Why some violins "carry" better than others

[Ganymede Piggot]

11 hours ago, Don Noon said:

I have never played a cheap but well made instrument.  The cheap ones I have are all poorly made, and not loud.

I have a cheap, well-made one that's loud as dynamite.  i'll temporary switch for one of yours if you'd like the experience...

[sospiri]

15 hours ago, Don Noon said:

I have never played a cheap but well made instrument.  The cheap ones I have are all poorly made, and not loud.

Sure the player matters a lot... but having attended many VMAAI competitions, listening to a single player play the same piece on a slew of violins, it is abundantly clear that violins can vary significantly as well.

You have never played a cheap but well made instrument? Oh dear, I shouldn't really respond to that because you are living in a separate reality to me.

And as for a single player playing the same piece on a slew of violins, can you not see how that would involve a certain degree of inconsistency of concentration and technique?

 

[sospiri]

4 hours ago, Bill Merkel said:

I have a cheap, well-made one that's loud as dynamite.  i'll temporary switch for one of yours if you'd like the experience...

Of course you do. Every good player know this to be true. But the violin world is extremely elitist and the higher up you go the more you have to deny this plain and simple fact.

Whether or not a cheap instrument sounds tonally good is to our ears is a different matter. Whether or not an expensive instrument sounds tonally good to us is again, a different matter. And anyway there are cheap well made instruments made in huge numbers in China. Are we expected to deny this here?

Just because this is Maestronet, doesn't mean we have to pander to the elitism does it?

[sospiri]

19 hours ago, sospiri said:

I'm going to state my case that this Holy Grail of Projection is largely in the skill of the player and their emotions and how they project their personality, and how the listener responds with their own emotion to the music.

 

19 hours ago, GeorgeH said:

In that case, I think it has more to do with how close in proximity the listener is to the player. 

Why? Can't you enjoy the concert equally as much in different parts of the auditorium?

[GeorgeH]

38 minutes ago, sospiri said:

Why? Can't you enjoy the concert equally as much in different parts of the auditorium?

No. This is why front row seats are more expensive than back row seats.

[sospiri]

11 minutes ago, GeorgeH said:

No. This is why front row seats are more expensive than back row seats.

That's your preference, not mine.

[Don Noon]

1 hour ago, sospiri said:

You have never played a cheap but well made instrument? Oh dear, I shouldn't really respond to that because you are living in a separate reality to me.

And as for a single player playing the same piece on a slew of violins, can you not see how that would involve a certain degree of inconsistency of concentration and technique?

"Cheap" does not have a universal definition.  My inventory of cheap violins (for experimental purposes) consists of under $100 junk off eBay.  Relative to professional violin makers, the ones I make are "cheap", and in my biased opinion are well-made and loud.

Yes, a player does have variability.  But if you have ever sat through one of these competitions, you'd hear differences in the instruments that are definitely not the player alone.

[Marty Kasprzyk]

I went to a Joshua Bell outdoor amphitheater concert and was seated so far away (I could afford that) that I estimated it might take about a second for the sound to reach me.  I thought it would be disconcerting to have a mismatch between Bell's fingering and the sound but he was so far away I couldn't hear him or see his hands anyway. 

So violin projection is important to me and I noticed nearby concert goes had binoculars.

[sospiri]

6 minutes ago, Don Noon said:

"Cheap" does not have a universal definition.  My inventory of cheap violins (for experimental purposes) consists of under $100 junk off eBay.  Relative to professional violin makers, the ones I make are "cheap", and in my biased opinion are well-made and loud.

I have bought ebay violins from $100  old German/Bohmian or Czech. Recently a Czech 1920s to 30s violin for $150 this was a very nice instrument but still cheap in its day. All of them with no problems in the loudness department. $200 up to $1000 for new Chinese instruments, same thing, none varied from the norm in loudness.

6 minutes ago, Don Noon said:

Yes, a player does have variability.  But if you have ever sat through one of these competitions, you'd hear differences in the instruments that are definitely not the player alone.

For that to be true there would have to be some character differences which you might refer to as Radiation Ratio, which has not met with universal approval as a reliable pre-condition of what constitutes the best spruce for the belly. I did believe this in 2016 but if it's true, why can't I hear the difference in thousand of hours of playing a variety of instruments?

[sospiri]

11 minutes ago, Marty Kasprzyk said:

I went to a Joshua Bell outdoor amphitheater concert and was seated so far away (I could afford that) that I estimated it might take about a second for the sound to reach me.  I thought it would be disconcerting to have a mismatch between Bell's fingering and the sound but he was so far away I couldn't hear him or see his hands anyway. 

So violin projection is important to me and I noticed nearby concert goes had binoculars.

A thousand feet away?

[Marty Kasprzyk]

15 minutes ago, sospiri said:

A thousand feet away?

Next time I go there I'll pace it off or take a range finder.  Its actually a large concert hall that can have its back opened up to an upward sloping lawn where people can sit on the grass.  I'm quite sure that high frequency sounds are attenuated more than low ones.   The orchestra sounded like dueling fog horns.

[Don Noon]

13 minutes ago, sospiri said:

For that to be true there would have to be some character differences which you might refer to as Radiation Ratio, which has not met with universal approval as a reliable pre-condition of what constitutes the best spruce for the belly. I did believe this in 2016 but if it's true, why can't I hear the difference in thousand of hours of playing a variety of instruments?

Radiation ratio only includes longitudinal stiffness and density.  There are many more wood properties that can be (and I think ARE) also important, like crossgrain stiffness and damping.  Arching, graduations, varnish, and all those other details matter too.  

At one time (not too long ago) I thought radiation ratio was a very important factor (even though the math couldn't show a significant dB advantage), as there was not much else to objectively hang on to.  Now I see that the math is right, and that all those other factors are far more important.

[sospiri]

6 minutes ago, Marty Kasprzyk said:

  I'm quite sure that high frequency sounds are attenuated more than low ones.   The orchestra sounded like dueling fog horns.

I believe this also. How do we relate is to instrument making for microphone free orchestral playing?

[Jim Bress]

13 minutes ago, Marty Kasprzyk said:

Next time I go there I'll pace it off or take a range finder.  Its actually a large concert hall that can have its back opened up to an upward sloping lawn where people can sit on the grass.  I'm quite sure that high frequency sounds are attenuated more than low ones.   The orchestra sounded like dueling fog horns.

Yes they are.  I provided a citation long ago for a similar conversation.

[sospiri]

5 minutes ago, Don Noon said:

Radiation ratio only includes longitudinal stiffness and density.  There are many more wood properties that can be (and I think ARE) also important, like crossgrain stiffness and damping.  Arching, graduations, varnish, and all those other details matter too.  

At one time (not too long ago) I thought radiation ratio was a very important factor (even though the math couldn't show a significant dB advantage), as there was not much else to objectively hang on to.  Now I see that the math is right, and that all those other factors are far more important.

I'm fascinated by it all.

[Greg Sigworth]

Reviewing my physics course work the speed of sound is dependent on the medium the sound is traveling through, not on the frequency of the sound. Mention was made to some differential equation which I did not look up, but I believe that is a misapplication of an equation. The sound radiating from a violin travels at the same speed regardless of the frequency. If the higher frequencies travel at a faster speed it is because at very high frequencies, above those we are concerned with, the wave length is close to the size of the air molecules in air. At that dimension the model for sound in air breaks down and the molecule size becomes a factor actually decreasing the distance the sound has to travel. But you are talking about very very high frequency sound waves.

[martin swan]

1 hour ago, Marty Kasprzyk said:

Next time I go there I'll pace it off or take a range finder.  Its actually a large concert hall that can have its back opened up to an upward sloping lawn where people can sit on the grass.  I'm quite sure that high frequency sounds are attenuated more than low ones.   The orchestra sounded like dueling fog horns.

Of course. High frequency sounds are attenuated over distance, but they travel at the same speed (unless as Greg Sigworth says you are talking about ultra-high frequencies).

So if you get further away you will lose a lot of high frequency information (depending on the nature of the acoustic you are in) but it's not to do with the speed of sound.

 

[Violadamore]

I feel that the answer is somewhere in this direction:

https://www.mathworks.com/company/newsletters/articles/making-all-the-right-noises-shaping-sound-with-audio-beamforming.html

https://en.wikipedia.org/wiki/Beamforming

IMHO, somehow the violin is acting as a multiple-emitter array with interference creating "sweet spots" in the 3-D emission amplitude pattern.  It certainly isn't an isotropic radiator.  

Investigating this is going to be complicated by the facts that creating a sufficiently dense near-field microphone array to test this will disturb the effects we are trying to measure, modeling the pattern instead will be darn near impossible with the tools available, and the probability that quite a bit of the observed/reported effects only occur between the ears of the audience. 

And, as we all know already, every fiddle is different, so what you get from one may contradict other tests.

But it's still some sort of beamforming.  Pardon the pun, it fits the pattern.

[David Burgess]

4 hours ago, sospiri said:

You have never played a cheap but well made instrument? Oh dear, I shouldn't really respond to that because you are living in a separate reality to me.

I can see both sides. Cheap and well-made instruments are hard to find,  and much more expensive instruments (into the multi-million category) can also have their weaknesses.

I would prefer that you wouldn't attempt to trash Don, since in my opinion, he has earned  a place among the most valuable contributors here.

[Don Noon]

8 minutes ago, Violadamore said:

I feel that the answer is somewhere in this direction:

https://www.mathworks.com/company/newsletters/articles/making-all-the-right-noises-shaping-sound-with-audio-beamforming.html

https://en.wikipedia.org/wiki/Beamforming

IMHO, somehow the violin is acting as a multiple-emitter array with interference creating "sweet spots" in the 3-D emission amplitude pattern.  It certainly isn't an isotropic radiator.  

I agree that the violin can be considered a multiple-emitter array at the higher frequencies, and although it isn't an isotropic radiator, I prefer to think of it as a chaotic radiator.  For most conditions, I don't think that beaming is that big of a deal.  For example, if you have ever switched speaker leads around to get them in phase, it's just the low frequencies that really stand out... the middle and upper frequencies sound about the same (to me, anyway) even though the phases are different and the beaming is all different... but it's too chaotic and with room reflections, it's not that significant.

The one case I mentioned before... in the mid-frequency ranges... I think that the antinode separation and frequency can line up to beam to the player, which CAN be significant, and make annoying "hot" notes that are not heard at other listening positions.

[martin swan]

32 minutes ago, Violadamore said:

 

But it's still some sort of beamforming.  Pardon the pun, it fits the pattern.

I don't think so.

Beamforming (in common parlance) is either to do with creating a beam that can be heard at a specific point only (the opposite of the omni-directionality you would need from a great violin) or to do with multiple speaker arrays that allow audiences at a massive outdoor gig to hear fully amplified sounds in phase.

For a violin to project well to all seats in a concert hall I am personally convinced that this is a combination of decibel level and the right frequency emphasis.

I still think the singer's formant is the best analogy. Operatic projection is something that has been very well studied, and while in a singer it's a muscular and voluntary process rather than (as with a violin) an inherent tonal characteristic, I think we could see it quite well on an FFT.

I had a recent experience with two violins in the Philharmonie which was a real eye-opener. I'll try to write a bit about it.

But certainly anyone who thinks this is "mythology" is either deaf or lacking in serious first-hand experience.

[Violadamore]

4 hours ago, martin swan said:

I don't think so.

Beamforming (in common parlance) is either to do with creating a beam that can be heard at a specific point only (the opposite of the omni-directionality you would need from a great violin) or to do with multiple speaker arrays that allow audiences at a massive outdoor gig to hear fully amplified sounds in phase.

For a violin to project well to all seats in a concert hall I am personally convinced that this is a combination of decibel level and the right frequency emphasis.

I still think the singer's formant is the best analogy. Operatic projection is something that has been very well studied, and while in a singer it's a muscular and voluntary process rather than (as with a violin) an inherent tonal characteristic, I think we could see it quite well on an FFT.

I had a recent experience with two violins in the Philharmonie which was a real eye-opener. I'll try to write a bit about it.

But certainly anyone who thinks this is "mythology" is either deaf or lacking in serious first-hand experience.

My understanding of beamforming (and other wave-based weirdness) comes from a background in military electronics on the one hand, and geophysical interpretation on the other.  In the first, you are usually aiming (P.I.) to make a compact, flat, fixed or semi-fixed array act like it was a giant parabolic reflector being swung freely on its axes.  In the latter, you are trying to reconstruct an image of visually hidden structures from how they reflect, transmit, and otherwise mess with sound or radio waves pulsed through them.  When dealing with emitters, one usually knows what specific frequency ( and what particular geometry) one needs to plug into the math to get a pretty chart that you can wave around the conference room.  With violins, we don't have that luxury.  As Don points out (and I like his term), the violin is a chaotic radiator.  It's emitting on multiple varying frequencies simultaneously, with variable amplitudes, from which of its surfaces for a given band we aren't sure of, with an internal delay circuit and matrix that we really haven't a clue about, except to know that it's different for all of them. Given a few, probably false, assumptions, I have tried to visualize what the standing wave patterns would superpose to be for the spectrum of a single bowed note, and keep coming up with directional patterns which resemble the anecdotal data posted here for concert hall listening experiences.  Here you hear it, there you don't sounds (P.I.) to me like what you get from interference patterns.  OTOH, inside a hall, reflection patterns have to be added in as well, so maybe going outdoors is where er need to be to investigate the violin's inherent emission directionality.  Once that's nailed down a bit (enough to model reliably), we can stick the construct in a hall and see what happens.

I very much agree with the formant hypothesis, but feel that it is a separable phenomenon which deserves its own thread.  I also feel that it will be easier to investigate with generally available gear.

"Projection", IMHO, isn't going to be due to any single effect, but due to a combination and interaction of several of them.  Let's be very tolerant of each other's ideas.  We may all be right.

Martin, I'm much looking forward to your account of the Strange Case of the Two Violins of the Philharmonic.  New data is always good.

In some ways, analyzing violins in general (and more so from a maker's or engineer's POV), is like trying to analyze some chunk of foreign technology dumped on your bench that the people who found it don't really know how it works or what it was used for.......but it does this useful thing, and it does that useful thing, and they'd really like to know how to make more just like it..........oh, and next week will be fine....   

[David Beard]

We are so much speculating in the dark with this topic.

Energy goes into a violin from the player.  Some of it is wasted quietly.  Some of it goes into unmusical noise.  But what we care about is how much energy goes into musically meaningful sound.  Improve that percentage, quality, and balance.

Some of that musical signal energy radiates out from the instrument.  Improve that percentage, quality, and balance.

This radiating signal travels some distance and is disipated and rebalanced in traveling.  Better understand how the signal modifies under various traveling circumstances.

Finally, some portion of the radiating signal reaches the listener at a distance. Better understand how the signal is perceived and judged musically.

 

We are blind for now.   What portion of input energy comes out as musically meangful?   What portions of a sound signal are muiscally meangful?  What portions of the signal convey what?  

Further, all sounds near a frequency aren't created equally.   Harmonically coherent sound energy gives us a 'pitched' experience.  A similar amount of sound energy but harmonically incoherent gives a noise impression. Etc.

We need studies that show frequency spectrum against time, in very small time increments, and that aim to make harmonic coherence very evident when present.  Perhaps some sort of color coding could do this.

With computer tools, this kind of sound analysis might let us begin to grapple with some aspects of what 'good' and 'bad' musical sound mean on physical and informational levels.

Then we might not be so blind in a discussion like this.   Amps generated in a mic might give some gross physical notion of 'volume', but it means nothing against the perceptions of a musical ear.  

 

*******

If instead of looking to science we look to culture, then articulation is something that fiddlers and singers focused on when wanting to reach the back of the auditorium (i.e. see discussion by Galamian).   Also, notions of open resonant Bel Canto tone and 'supported power' come into play.

In other words, the bulk of energy is sent into very harmonious consonant broad core tone, but dressed very abundantly in the noises and timing of enhanced articulation.

******

I'd love to see much more detailed science about what we hear as good and bad sound.   

 

 

[Jim Bress]

1 hour ago, David Beard said:

We are so much speculating in the dark with this topic.

Energy goes into a violin from the player.  Some of it is wasted quietly.  Some of it goes into unmusical noise.  But what we care about is how much energy goes into musically meaningful sound.  Improve that percentage, quality, and balance.

Some of that musical signal energy radiates out from the instrument.  Improve that percentage, quality, and balance.

This radiating signal travels some distance and is disipated and rebalanced in traveling.  Better understand how the signal modifies under various traveling circumstances.

Finally, some portion of the radiating signal reaches the listener at a distance. Better understand how the signal is perceived and judged musically.

 

We are blind for now.   What portion of input energy comes out as musically meangful?   What portions of a sound signal are muiscally meangful?  What portions of the signal convey what?  

Further, all sounds near a frequency aren't created equally.   Harmonically coherent sound energy gives us a 'pitched' experience.  A similar amount of sound energy but harmonically incoherent gives a noise impression. Etc.

We need studies that show frequency spectrum against time, in very small time increments, and that aim to make harmonic coherence very evident when present.  Perhaps some sort of color coding could do this.

With computer tools, this kind of sound analysis might let us begin to grapple with some aspects of what 'good' and 'bad' musical sound mean on physical and informational levels.

Then we might not be so blind in a discussion like this.   Amps generated in a mic might give some gross physical notion of 'volume', but it means nothing against the perceptions of a musical ear.  

 

*******

If instead of looking to science we look to culture, then articulation is something that fiddlers and singers focused on when wanting to reach the back of the auditorium (i.e. see discussion by Galamian).   Also, notions of open resonant Bel Canto tone and 'supported power' come into play.

In other words, the bulk of energy is sent into very harmonious consonant broad core tone, but dressed very abundantly in the noises and timing of enhanced articulation.

******

I'd love to see much more detailed science about what we hear as good and bad sound.   

 

 

Excellent post.  Too often people are stuck on whether something is significant without asking whether the measured difference is meaningful.

-Jim

[A. Strelnikov-Resch]

If this is of some help I can state that over more than 40 years of orchestra work I played with most of the 1960s, 1970s, 1980s, soloists. This question of projection was discussed often and there was consensus that it depends on the soloist and the violin each maybe about half. I can give many many examples but first one very enlightening to me was Maestro Menuhin in the Beethoven concerto in around 1974 when I listen from the back of the hall the first night and played in the orchestra in the second. At the back of the hall his violin sounded like a huge speaker over orchestra. Three meters on stage from him it was nothing and I could not believe people in the audience can hear it. On next Saturday morning the concert was recorded for television and there was the most interesting experiment when Maestro Menuhin and two other great players tested their violins and it became very clear that it is part the player and part the violin. My impression is that it was mostly the violin but others were thinking the player to be about 40%. I have doubts as when Maestro Menuhin tried other indifferent violins it was clear the orchestra must reduce volume a lot and the sound did not go to the end of the hall. My own violin which I used my entire professional life is loud enough and serves well up to a certain hall size but then no more. Very interesting was that on his violin Maestro Menuhin seemed to use almost no bow pressure. I heard other very good soloists playing with high bow pressure and result was not even half as loud.

Of course, the above is just my own personal experience and ideas. It is very good to see how talented makers with scientific minds are trying to decode this problem.