How tunings sound: Mechanically real verses Electronically synthesised

I find that with Scala and a synth I very often have trouble distinguishing (by ear) between tunings that sound very obviously different to me when played on my lyre guitar (which has exchangeable fingerboards with frets in different places).

I have theorised that this is because when I play chords on the guitar, the vibrations in the strings physically interact with each other (the soundboard is light so the vibrations can flow in and out of strings and between stings easily), whereas the synth behaves asthough chords are a group of notes played at the same time, but each note on a different instrument so that they cannot interact in the same way.

But, on the other hand, it may be just that I am more used to how a guitar sounds and have been tuning guitars by ear for over 20 years.

Someone on this forum mentionned that they didn't think differences of less than 13 cents mattered to how notes sound, but Pythagorean major thirds (at about 408 cents) sound obviously different to me on guitar than 12 edo major thirds (400 cents). (As a note there is another major third in the 12 note pythagorean scale, which is about 384 cents, a little flatter than 5/4).

But Marcel (if I have the name right) wrote another time that when he tested pythagorean versus 12 edo on listeners, they prefered the 12 edo major chords.

What do other people think? Are differences between scales more obvious when the notes physically interact with one another before being amplified? Is it the timbre of the notes that makes the difference between scales more obvious? Or do I just hear the differences as more obvious because I'm used to guitar?

If it helps, the guitar I use is more or less semiacoustic. You can see and hear it at:
http://www.youtube.com/user/MarkAllanBarnes#p/a/u/2/AF1ya6zwx_s
http://www.youtube.com/user/MarkAllanBarnes#p/u/7/sLXvP0c-zPQ
(Quarter Comma Meantone)
and
http://www.youtube.com/user/MarkAllanBarnes#p/u/8/x5SxQSUYnw0
(Pythagorean Intonation)
Unfortunately the sound quality is not very good.

On 10 May 2010 21:21, markallanbarnes <mark.barnes3@...> wrote:

> I find that with Scala and a synth I very often have trouble distinguishing
> (by ear) between tunings that sound very obviously different to me when
> played on my lyre guitar (which has exchangeable fingerboards with frets in
> different places).
>
> I have theorised that this is because when I play chords on the guitar, the
> vibrations in the strings physically interact with each other (the
> soundboard is light so the vibrations can flow in and out of strings and
> between stings easily), whereas the synth behaves asthough chords are a
> group of notes played at the same time, but each note on a different
> instrument so that they cannot interact in the same way.
>
> But, on the other hand, it may be just that I am more used to how a guitar
> sounds and have been tuning guitars by ear for over 20 years.
>
> Someone on this forum mentionned that they didn't think differences of less
> than 13 cents mattered to how notes sound, but Pythagorean major thirds (at
> about 408 cents) sound obviously different to me on guitar than 12 edo major
> thirds (400 cents). (As a note there is another major third in the 12 note
> pythagorean scale, which is about 384 cents, a little flatter than 5/4).
>

Differences of 13 cents are huge in my opinion.
But it depends on the sound how well this is heard.
Perhaps indeed a soundboard on a real instrument makes pitch differences
more audible, sound very plausible to me.

As far as synth sounds not beeing as clear.
I think this depends on the sound a lot. For me some soundfonts are much
more clear to me than a straight synth saw wave.
And I have my prefered speakers for listening closely to pitch (and allways
prefer small speakers for this with little bass, clear mids and enough
highs)
Probably amps matter too.
When I listen to microtonal things on my main speaker rig, which is actually
very good quality full range old Tannoy speakers, good solid state amp and
high quality dac, but still I don't hear microtonal pitces nearly as well as
with my laptop built in speakers or my little old apple external computer
speakers.
So It's a matter of finding the right computer sound, and right speakers for
maximum pitch/tuning audibility.

>
> But Marcel (if I have the name right) wrote another time that when he
> tested pythagorean versus 12 edo on listeners, they prefered the 12 edo
> major chords.
>

Yes that was me.
But this is almost common knowledge I thought. I don't know anybody on this
list either who has ever spoken out to prefer a Pythagorean major traid to a
12tet major triad.
The nicest major triad is 1/1 5/4 3/2, and Pythagorean is further away from
this than 12tet.

>
> What do other people think? Are differences between scales more obvious
> when the notes physically interact with one another before being amplified?
> Is it the timbre of the notes that makes the difference between scales more
> obvious? Or do I just hear the differences as more obvious because I'm used
> to guitar?
>
> If it helps, the guitar I use is more or less semiacoustic. You can see and
> hear it at:
> http://www.youtube.com/user/MarkAllanBarnes#p/a/u/2/AF1ya6zwx_s
> http://www.youtube.com/user/MarkAllanBarnes#p/u/7/sLXvP0c-zPQ
> (Quarter Comma Meantone)
> and
> http://www.youtube.com/user/MarkAllanBarnes#p/u/8/x5SxQSUYnw0
> (Pythagorean Intonation)
> Unfortunately the sound quality is not very good.
>
I didn't find it very clear in pitch/tuning actually, but I'm sure that in
real life pitch/tuning is much more audible.

Marcel

[Non-text portions of this message have been removed]

> Marcel: "Differences of 13 cents are huge in my opinion."

Mark: To me aswell, but it depends on the context. Someone had posted that they didn't see the point in differences less than 13 cents, giving the example of tuning a fifth to 74/49 instead of 3/2. That is more than 11 cents. I can hear differences fairly easily when playing a mechanical guitar, but (and I am embarrassed about this) with Scala and a synth, I often cannot tell by ear whether a scale is 12 edo or quarter comma meantone, even though the difference is glaringly obvious to me when I play the scales on a mechanical guitar (usually a semi acoustic guitar without an amplifier)

>Marcel: But it depends on the sound how well this is heard.
> Perhaps indeed a soundboard on a real instrument makes pitch differences
> more audible, sound very plausible to me.
>
> As far as synth sounds not beeing as clear.
> I think this depends on the sound a lot. For me some soundfonts are much
> more clear to me than a straight synth saw wave.
> And I have my prefered speakers for listening closely to pitch (and allways
> prefer small speakers for this with little bass, clear mids and enough
> highs)
> Probably amps matter too.
> When I listen to microtonal things on my main speaker rig, which is actually
> very good quality full range old Tannoy speakers, good solid state amp and
> high quality dac, but still I don't hear microtonal pitces nearly as well as
> with my laptop built in speakers or my little old apple external computer
> speakers.
> So It's a matter of finding the right computer sound, and right speakers for
> maximum pitch/tuning audibility.
>
Mark: Thank you for that, I think that will help a lot. In my case, I think it may be the speakers and amp then. I've been having trouble recently when using my admittedly dodgy and distorting headphones (because I can't be bothered to keep getting down on my hands and knees to swap the jack plugs over between the monitor amp and the headphones. I need to use headphones to record audio and I've found I can hear the metronome easier with headphones than without them when recording electric instruments)
> >
> > Mark Previously: But Marcel (if I have the name right) wrote another time that when he
> > tested pythagorean versus 12 edo on listeners, they prefered the 12 edo
> > major chords.
> >
>
>Marcel: Yes that was me.
> But this is almost common knowledge I thought. I don't know anybody on this
> list either who has ever spoken out to prefer a Pythagorean major traid to a
> 12tet major triad.
> The nicest major triad is 1/1 5/4 3/2, and Pythagorean is further away from
> this than 12tet.

Mark: I think it is worth bearing in mind that the 12 note pythagorean scale gives two different sizes for all the intervals except the octave (as all linear temperaments usually do). There are 8 major thirds that are 81/64, and 4 major thirds that are 8192/6561. 8192/6561 is less than 2 cents flatter than 5/4. This means that there are 3 major triads in 12 note pythagorean intonation that are closer to 1/1 5/4 3/2 than 12 edo is (the other with 5/4 has a wolf fifth).

Whereas I usually prefer 1/1 5/4 3/2 to 1/1 81/64 3/2, I find I enjoy the sound contrast between the "rougher" 81/64 major triad and the almost exact 5/4 triad given in 12 note pythagorean. For example, if the wolf 5th is B flat to F, then the chord sequence A G A G A G A G A G F G# (all major chords) combines the jump out of key when moving from F to G# with a move to a more consonant form of major triad. I personally find this pleasing. A similar effect is obtained in quarter comma and sixth comma, but with those there are 8 more consonant major 3rds and 4 less consonant major thirds.

Also, if you choose the key to match the position of the wolf fifth (or the other way round), you can use 12 note pythagorean to play in very close approximation to 5 limit just versions of the natural major scale, natural minor scale and others, but still have the facility to then change to other chords not in these keys without them sounding too dissonant (you have the possibilty of more tension in the chords outside the key, which is then resolved when you return to the key)

Mark Previously: > > If it helps, the guitar I use is more or less semiacoustic. You can see and
> > hear it at:
> > http://www.youtube.com/user/MarkAllanBarnes#p/a/u/2/AF1ya6zwx_s
> > http://www.youtube.com/user/MarkAllanBarnes#p/u/7/sLXvP0c-zPQ
> > (Quarter Comma Meantone)
> > and
> > http://www.youtube.com/user/MarkAllanBarnes#p/u/8/x5SxQSUYnw0
> > (Pythagorean Intonation)
> > Unfortunately the sound quality is not very good.
> >

> Marcel: I didn't find it very clear in pitch/tuning actually, but I'm sure that in
> real life pitch/tuning is much more audible.
>

Mark: The first 16 tracks on the following page are all in pythagorean intonation or quarter comma, but with better sound quality than the videos:
http://www.reverbnation.com/markbarnes
I still find it easier to tell when playing than listening to recordings because I can deliberately test wolf 5ths, major and minor semitones and different versions of the major 3rd to check which is which.

> I think it is worth bearing in mind that the 12 note pythagorean scale
> gives two different sizes for all the intervals except the octave (as all
> linear temperaments usually do). There are 8 major thirds that are 81/64,
> and 4 major thirds that are 8192/6561. 8192/6561 is less than 2 cents
> flatter than 5/4. This means that there are 3 major triads in 12 note
> pythagorean intonation that are closer to 1/1 5/4 3/2 than 12 edo is (the
> other with 5/4 has a wolf fifth).
>

You mean 2 that are closer than 12tet if you include the wolf one.
But I don't think the wolf one is "closer".
And both defeat the whole purpose of Pythagorean.
If the fifth isn't 3/2 and the major third isn't 81/64 then why
Pythagorean??
The whole reason for the 81/64 is because you can then make chord
progressions where 4 pure fifths make a major third.
To use the other "major third" makes no sense at all in Pythagorean, one can
better tune it 5/4 then because the 8192/6561 is of no use in any other way
than a "major third", it's no longer linked by pure fifths in any reasonable
way (which is the whole basis of Pythagorean)

Marcel

[Non-text portions of this message have been removed]

I think you have made mistakes in your reply to my earlier post.
First, an intonation is a mathematical construction that has no innate purpose or meaning. A person may design an intonation with a purpose in mind, but the purpose is in the person, not the intonation.

Secondly, the 81/64 major third is not only found in pythagorean intonation. It is also found in systems such as Indian classical music which do not use chord progressions and inwhich the ability for a series of fifths to add up to a major third and two octaves is irrelevant.

Thirdly, you have the maths wrong for 12 note pythagorean intonation.
I have my own reasons for choosing to use this scale. Some are historical, for example I like to play renaissance music and also pythagorean intonation is supposedly the tuning for which we have the oldest evidence. Some are practical in that I play guitar and use a lot of barre chords which means that pythagorean intonation is easier for me than 12 note quarter comma or 12 note just intonation with a 5 limit. (7 edo is easier for me than pythagorean, but there are other reasons for not always choosing 7 edo.)

I may be using the scale in a way that the ancient Summerians would have disapproved of. That doesn't worry me much. If they have a problem with it they will have to find a medium they can badger into emailing me.
Since I have studied the scale and play it often, here is the genuine mathematics of 12 note pythagorean intonation:

Here I have chosen the wolf fifth to be B flat to F
All the "perfect" fifths are 3/2 two except B flat to F which is within 2 cents of 40/27
All the major 2nds are 9/8 except B flat to C and E flat to F, which are within 2 cents of 10/9
All the minor 3rds are 32/27 except C to E flat, F to A flat and G to B flat, which are within 2 cents of 6/5
All the major thirds are 81/64 except B flat to D, E flat to G, A flat to C and D flat to F, which are within 2 cents of 5/4
All the augmented fourths are either within 2 cents of 45/32 or within 2 cents of 64/45.
5 of the semitones are within 2 cents of 16/15. The other 7 semitones are within 2 cents of 135/128

(It may be clearer to refer to "2 step intervals" rather than "major 2nds" and so on. I apologise if the way I have written the above is annoying)

The triads A flat Major, D flat Major and E flat Major all have pure 3/2 fifths and thirds that are within 2 cents of 5/4.
The triads C minor, F minor and G minor all have pure 3/2 fifths and thirds that are within 2 cents of 6/5

Choosing A flat as the tonic "Sa" makes the scale A flat, B flat, C, D flat, E flat, F, G an almost perfect rendition of the Indian "Ma Grammar", which in perfect form is 1, 9/8, 5/4, 4/3, 3/2, 5/3, 15/8

Choosing D flat as the tonic "Sa" makes the scale D flat, E flat, F, G flat, A flat, B flat, C an almost perfect rendition of the Indian "Sa Grammar", which in perfect form is 1, 9/8, 5/4, 4/3, 3/2, 27/16, 15/8.

If you extend the scale further by adding more pure fifths, you can achieve a 22 note scale that gives all the standard Indian Srutis to an accuracy of 2 cents (though the pitches of the srutis can infact vary, I mean that you get the pitches as they are usually listed to an accuracy of 2 cents)

If you want exact 5/4 fifths combined with the flexibility and ease of playing on guitar of 12 note pythagorean intonation, you can flatten the non wolf fifths to (2^(7/8))/(5^(1/8)) which is about 701.711 cents. This means flattening them by less than a quarter of a cent and I think it would be called a schismatic temperament.

Mark