Can Old Italian sound be seeen in violin body impact spectra?

[Don Noon]

Hi Anders

Bit off topic but related.

Do you have any thoughts or data re violin sound and fittings?

I would be interested to see if there is a measurable difference in tone

when a violin is played without a chinrest or shoulder rest.

Thanks

I can say absolutely positively yes, chin and shoulder rests do make significant measurable differences.

What you point out with your post, is that a basic fallacy might exist within the PREMISS of what is being proposed here, of using this particular "diagnostic method" to identify Old Italian Sound. Well, if the answer is simply "no", then the discussion ends there, right?

I think the answer is yes, there are spectral features that might be used to identify "Old Italian Sound'. I'm not sure everyone will agree what that is, but the distinguishing characteristics are, I think:

-unususally weak response ~1000 to 1500 Hz

-unusually high response 2000 - 4000 Hz (a solid "hill", not peaky and dippy)

-steep dropoff in response above 3000 or 4000 Hz.

These characteristics will give very high ratings in the Dunnwald parameters.

This response shape has been a mystery to me... how any arching, graduation, or wood properties could give a strong, solid response in one zone (obviously not just one or two modes, but a bunch of them), yet depress response just below and above that zone. There may be enough clues around to figure it out now, I hope... I'll let you know.

[Craig Tucker]

Sorry, I was editing the grammer in my original post, and hit the wrong button.

[Don Noon]

We can measure my height, but it doesn't really say anything about "me".

True. But if we didn't measure your height, we would know even less.

I don't think anyone claims (or, at least, they shouldn't) that impact tests tell us everything. But at least it's something quantitative that we might use to make some progress in understanding.

[Don Noon]

Sorry, I was editing the grammer in my original post, and hit the wrong button.

Even subconsiously, you seem to be a multi-poster

[Craig Tucker]

Even subconsiously, you seem to be a multi-poster

Shhh!

I don't want THAT getting out!

[stradofear]

I don't think anyone claims (or, at least, they shouldn't) that impact tests tell us everything. But at least it's something quantitative that we might use to make some progress in understanding.

Sure, but the danger is always thinking we know more than we do, because numbers don't lie.

[Don Noon]

"Thinking we know more than we do" certainly isn't reserved only for "science types". That's been a common human trait forever, and one needs to be careful not to fall into that trap.

[JimMurphy]

... there are spectral features that might be used to identify "Old Italian Sound'. ... This response shape has been a mystery to me... how any arching, graduation, or wood properties could give a strong, solid response in one zone (obviously not just one or two modes, but a bunch of them), yet depress response just below and above that zone.

Impact spectra can reveal quite a bit actually, about Cremonese acoustic design concepts, i.e., how they were trying to blend frequencies mathematically [using geometry].

Maybe the Old Italian ~1000-1500Hz 'dip' is merely a function of missed design opportunity - as revealed by lack of correlating

spectra frequencies [to instrument tuning]. Something 'off' in their compression & rarefication scheme [Modern acoustic terms].

There may be more info garnered by studying different instrument types from the same Maker. If you've ever closely analyzed spectra from both an original [no regrads, etc.] Old Master Violin & Viola for example, you may be tempted to "question" whether that one Maker was consistent in his acoustic design approach [iF he had one], or whether he was actually more of an Artiste.

Jim

[Salve Håkedal]

..

I think the answer is yes, there are spectral features that might be used to identify "Old Italian Sound'. I'm not sure everyone will agree what that is, but the distinguishing characteristics are, I think:

-unususally weak response ~1000 to 1500 Hz

-unusually high response 2000 - 4000 Hz (a solid "hill", not peaky and dippy)

-steep dropoff in response above 3000 or 4000 Hz.

..

Hi all!

I really want to be with the science guys, but when I'm actually carving I fall back to my old intuitive approach! F#*!

Don. Looking at the spectra from the Strad3D dvd your statements above don't seem to me to be true. I can't see a big dip around 1000Hz for the Titian and the Plowden(?!)

[stradofear]

I don't remember where the Dunnewald parameters were derived from, but if they were from anything Italian, anything from 1600 to 1800, that could bring in a lot of contradictory data, and still not show the cream rising. There are, after all, thousands and thousands of violins of mediocre performance fitting that criteria, but only 150-175 del Gesus, which people would choose in preference to any of those thousands of others.

Likewise with Strads--a larger number, but still encompassing the desired sound. . . . and a very different sound than a del Gesu. Even if you mixed the sounds of the best Strads and del Gesus in a blender, would the resulting tone be considered as desirable as either one? I'm not so sure of that.

This is the problem I have with averaging exercises where care isn't taken to separate things into legitimate types and measure them as separate things. It's similar to how Sacconi seems to have blended all Strad traits together and arrived at an idealized Strad model which seems in some respects not to very closely resemble what any real Strad actually is.

[Johnmasters]

Impact spectra can reveal quite a bit actually, about Cremonese acoustic design concepts, i.e., how they were trying to blend frequencies mathematically [using geometry].

Maybe the Old Italian ~1000-1500Hz 'dip' is merely a function of missed design opportunity - as revealed by lack of correlating

spectra frequencies [to instrument tuning]. Something 'off' in their compression & rarefication scheme [Modern acoustic terms].

There may be more info garnered by studying different instrument types from the same Maker. If you've ever closely analyzed spectra from both an original [no regrads, etc.] Old Master Violin & Viola for example, you may be tempted to "question" whether that one Maker was consistent in his acoustic design approach [iF he had one], or whether he was actually more of an Artiste.

Jim

I found this Phase Shifter and think I will try to make it. I have a pile of op-amps.

If there is anything beyond power spectra, it would be nice to know. One could hear if this changed the sound of a sawtooth from a signal generator and also run a played violin signal through it.

This should be simple to do, and maybe would turn up something interesting.

Perhaps the signal "Vin" coule be changed to "Vln",

[Don Noon]

Hi all!

I really want to be with the science guys, but when I'm actually carving I fall back to my old intuitive approach! F#*!

Don. Looking at the spectra from the Strad3D dvd your statements above don't seem to me to be true. I can't see a big dip around 1000Hz for the Titian and the Plowden(?!)

I agree. According to what I think defines "the Old Italian Sound", these don't fit. There is not concrete definition of "Old Italian Sound", and it is not a certainty that all violins that are old and Italian will have the OIS... whatever it is. I recall Fan Tao at Oberlin this past year commenting that when he played the Titian, he thought it sounded like a modern instrument. He also didn't think much of the Plowden, but I don't recall any specific description of the sound.

We need and accurate OIS meter, so we can know positively if a fiddle has 98% OIS sound, or only 84%.

Until then, we will just have to muddle along in the fog, as usual.

[JimMurphy]

If there is anything beyond power spectra, it would be nice to know.

I'm both acknowledging the greatness of benchmark Old Italian Sound AND questioning whether the low 'dip' is a required precursor

[as some are interpreting the data].

Perhaps there's room for 'sweet-mids' too [without the dip] between buttery-lows & laser-like highs.

Jim

[stradofear]

I recall Fan Tao at Oberlin this past year commenting that when he played the Titian, he thought it sounded like a modern instrument.

That's why I don't like the very broad "Old Italian" category. One of the things I've noticed about Strads is that relative to almost all other instruments in the "Old Italian" category, they often have an extremely open tone quality that's associated with, for instance, makers like Carl Becker (a very extreme example, and why I suspect the Beckers have gone for this attribute) more than makers like Grancino, Testore, Gagliano, Cappa, and [endless list of "Old Italians"]. You hardly ever meet a player who complains that a violin is "too open sounding."

[JimMurphy]

I don't remember where the Dunnewald parameters were derived from, but if they were from anything Italian, anything from 1600 to 1800, that could bring in a lot of contradictory data, and still not show the cream rising.

There's also a chance Dünnwald's choice of normalizing-to-25dB has skewed certain harmonic-based frequency relationships.

But it was his mathematical choice to make nonetheless.

Jim

[Johnmasters]

That's why I don't like the very broad "Old Italian" category. One of the things I've noticed about Strads is that relative to almost all other instruments in the "Old Italian" category, they often have an extremely open tone quality that's associated with, for instance, makers like Carl Becker (a very extreme example, and why I suspect the Beckers have gone for this attribute) more than makers like Grancino, Testore, Gagliano, Cappa, and [endless list of "Old Italians"]. You hardly ever meet a player who complains that a violin is "too open sounding."

I agree 100% and it is not necessarily just being "loud." I keep mentioning my planar speakers... They are very open and clear even though they are inefficient. They sound "open" at low volumes too. That is one reason I am interested in the phase shifting of components and the possible role of this effect.

The phase shift in question is not from the radiation geometry but from the way the entire system is driven by the strings. I would like to see the phase-shift filter in use. I can make one, but perhaps Jim or Anders will take an interest too.

Being "open" seems to me to be rather independent of timbre. Perhaps fairly independent of response curves also.

The thing to know is what is the physical and mechanical model that pertains to being "open." Or acoustical thing that is happening. Many ways to say it.

I played a 1947 Becker for several years.