blurry unison versus bull's eye unison

Dear tuning list members,

Here is an interesting phenomenon: Two mp3 files created in Logic Pro
both featuring an orchestral unison of the pitch of C, the first a
"blurry unison" 30 cents wide, the second a "bull's eye unison" with
zero pitch bends, appear to be somewhat indiscernable. That is to say,
a listener can tolerate the blurry unison as much as the bull's eye
unison, even though the blurry unison is spread within a 30 cent
region around the C pitch.

Link to the blurry unison mp3 file.

Link to the bull's eye unison mp3 file.

Note that these audio files were created using the Logic Studio
orchestral soundset.

Please share your thoughts. Specifically, do you find that the blurry
unison is objectionable?

Cordially,
Oz.

✩ ✩ ✩
www.ozanyarman.com

Hi Ozan,

Please share your thoughts. Specifically, do you find that the blurry unison
> is objectionable?

I find the blurry unison a little bit objectionable.
Though I think the sound you chose "blurs" the blurry unison.
With other instruments or perhaps thesame instrument played / recorded
differently could very well give a much more objectionable blurry unison.
I know that if you repeat this test with straight saw waves for instance,
the difference will be much more audible.

Marcel

I prefer the sound of the "blurry" unison, actually, as the slight
chorusing involved makes it sound much more like a real orchestra to
my ears. On the other hand, even the "correct unison" didn't sound
like it was all 100% in perfect unison, and there's no guarantee that
the samples themselves are tuned with 100% accuracy to C.

-Mike

On Sun, Aug 30, 2009 at 4:41 PM, Ozan Yarman<ozanyarman@...> wrote:
>
>
> Dear tuning list members,
> Here is an interesting phenomenon: Two mp3 files created in Logic Pro both
> featuring an orchestral unison of the pitch of C, the first a "blurry
> unison" 30 cents wide, the second a "bull's eye unison" with zero pitch
> bends, appear to be somewhat indiscernable. That is to say, a listener can
> tolerate the blurry unison as much as the bull's eye unison, even though the
> blurry unison is spread within a 30 cent region around the C pitch.
> Link to the blurry unison mp3 file.
> Link to the bull's eye unison mp3 file.
> Note that these audio files were created using the Logic Studio orchestral
> soundset.
> Please share your thoughts. Specifically, do you find that the blurry unison
> is objectionable?
> Cordially,
> Oz.
> ✩ ✩ ✩
> www.ozanyarman.com
>

I think this whole issue clouds the problem with accurate intonation when dealing with larger ensembles, and to some extent does question some tuning schemes' use in these contexts.

But this is opinion. It would be interesting to see if an orchestra could perform effectively in 1/8 tones (25-cent) divisions, as this step is of the same size as the 'blur' in the example mp3. From the example, I think that it would be 'doubtful', unless the 'blur' can be significantly reduced.

--- In tuning@yahoogroups.com, Marcel de Velde <m.develde@...> wrote:
>
> Hi Ozan,
>
> Please share your thoughts. Specifically, do you find that the blurry unison
> > is objectionable?
>
>
> I find the blurry unison a little bit objectionable.
> Though I think the sound you chose "blurs" the blurry unison.
> With other instruments or perhaps thesame instrument played / recorded
> differently could very well give a much more objectionable blurry unison.
> I know that if you repeat this test with straight saw waves for instance,
> the difference will be much more audible.
>
> Marcel
>

Everyone, please feel comfortable calling me Oz..

I have gathered responses from Marcel de Velde, Mike Battaglia and
microtonalist on this subject and shall respond in between the lines.

Oz.

✩ ✩ ✩
www.ozanyarman.com

On Aug 31, 2009, at 5:53 PM, Marcel de Velde wrote:

>
>
> Hi Ozan,
>
> Please share your thoughts. Specifically, do you find that the
> blurry unison is objectionable?
>
> I find the blurry unison a little bit objectionable.
> Though I think the sound you chose "blurs" the blurry unison.

That was the point exactly. Orchestral timbres appear to smudge the
pitch so that tones more than a Pythagorean comma apart are still
regarded as the same tone. That is to say, our ability to
differentiate between two pitches seperated by a small interval is
obscured when a complex timbre-structure is involved.

This puts in question the validity of applying voluminous tunings to
orchestras and similar ensembles.

Case in point, the Mezzo channel broadcasts Baroque performances which
involve theorbos tuned in a Pythagorean fashion and harpsichord/organ
in meantone temperament and the clash of tunings do not greatly bother
listeners. In fact the effect is quite the opposite: one of joy.

> With other instruments or perhaps thesame instrument played /
> recorded differently could very well give a much more objectionable
> blurry unison.

Precisely. I think the same way. But is there any established theory
on this subject? I don't know of any myself.

> I know that if you repeat this test with straight saw waves for
> instance, the difference will be much more audible.
>

Please suggest how I can achieve such an experiment with Logic Pro.
Nothing supercedes empricism when it comes to judging the correctness
of music theory assertions.

> Marcel
>
> *
>
> I prefer the sound of the "blurry" unison, actually, as the slight
> chorusing involved makes it sound much more like a real orchestra to
> my ears.

This is an interesting observation. I tend to agree that a "bull's eye
unison" is somewhat boring and unrealistic compared to a slightly
blurry unison. But I suppose, a 30-cent "honky-tonky" smudge is akin
to a student orchestra compared to the unisons of, say, Berlin
Philarmonic. I suspect that even larger (quarter-tonal) smudges occur
with student or beginner orchestras.

> On the other hand, even the "correct unison" didn't sound
> like it was all 100% in perfect unison, and there's no guarantee that
> the samples themselves are tuned with 100% accuracy to C.
>

I have noticed once too often that certain sampled sounds deviate from
the tones of 12-equal (either based on pure or stretched octaves).
Sometimes the deviation is significant and audible in the case of the
lowest and highest registers of pianos, or world/ethnic instruments.
However, I took for granted the tuning of Logic Pro samples and tried
to judge their correctness by ear alone. I think it is fair to say
that the bull's eye unison is in truth a smear just a few cents wide.

I am open to suggestions on how to improve the test.

> -Mike
>
>
> *
>
>
> I think this whole issue clouds the problem with accurate intonation> when dealing with larger ensembles, and to some extent does question
> some tuning schemes' use in these contexts.
>

This is in fact an observation similar to that of Prof. Yalcin Tura,
who claims that commatic differences cannot be discerned with the
orchestras, and the whole issue of maqam polyphony is therefore mute.

> But this is opinion. It would be interesting to see if an orchestra
> could perform effectively in 1/8 tones (25-cent) divisions, as this
> step is of the same size as the 'blur' in the example mp3. From the
> example, I think that it would be 'doubtful', unless the 'blur' can
> be significantly reduced.
>

Smaller by 5 cents in fact. In this case, it is safe to say that one
eigth tone resolution is lost within the unison in question.

> (microtonalist)
>

Cordially,
Oz.

The 'blurry' one sounds less fake, but at the expense of
sounding more like the NY Philharmonic under the direction
of Leonard Bernstein. That is to say, it does certainly sound
'out of tune'.

Cheers,
AAH
=====

--- In tuning@yahoogroups.com, Ozan Yarman <ozanyarman@...> wrote:
>
> Dear tuning list members,
>
> Here is an interesting phenomenon: Two mp3 files created in Logic Pro
> both featuring an orchestral unison of the pitch of C, the first a
> "blurry unison" 30 cents wide, the second a "bull's eye unison" with
> zero pitch bends, appear to be somewhat indiscernable. That is to say,
> a listener can tolerate the blurry unison as much as the bull's eye
> unison, even though the blurry unison is spread within a 30 cent
> region around the C pitch.
>
> Link to the blurry unison mp3 file.
>
> Link to the bull's eye unison mp3 file.
>
> Note that these audio files were created using the Logic Studio
> orchestral soundset.
>
> Please share your thoughts. Specifically, do you find that the blurry
> unison is objectionable?
>
> Cordially,
> Oz.
>
> ✩ ✩ ✩
> www.ozanyarman.com
>

> > I prefer the sound of the "blurry" unison, actually, as the slight
> > chorusing involved makes it sound much more like a real orchestra to
> > my ears.
>
> This is an interesting observation. I tend to agree that a "bull's eye
> unison" is somewhat boring and unrealistic compared to a slightly blurry
> unison. But I suppose, a 30-cent "honky-tonky" smudge is akin to a student
> orchestra compared to the unisons of, say, Berlin Philarmonic. I suspect
> that even larger (quarter-tonal) smudges occur with student or beginner
> orchestras.

Well, it's a 30 cent smudge in theory, but listen to it! I doubt that
if the Berlin Philharmonic played that unison, the listeners in the
audience would suspect that it was 30 cents wide at all.

It would have been interesting to see which one people prefer without
giving away the secret that the 30 cent wide one is 30 cents wide...
To my ears it really just sounds like the instruments are playing with
a slight vibrato or tremolo in the first example, and at different
rates, which makes it sound more "real" to my ears, as I said. Perhaps
this is due to the beating between notes that are close together in
pitch.

> This is in fact an observation similar to that of Prof. Yalcin Tura, who claims that
> commatic differences cannot be discerned with the orchestras, and the whole
> issue of maqam polyphony is therefore mute.

I've always thought that hearing tempered intervals as being
commatically different from one another was a psychological thing more
than an issue of tuning anyway... Provided the tempering in question
isn't too far from the original interval (Under what circumstances
will a 12tet major third fit into the harmony like a 5:4 ratio, and
under what circumstances can you get it to sound like 81:64?)
Likewise, if you write a triadic piece of music in which every 5:4 is
replaced by an 81:64, you'll likely hear those 81:64's as very wide
"versions" of 5:4's. You can think loosely of the orchestra's detuning
of notes as kind of a temperament in which a certain amount of pitch
resolution is lost, but I would expect that commatic conceptual
differences would translate over into the end result nonetheless.

This also begs the question of which 12-tet chords, when played with
an orchestra, perceptually sound like higher-limit chords? Perhaps a
#11 sounds like 11:4, or a b13 sounds like 13:4? Who knows.

-Mike

Thank you for voicing your opinion Aaron Andrew. I also tend to agree
that the blurry unison is more lively and realistic compared to the
"correct" unison. Nevertheless, I think that the 30 cent smudge is too
much out-of-tuneness. I should say, the "bad" unison in the example
sounds more like a novice orchestra. I believe modern orchestras tend
to minimize the smudge to about half of what has been given in the
example.

But the question remains: Is there a theory we can apply as a measure
to what tuning resolution is the upper discernable limit for large
orchestras?

Cordially,
Oz.

✩ ✩ ✩
www.ozanyarman.com

On Sep 1, 2009, at 6:21 PM, hpiinstruments wrote:

> The 'blurry' one sounds less fake, but at the expense of
> sounding more like the NY Philharmonic under the direction
> of Leonard Bernstein. That is to say, it does certainly sound
> 'out of tune'.
>
> Cheers,
> AAH
> =====
>
>
> --- In tuning@yahoogroups.com, Ozan Yarman <ozanyarman@...> wrote:
>>
>> Dear tuning list members,
>>
>> Here is an interesting phenomenon: Two mp3 files created in Logic Pro
>> both featuring an orchestral unison of the pitch of C, the first a
>> "blurry unison" 30 cents wide, the second a "bull's eye unison" with
>> zero pitch bends, appear to be somewhat indiscernable. That is to
>> say,
>> a listener can tolerate the blurry unison as much as the bull's eye
>> unison, even though the blurry unison is spread within a 30 cent
>> region around the C pitch.
>>
>> Link to the blurry unison mp3 file.
>>
>> Link to the bull's eye unison mp3 file.
>>
>> Note that these audio files were created using the Logic Studio
>> orchestral soundset.
>>
>> Please share your thoughts. Specifically, do you find that the blurry
>> unison is objectionable?
>>
>> Cordially,
>> Oz.
>>
>> ✩ ✩ ✩
>> www.ozanyarman.com
>>
>
>
>
>
> ------------------------------------
>
> You can configure your subscription by sending an empty email to one
> of these addresses (from the address at which you receive the list):
> tuning-subscribe@yahoogroups.com - join the tuning group.
> tuning-unsubscribe@yahoogroups.com - leave the group.
> tuning-nomail@yahoogroups.com - turn off mail from the group.
> tuning-digest@yahoogroups.com - set group to send daily digests.
> tuning-normal@yahoogroups.com - set group to send individual emails.
> tuning-help@yahoogroups.com - receive general help information.
> Yahoo! Groups Links
>
>
>

Dear Mike,

I respond in between the lines.

✩ ✩ ✩
www.ozanyarman.com

On Sep 1, 2009, at 7:45 PM, Mike Battaglia wrote:

>>> I prefer the sound of the "blurry" unison, actually, as the slight
>>> chorusing involved makes it sound much more like a real orchestra to
>>> my ears.
>>
>> This is an interesting observation. I tend to agree that a "bull's
>> eye
>> unison" is somewhat boring and unrealistic compared to a slightly
>> blurry
>> unison. But I suppose, a 30-cent "honky-tonky" smudge is akin to a
>> student
>> orchestra compared to the unisons of, say, Berlin Philarmonic. I
>> suspect
>> that even larger (quarter-tonal) smudges occur with student or
>> beginner
>> orchestras.
>
> Well, it's a 30 cent smudge in theory,

No, it was a practical example. The smudge is not merely theoretical,
it is actual.

> but listen to it!

I already have listened to it for more than I can account before
having brought it to the attention of the tuning list. I am the one
who devised the examples.

> I doubt that
> if the Berlin Philharmonic played that unison, the listeners in the
> audience would suspect that it was 30 cents wide at all.
>

A more appropriate example is to measure the out-of-tuneness of the
unisons of modern orchestras such as Berlin and NY Philarmonic. Has
this been done? I suspect not.

> It would have been interesting to see which one people prefer without
> giving away the secret that the 30 cent wide one is 30 cents wide...

I fail to see what that would accomplish. An appropriate scenario
would be where a 20-cent blur was compared to a 40-cent blur. The
listeners would than be requested to identify each blur. But how would
this help in determining the upper-limit tuning resolution that one
can discern with large orchestras?

> To my ears it really just sounds like the instruments are playing with
> a slight vibrato or tremolo in the first example, and at different
> rates, which makes it sound more "real" to my ears, as I said. Perhaps
> this is due to the beating between notes that are close together in
> pitch.
>

That "tremolo" or "vibrato" you hear is the beating of the "badly
tuned" unisons and octaves. Of course, some instruments such as the
horns and oboes have additional vibratos that complicate matters. The
shear result is that of oscillating beat patterns of 1:2:4:8:16:32 I
surmise.

>
>> This is in fact an observation similar to that of Prof. Yalcin
>> Tura, who claims that
>> commatic differences cannot be discerned with the orchestras, and
>> the whole
>> issue of maqam polyphony is therefore mute.
>
> I've always thought that hearing tempered intervals as being
> commatically different from one another was a psychological thing more
> than an issue of tuning anyway... Provided the tempering in question
> isn't too far from the original interval (Under what circumstances
> will a 12tet major third fit into the harmony like a 5:4 ratio, and
> under what circumstances can you get it to sound like 81:64?)

I always wondered the same thing. JI theory would dictate that we are
always desiring to hear just or near-just intervals. But I think the
validity of this theory is in question. Nevertheless, this is another
topic of discussion.

> Likewise, if you write a triadic piece of music in which every 5:4 is
> replaced by an 81:64, you'll likely hear those 81:64's as very wide
> "versions" of 5:4's.

Perhaps. Or else, there are different major third qualities that one
desires in triadic music, depending on the harmonic progression. This
is a phenomenon I have experienced often with irregular cyclic
temperaments I set on my grandpiano. The mood or the force of the
harmony dictates the use of either Pyth/super-Pyth triads or near JI/
JI triads. Again this is the subject of another discussion.

> You can think loosely of the orchestra's detuning
> of notes as kind of a temperament in which a certain amount of pitch
> resolution is lost, but I would expect that commatic conceptual
> differences would translate over into the end result nonetheless.
>

I don't get this part. Please elaborate.

> This also begs the question of which 12-tet chords, when played with
> an orchestra, perceptually sound like higher-limit chords? Perhaps a
> #11 sounds like 11:4, or a b13 sounds like 13:4? Who knows.
>

Again, the subject of another discussion.

> -Mike
>
>

Cordially,
Oz.

Ozan,

Both files did sound like octaves, but I like the "bull's eye unison" much, much better. A 30 cent error in octaves or fifths is too rough for my tastes, but I could tolerate maybe 15-20 cents.

A lot of it is personal taste; I do prefer dry sounds, with intervals no more than 8 cents or so off of just. Heavier detuning is something I use purely for effect; same goes for chorus effect.

~D.

----- Original Message ----- From: Ozan Yarman
To: Tuning List
Sent: Sunday, 30 August, 2009 15:41
Subject: [tuning] blurry unison versus bull's eye unison

Dear tuning list members,

Here is an interesting phenomenon: Two mp3 files created in Logic Pro both featuring an orchestral unison of the pitch of C, the first a "blurry unison" 30 cents wide, the second a "bull's eye unison" with zero pitch bends, appear to be somewhat indiscernable. That is to say, a listener can tolerate the blurry unison as much as the bull's eye unison, even though the blurry unison is spread within a 30 cent region around the C pitch.

Link to the blurry unison mp3 file.

Link to the bull's eye unison mp3 file.

Note that these audio files were created using the Logic Studio orchestral soundset.

Please share your thoughts. Specifically, do you find that the blurry unison is objectionable?

Cordially,
Oz.

✩ ✩ ✩
www.ozanyarman.com

>> Well, it's a 30 cent smudge in theory,
>
> No, it was a practical example. The smudge is not merely theoretical,
> it is actual.
>
>> but listen to it!
>
> I already have listened to it for more than I can account before
> having brought it to the attention of the tuning list. I am the one
> who devised the examples.

Is there some kind of unwritten rule that every discussion on the
tuning list has to be conducted as condescendingly as possible, as if
the other party were an idiot? Please don't respond to small fractions
of my sentences out of context like that.

>> I doubt that
>> if the Berlin Philharmonic played that unison, the listeners in the
>> audience would suspect that it was 30 cents wide at all.
>>
>
> A more appropriate example is to measure the out-of-tuneness of the
> unisons of modern orchestras such as Berlin and NY Philarmonic. Has
> this been done? I suspect not.

If you'd like to supply recordings of unisons from any major modern
orchestra, I'd be happy to play around with them in MATLAB and see how
spread out they are. Keep in mind that if different instruments in the
unison start playing with vibrato or tremolo, that the results will
quickly become more and more difficult to tell from mistunings of the
same note.

>> It would have been interesting to see which one people prefer without
>> giving away the secret that the 30 cent wide one is 30 cents wide...
>
> I fail to see what that would accomplish. An appropriate scenario
> would be where a 20-cent blur was compared to a 40-cent blur. The
> listeners would than be requested to identify each blur. But how would
> this help in determining the upper-limit tuning resolution that one
> can discern with large orchestras?

It would help to offset any bias introduced that one might have with
after learning in advance that the first example was out of tune and
30 cents wide. That type of bias would certainly affect the results of
any test that hopes to measure ones self-diagnosed perceptual
"tolerance" of a mistuning.

>> You can think loosely of the orchestra's detuning
>> of notes as kind of a temperament in which a certain amount of pitch
>> resolution is lost, but I would expect that commatic conceptual
>> differences would translate over into the end result nonetheless.
>
> I don't get this part. Please elaborate.

If it turns out that the ability to discern a 40-cent wide unison from
a perfect unison isn't that good, then you could loosely think of that
40-cent interval as being "tempered out" of the tuning used.
Nonetheless, you'd still be able to imply different comma-adjacent
intervals in the same way you would in 12-tet, or any other tempered
tuning, so I hardly think you can just disregard their existence
entirely for orchestral works.

-Mike

--- In tuning@yahoogroups.com, Ozan Yarman <> wrote:
>
> Thank you for voicing your opinion Aaron Andrew. I also
> tend to agree that the blurry unison is more lively and
> realistic compared to the "correct" unison. Nevertheless,
> I think that the 30 cent smudge is too
> much out-of-tuneness. I should say, the "bad" unison in
> the example sounds more like a novice orchestra.

Indeed, perhaps instead of the NY Phil under Bernstein (although
the tuning on those recordings is remarkably bad) probably I
should have said the Podunk Philharmonic directed by
Horhey Snappypants; you see, I did not want to offend
Mr. Snappypants or the residents of Podunk or the members
of its Philharmonic.

> I believe modern orchestras tend
> to minimize the smudge to about half of what has been
> given in the
> example.

For unisons, agreed, but in diverse situations, the spread
may vary more widely. Of course the idea of 'chorus'
('celeste' on pipe organs) is a deliberate mistuning,
as obviously this is what happens unintentionally, that is,
naturally, in a chorus of instruments or voices in real life.

BTW, Martin Vogel claimed in 'On The Relations of Tone'
that orchestras in Germany were performing with tuning
errors on the order of a quartertone, when playing atonal
works. He had no direct evidence for this, however, only
his own insight into tuning, which was very deep.

> But the question remains: Is there a theory we can
> apply as a measure to what tuning resolution is the
> upper discernable limit for large orchestras?

As you know, Oz, as part of my own theory called
H-System, I give 2 answers for this. First, the
_Iintonation _Quality system (IQ System), based
on the perceptual unit of the average JND, states
that an attempt at correct intonation qualifies as
a 'hit' within + / - 12 cents with varying degrees of
accuracy:
<http://www.h-pi.com/theory/huntsystem2.html#2>

See also: an online calculator giving an IQGPA
(Intonation Quality Grade Point Average) for how
well a given tuning will produce a set of target pitches:
<http://www.h-pi.com/theory/huntsystemcalc.html>

Second, the theoretical measure of the average comma
states that a pitch changes categorically at around 27 cents,
so pitches within this boundary should all sound as if they
are different versions of one another.
<http://www.h-pi.com/theory/huntsystem2.html#4>

Each of these measures supports the argument that
intonation is deemed acceptable within the range of an
eighth of a tone (within a Pythagorean comma). Critical
band theory, as well as research in physical measurement
of actual performance pitches by both instruments and
voices also support these ideas.

Cheers,
AAH
=====

... oops, I misspoke my own system :D
IQ hit boundary is + / - 9 cents, as you can see here:
<http://www.h-pi.com/theory/huntsystemcalc.html>

--- In tuning@yahoogroups.com, "hpiinstruments" <aaronhunt@...> wrote:
>
> --- In tuning@yahoogroups.com, Ozan Yarman <> wrote:
> >
> > Thank you for voicing your opinion Aaron Andrew. I also
> > tend to agree that the blurry unison is more lively and
> > realistic compared to the "correct" unison. Nevertheless,
> > I think that the 30 cent smudge is too
> > much out-of-tuneness. I should say, the "bad" unison in
> > the example sounds more like a novice orchestra.
>
>
> Indeed, perhaps instead of the NY Phil under Bernstein (although
> the tuning on those recordings is remarkably bad) probably I
> should have said the Podunk Philharmonic directed by
> Horhey Snappypants; you see, I did not want to offend
> Mr. Snappypants or the residents of Podunk or the members
> of its Philharmonic.
>
>
> > I believe modern orchestras tend
> > to minimize the smudge to about half of what has been
> > given in the
> > example.
>
>
> For unisons, agreed, but in diverse situations, the spread
> may vary more widely. Of course the idea of 'chorus'
> ('celeste' on pipe organs) is a deliberate mistuning,
> as obviously this is what happens unintentionally, that is,
> naturally, in a chorus of instruments or voices in real life.
>
> BTW, Martin Vogel claimed in 'On The Relations of Tone'
> that orchestras in Germany were performing with tuning
> errors on the order of a quartertone, when playing atonal
> works. He had no direct evidence for this, however, only
> his own insight into tuning, which was very deep.
>
>
> > But the question remains: Is there a theory we can
> > apply as a measure to what tuning resolution is the
> > upper discernable limit for large orchestras?
>
>
> As you know, Oz, as part of my own theory called
> H-System, I give 2 answers for this. First, the
> _Iintonation _Quality system (IQ System), based
> on the perceptual unit of the average JND, states
> that an attempt at correct intonation qualifies as
> a 'hit' within + / - 12 cents with varying degrees of
> accuracy:
> <http://www.h-pi.com/theory/huntsystem2.html#2>
>
> See also: an online calculator giving an IQGPA
> (Intonation Quality Grade Point Average) for how
> well a given tuning will produce a set of target pitches:
> <http://www.h-pi.com/theory/huntsystemcalc.html>
>
> Second, the theoretical measure of the average comma
> states that a pitch changes categorically at around 27 cents,
> so pitches within this boundary should all sound as if they
> are different versions of one another.
> <http://www.h-pi.com/theory/huntsystem2.html#4>
>
> Each of these measures supports the argument that
> intonation is deemed acceptable within the range of an
> eighth of a tone (within a Pythagorean comma). Critical
> band theory, as well as research in physical measurement
> of actual performance pitches by both instruments and
> voices also support these ideas.
>
> Cheers,
> AAH
> =====
>

On Sep 2, 2009, at 1:30 AM, Mike Battaglia wrote:

>>> Well, it's a 30 cent smudge in theory,
>>
>> No, it was a practical example. The smudge is not merely theoretical,
>> it is actual.
>>
>>> but listen to it!
>>
>> I already have listened to it for more than I can account before
>> having brought it to the attention of the tuning list. I am the one
>> who devised the examples.
>
> Is there some kind of unwritten rule that every discussion on the
> tuning list has to be conducted as condescendingly as possible, as if
> the other party were an idiot? Please don't respond to small fractions
> of my sentences out of context like that.

??? You should not compound a rough remark with unnecessary rudeness in a discussion. I was not condescending in my response, but was trying to be gentle. So please be kind enough to refrain from emotional upheavals. Let us not escalate tensions any further.

>
>>> I doubt that
>>> if the Berlin Philharmonic played that unison, the listeners in the
>>> audience would suspect that it was 30 cents wide at all.
>>>
>>
>> A more appropriate example is to measure the out-of-tuneness of the
>> unisons of modern orchestras such as Berlin and NY Philarmonic. Has
>> this been done? I suspect not.
>
> If you'd like to supply recordings of unisons from any major modern
> orchestra, I'd be happy to play around with them in MATLAB and see how
> spread out they are. Keep in mind that if different instruments in the
> unison start playing with vibrato or tremolo, that the results will
> quickly become more and more difficult to tell from mistunings of the
> same note.

I may not be the best person to supply you with such samples. Maybe other members of the list may.

>
>>> It would have been interesting to see which one people prefer >>> without
>>> giving away the secret that the 30 cent wide one is 30 cents wide...
>>
>> I fail to see what that would accomplish. An appropriate scenario
>> would be where a 20-cent blur was compared to a 40-cent blur. The
>> listeners would than be requested to identify each blur. But how >> would
>> this help in determining the upper-limit tuning resolution that one
>> can discern with large orchestras?
>
> It would help to offset any bias introduced that one might have with
> after learning in advance that the first example was out of tune and
> 30 cents wide. That type of bias would certainly affect the results of
> any test that hopes to measure ones self-diagnosed perceptual
> "tolerance" of a mistuning.
>

Possibly.

>
>>> You can think loosely of the orchestra's detuning
>>> of notes as kind of a temperament in which a certain amount of pitch
>>> resolution is lost, but I would expect that commatic conceptual
>>> differences would translate over into the end result nonetheless.
>>
>> I don't get this part. Please elaborate.
>
> If it turns out that the ability to discern a 40-cent wide unison from
> a perfect unison isn't that good, then you could loosely think of that
> 40-cent interval as being "tempered out" of the tuning used.

I don't see how a stand-alone 40 cent interval can be tempered out from, say, 12-tET. This analogy is not a very attractive one to me.

> Nonetheless, you'd still be able to imply different comma-adjacent
> intervals in the same way you would in 12-tet, or any other tempered
> tuning, so I hardly think you can just disregard their existence
> entirely for orchestral works.
>

I do not get this part either. Please elaborate further.

> -Mike
>

Cordially,
Oz.

Dear Danny, hello!

While I agree with your views, maybe a kind member of the tuning list
will prepare a 20-cent blurry orchestral unison for testing purposes?
I can send the Logic Pro file of the correct unison to work with.

Cordially,
Oz.

✩ ✩ ✩
www.ozanyarman.com

On Sep 2, 2009, at 1:11 AM, Danny Wier wrote:

> Ozan,
>
> Both files did sound like octaves, but I like the "bull's eye unison"
> much, much better. A 30 cent error in octaves or fifths is too rough
> for
> my tastes, but I could tolerate maybe 15-20 cents.
>
> A lot of it is personal taste; I do prefer dry sounds, with
> intervals no
> more than 8 cents or so off of just. Heavier detuning is something I
> use
> purely for effect; same goes for chorus effect.
>
> ~D.
>

On Sep 2, 2009, at 1:36 AM, hpiinstruments wrote:

> --- In tuning@yahoogroups.com, Ozan Yarman <> wrote:
>>
>> Thank you for voicing your opinion Aaron Andrew. I also
>> tend to agree that the blurry unison is more lively and
>> realistic compared to the "correct" unison. Nevertheless,
>> I think that the 30 cent smudge is too
>> much out-of-tuneness. I should say, the "bad" unison in
>> the example sounds more like a novice orchestra.
>
>
> Indeed, perhaps instead of the NY Phil under Bernstein (although
> the tuning on those recordings is remarkably bad) probably I
> should have said the Podunk Philharmonic directed by
> Horhey Snappypants; you see, I did not want to offend
> Mr. Snappypants or the residents of Podunk or the members
> of its Philharmonic.
>

:)

>
>> I believe modern orchestras tend
>> to minimize the smudge to about half of what has been
>> given in the
>> example.
>
>
> For unisons, agreed, but in diverse situations, the spread
> may vary more widely. Of course the idea of 'chorus'
> ('celeste' on pipe organs) is a deliberate mistuning,
> as obviously this is what happens unintentionally, that is,
> naturally, in a chorus of instruments or voices in real life.
>
> BTW, Martin Vogel claimed in 'On The Relations of Tone'
> that orchestras in Germany were performing with tuning
> errors on the order of a quartertone, when playing atonal
> works. He had no direct evidence for this, however, only
> his own insight into tuning, which was very deep.
>

Interesting.

>
>> But the question remains: Is there a theory we can
>> apply as a measure to what tuning resolution is the
>> upper discernable limit for large orchestras?
>
>
> As you know, Oz, as part of my own theory called
> H-System, I give 2 answers for this. First, the
> _Iintonation _Quality system (IQ System), based
> on the perceptual unit of the average JND, states
> that an attempt at correct intonation qualifies as
> a 'hit' within + / - 12 cents with varying degrees of
> accuracy:
> <http://www.h-pi.com/theory/huntsystem2.html#2>
>

Yes, I have arrived at similar conclusions: That a 9 cent error is a poor hit, while 10-11 cent error is a near-miss. But this doesn't seem to work for complex timbres as in the case of a symphony orchestra. Such a theoretical approach may be constrained to solo instruments and small ensembles.

> See also: an online calculator giving an IQGPA
> (Intonation Quality Grade Point Average) for how
> well a given tuning will produce a set of target pitches:
> <http://www.h-pi.com/theory/huntsystemcalc.html>
>

Splendid. But have you tried 171-tET? Three steps of this temperament yield practically a syntonic comma.

> Second, the theoretical measure of the average comma
> states that a pitch changes categorically at around 27 cents,
> so pitches within this boundary should all sound as if they
> are different versions of one another.
> <http://www.h-pi.com/theory/huntsystem2.html#4>
>

In the case of the symphonic blurry unison 30-cents wide, the threshold seems to increase with complex timbres.

> Each of these measures supports the argument that
> intonation is deemed acceptable within the range of an
> eighth of a tone (within a Pythagorean comma). Critical
> band theory, as well as research in physical measurement
> of actual performance pitches by both instruments and
> voices also support these ideas.
>

This would inadvertently mean that you agree to replacing 48-tET with Just Intonation, which I believe you don't. Then again, in a symphony orchestra, 48-tET might not be executed much differently from 41-tET.

> Cheers,
> AAH
> =====
>

Cordially,
Oz.

--- In tuning@yahoogroups.com, Ozan Yarman <ozanyarman@...> wrote:
> On Sep 2, 2009, at 1:36 AM, hpiinstruments wrote:
> > See also: an online calculator giving an IQGPA
> > (Intonation Quality Grade Point Average) for how
> > well a given tuning will produce a set of target pitches:
> > <http://www.h-pi.com/theory/huntsystemcalc.html>
> >
>
> Splendid. But have you tried 171-tET? Three steps of this temperament
> yield practically a syntonic comma.

Yes, of course; 171ET was Martin Vogel's ET of choice, it
being excellent for 7-Limit JI. But unfortunately, 171 is
3 x 3 x 19 which makes it a meantone derived tuning
having flat fifths from 19 which get corrected by 171,
so it's no good taxonomically, from a circle-of-fifths
point of view.

> > Each of these measures supports the argument that
> > intonation is deemed acceptable within the range of an
> > eighth of a tone (within a Pythagorean comma). Critical
> > band theory, as well as research in physical measurement
> > of actual performance pitches by both instruments and
> > voices also support these ideas.
> >
>
>
> This would inadvertently mean that you agree to replacing 48-tET with
> Just Intonation,

I'm sure you meant to say this vice versa.

> which I believe you don't.

Correct; I don't.

> Then again, in a symphony
> orchestra, 48-tET might not be executed much differently from 41-tET.

Who knows...

I would like to see instruments redesigned for playing 41ET
as suggested by POL's 21st C Orchestral Instruments, with
JND adjustments as in my work.

Cheers,
Aaron
=====

Dear Aaron Andrew,

✩ ✩ ✩
www.ozanyarman.com

On Sep 2, 2009, at 3:37 AM, hpiinstruments wrote:

> --- In tuning@yahoogroups.com, Ozan Yarman <ozanyarman@...> wrote:
>> On Sep 2, 2009, at 1:36 AM, hpiinstruments wrote:
>>> See also: an online calculator giving an IQGPA
>>> (Intonation Quality Grade Point Average) for how
>>> well a given tuning will produce a set of target pitches:
>>> <http://www.h-pi.com/theory/huntsystemcalc.html>
>>>
>>
>> Splendid. But have you tried 171-tET? Three steps of this temperament
>> yield practically a syntonic comma.
>
>
>
> Yes, of course; 171ET was Martin Vogel's ET of choice, it
> being excellent for 7-Limit JI. But unfortunately, 171 is
> 3 x 3 x 19 which makes it a meantone derived tuning
> having flat fifths from 19 which get corrected by 171,
> so it's no good taxonomically, from a circle-of-fifths
> point of view.
>

Possibly. But all the same...

>
>
>>> Each of these measures supports the argument that
>>> intonation is deemed acceptable within the range of an
>>> eighth of a tone (within a Pythagorean comma). Critical
>>> band theory, as well as research in physical measurement
>>> of actual performance pitches by both instruments and
>>> voices also support these ideas.
>>>
>>
>>
>> This would inadvertently mean that you agree to replacing 48-tET with
>> Just Intonation,
>
>
> I'm sure you meant to say this vice versa.
>

Yes, you are right.

>
>> which I believe you don't.
>
>
> Correct; I don't.
>

Good.

>
>> Then again, in a symphony
>> orchestra, 48-tET might not be executed much differently from 41-tET.
>
>
> Who knows...
>
> I would like to see instruments redesigned for playing 41ET
> as suggested by POL's 21st C Orchestral Instruments, with
> JND adjustments as in my work.
>

In my opinion, it does not seem possible to have listeners appreciate
the JND distinctions in such an orchestra, given that even a 30-cent
smudge of a unison is perceived as the same pitch as a bull's eye
unison.

> Cheers,
> Aaron
> =====
>
>
>

Cordially,
Oz.

--- In tuning@yahoogroups.com, Ozan Yarman <ozanyarman@...> wrote:
> > I would like to see instruments redesigned for playing 41ET
> > as suggested by POL's 21st C Orchestral Instruments, with
> > JND adjustments as in my work.
> >
>
>
> In my opinion, it does not seem possible to have listeners appreciate
> the JND distinctions in such an orchestra, given that even a 30-cent
> smudge of a unison is perceived as the same pitch as a bull's eye
> unison.

The idea is that instruments would be built for 41 and players
would fine tune in JNDs.

Yours,
Aaron
=====

I listened to both and thought the wide version less desirable but much more
realistic.

This conversation seems to be driving towards one of many conclusions - 12
TET may have been arrived at in part by inaccuracies in ensemble
performance?

I thought of an experiment, if anyone is interested - With a tracker I have
fine control over the pitch of samples.
I can easily compare from sine wave to squarewave (or other "stock"
waveform) with 1 cent deviation control.
Actually... since the tracker has filtering I can do a comparison with a
sample of strings and sweep from near sinusoidal to full string sound.
I think this would shed some light on the timbre/pitch discrimination
question - if others think it could be useful.

Thanks for the insight Oz,

Chris

On Sun, Aug 30, 2009 at 4:41 PM, Ozan Yarman <ozanyarman@ozanyarman.com>wrote:

>
>
> Dear tuning list members,
>
> Here is an interesting phenomenon: Two mp3 files created in Logic Pro both
> featuring an orchestral unison of the pitch of C, the first a "blurry
> unison" *30 cents wide*, the second a "bull's eye unison" with *zero pitch
> bends*, appear to be somewhat indiscernable. That is to say, a listener
> can tolerate the blurry unison as much as the bull's eye unison, even though
> the blurry unison is spread within a 30 cent region around the C pitch.
>
> Link to the blurry unison <http://www.ozanyarman.com/Media/badunison.mp3> mp3
> file.
>
> Link to the bull's eye unison<http://www.ozanyarman.com/Media/correctunison.mp3> mp3
> file.
>
> Note that these audio files were created using the Logic Studio orchestral
> soundset.
>
> Please share your thoughts. Specifically, do you find that the blurry
> unison is objectionable?
>
> Cordially,
> Oz.
>
> ✩ ✩ ✩
> www.ozanyarman.com
>
>
>

Thank you for your input Chris. The experiment you are suggesting
seems to be a good one. Be sure to randomize the pitch-bends of
strings, woodwinds and brass in a unison.

Logic has excellent synth modelling capabilities but I know little to
nothing about. With some guidance, I might be tempted to try something
out in the future myself.

Cordially,
Oz.

P.S. Your new musics are very charming.

✩ ✩ ✩
www.ozanyarman.com

On Sep 2, 2009, at 4:33 AM, Chris Vaisvil wrote:

>
>
> I listened to both and thought the wide version less desirable but
> much more realistic.
>
> This conversation seems to be driving towards one of many
> conclusions - 12 TET may have been arrived at in part by
> inaccuracies in ensemble performance?
>
> I thought of an experiment, if anyone is interested - With a tracker
> I have fine control over the pitch of samples.
> I can easily compare from sine wave to squarewave (or other "stock"
> waveform) with 1 cent deviation control.
> Actually... since the tracker has filtering I can do a comparison
> with a sample of strings and sweep from near sinusoidal to full
> string sound.
> I think this would shed some light on the timbre/pitch
> discrimination question - if others think it could be useful.
>
> Thanks for the insight Oz,
>
> Chris
>
> On Sun, Aug 30, 2009 at 4:41 PM, Ozan Yarman <ozanyarman@...
> > wrote:
>
>
> Dear tuning list members,
>
> Here is an interesting phenomenon: Two mp3 files created in Logic
> Pro both featuring an orchestral unison of the pitch of C, the first
> a "blurry unison" 30 cents wide, the second a "bull's eye unison"
> with zero pitch bends, appear to be somewhat indiscernable. That is
> to say, a listener can tolerate the blurry unison as much as the> bull's eye unison, even though the blurry unison is spread within a
> 30 cent region around the C pitch.
>
> Link to the blurry unison mp3 file.
>
> Link to the bull's eye unison mp3 file.
>
> Note that these audio files were created using the Logic Studio
> orchestral soundset.
>
> Please share your thoughts. Specifically, do you find that the
> blurry unison is objectionable?
>
> Cordially,
> Oz.
>
> ✩ ✩ ✩
> www.ozanyarman.com
>
>
>
>
>