> (or the ranks of pipes), at various tempos, plus noticing how the
> tensions in the tuning affect one's timing and phrasing as a performer.
>
> As for "giving" anything, what more can I do than I've already done,
> which is to record these Duetti on organ:
> http://www-personal.umich.edu/~bpl/larips/cd1002.html
> and make the samples of that album freely available too; click through
> "disc 2" there.
>
>
> Brad Lehman
> http://www.larips.com
>

'tis but the acid test of tuners! Claims vs Reality...

Oz.

----- Original Message -----
From: "Aaron Krister Johnson" <aaron@dividebypi.com>
To: <tuning@yahoogroups.com>
Sent: 18 Ocak 2007 Per�embe 1:38
Subject: [tuning] The popcorn and Starbucks analogy (was Re:
"well-temperaments sound alike")

>
> Hey,
>
> Fair enough, but Brad--all else being equal, are you saying you
> couldn't recognize your tuning among others if they were on
> synthesizer (or sampler or physically modelled instrument)?
>
> I'm get suspect of all the large claims people make about various WTs
> when the first thing that happens when I propose an actual _hardcore_
> listening test is that people start back-pedalling and making excuses
> as to why they wouldn't be able to hear such-and-such an effect. It's
> like you can *feel* people getting nervous about making a mistake
> publicly!
>
> As for upper partials, like I said, we don't have to go realistic--a
> sawtooth timbre on a synth is great for upper partials, and they're
> perfectly harmonic like on a harpsichord. I dont see why you wouldn't
> take the challenge---one certainly hears the differences between
> temperaments, all else being equal. The question is, could we identify
> our favorite from among them, randomly? I too prefer acoustic music,
> but c'mon, let's not hide our bold claims behind hot air about
> synthesized tones.
>
> I'm willing to bet the farm that anyone who took the time to actually
> tune a harpsichord or organ to Brad's satisfaction (or Johnny's or
> whoever's) in the comparison set I proposed would still experience the
> same general result---if they are confused by synthesized tones, they
> will be also confused by acoustic ones.
>
> Also, the reverse I think would be true--if it's easy to recognize
> synthetically, it would be easy acoustically (assuming of course that
> amazingly accurate acoustic tuning was done).
>
> I'll never forget a basic experiment I did with my wife. She's a
> popcorn fanatic, and always complained when I bought generic Jewel
> popcorn. She wanted Orville Redenbacker all the time, only Orville. So
> one day, without her expecting, I cooked up two bowls, one Jewel, one
> Orville, and gave her a simple blind test, and you know what? She
> prefered the Jewel! So now she says---always Jewel popcorn!
>
> My father-in-law did the same with my mother-in-law. My mother-in-law
> claimed Starbucks from the store was better than brewing Starbucks at
> home with their beans. She was wrong, because she couldn't tell the
> difference. We concluded that the *irrational experience* of going to
> Starbucks, and having a paper disposable cup was comforting and
> rewarding for her, so she preferred it!
>
> Get my drift?
>
> -Aaron.
>
>

I wrote:
> It's not merely a *listening* experiment. It's a *playing* experiment > of playing the pieces with one's head right there next to the soundboard > (or the ranks of pipes), at various tempos, plus noticing how the > tensions in the tuning affect one's timing and phrasing as a performer.
> > As for "giving" anything, what more can I do than I've already done, > which is to record these Duetti on organ:
> http://www-personal.umich.edu/~bpl/larips/cd1002.html
> and make the samples of that album freely available too; click through > "disc 2" there.

Plus the Duetti scores, in case anybody missed them, available here for the past 18 months at least:
http://www-personal.umich.edu/~bpl/larips/outline.html

Those are scans from the original print, in facsimile: plus annotations and several pages of commentary by me, showing why I believe Bach wrote (and published) these four little inventions to serve as temperament-test pieces.

Brad Lehman
http://www.larips.com

--- In tuning@yahoogroups.com, "Keenan Pepper" <keenanpepper@...> wrote:
>
> On 1/17/07, Brad Lehman <bpl@...> wrote:
> > And MIDI fake harpsichord or organ tone just doesn't sound convincing
> > enough...nor do fake organ tones do what real organ pipes do, which is
> > to "draw" slightly better into tune on sustained pitches.
>
> Quibble: A MIDI file doesn't contain any timbre information; that all
> depends on the synth used to render it, which can vary over the whole
> range of quality.
>
> The organ pipe thing is an excellent point though, the pipes are
> coupled together so they perturb each other and can even lock in
> phase, if the interval is close enough to just to begin with. I wonder
> if anyone's ever tried to mimic it with software?
>
> Keenan
>

Interesting point, Keenan. I've actually done a little 'research' into
the current state of the art of acoustic simulation, which is what
you'd need in order to do this. I haven't yet got my hands on any of
the tools and engines that I think are currently being used in
immersive multiplayer video games, but I believe that a pipe organ
could be simulated in one of them by defining the pipes and using
appropriate simple sound sources to simulate the reeds or lips of the
pipes. http://citeseer.ist.psu.edu/funkhouser99realtime.html is a
research paper from a few years back - I've been meaning to contact
the researchers, but haven't got around to it.

In case you're wondering, my interest in real-time acoustic modeling
is as a singer - to provide me with a virtual practice room with good
acoustics. Quite apart from boosting some harmonics and generally
mellowing the sound, a good reverb chamber makes precise integer-ratio
tuning stand out.

Patrick

--- In tuning@yahoogroups.com, "Aaron Krister Johnson" <aaron@...>
wrote:

> I'm get suspect of all the large claims people make about various
WTs
> when the first thing that happens when I propose an actual
_hardcore_
> listening test is that people start back-pedalling and making
excuses
> as to why they wouldn't be able to hear such-and-such an effect.
It's
> like you can *feel* people getting nervous about making a mistake
> publicly!

I'm uploading a big listening test now, for BWV 542. Two of the
tunings stood out clearly--my transcription to 31-et (the most
musical in my opinion) and grail, which is a screwball circulating
tuning with two exact 14/11s and one exact 9/7 in place of major
thirds. I thought it was exciting stuff, but of course neither of
these are authentic.

Of the more authentic tunings, you can pick 12-et out by trying to
choose the one that sounds the most boring. Unequal circulating
temperaments sound generally pretty similar, but Werck3 seems to me
to be something you can distinguish from the rest I tried--Brad's
squiggle temperament, Valotti-Young, and Neidhardt1.

My conclusion from all of this is that Bach should be played on an
organ with 31 equal temperament. It sounds good. If you want to use a
historical circulating temperament instead, lots of choices will work
but Werck3 is fine and has a bit more oomph than most, I think.

> As for upper partials, like I said, we don't have to go realistic--a
> sawtooth timbre on a synth is great for upper partials, and they're
> perfectly harmonic like on a harpsichord.

I used a very large, high-quality organ font.