Why some violins "carry" better than others

[Johnmasters]

A possible explanation for why some "soft" violins carry well.

The physicists and engineers here can comment...  a signal of finite length (in time) can be considered as a summation of single-freqency components.  (More exactly, a Fourrier integral of retarded components)  These components all travel at slightly different speeds.  The effect is more prominent in optical waves in, for example, glass.  This is what allows a prism to split white light into its component colors.  The fact that this was first seen in optics is likely responsible for the term "Dispersion" although it is not confined to waves in glass.  AIR IS A DISPERSIVE MEDIUM FOR ACOUSTICAL WAVES ALSO.  

Air is almost certainly less dispersive for acoustical waves than is the case for optical waves in glass.  But it exists.  A wave packet will spread in time...  this causes a smearing of the signal in time (distance traveled).  It would be dependent on the bandwidth of the components of the wave packet.  The two central ideas are phase velocity (velocities of the single-frequency components) and group velocity, the average velocity of the entire packet.  The wider the band-width of the packet, the worse the dispersion.  I have seen some statements that perhaps a sharp rolloff of response in a violin at about 7k or less is a good thing.  This would lessen the dispersion of any wave packet signal.  Also, under the ear, the violin might sound less loud.

I consulted an engineer friend in California, and indeed, "phase lag" causes a difficulty in broad-band amplifiers.  I believe it all goes back to considering the Fourrier integral for the signal and then taking the time representation of a signal heard at a distance.  With too much band-width, the signal gets smeared over time.  The higher frequency components are lost to the ear.  And perhaps that is why a violin that "carries" well sounds a bit soft under the ear.  (The player may unconsciously diminish his efforts because of the sensation of loudness under the ear.)

Comments?

[jezzupe]

I think it's a realistic way to look at it. My take on it is that LOTS of these types of perceptions are based off of the "air" in the room particularly the size and airflow in that room.

In a smaller room these effects are not so noticeable as the conditions to present these "personal" acoustic experiences are not really present.

When we get into larger rooms,with stages and large expansive walls/ceilings the ability for the air to filter/dissect elements of the sound as well as push it around in an almost "mish mosh rotary speaker" kind of way dramatically increase the potential for events

I don't think the medium "air" can be anything other than the culprit in these cases. Conditions in large buildings are so variable from moment to moment that it certainly can explain the sometimes you hear it, sometimes you don't

In my early days I did tons experimenting with "carry", lots can be learned by using a powerful fan, a fellow listener, and either a large room,or outside area with some reflective walls, sound can be pushed around,or certainly it can be made to seem that way.

[Don Noon]

2 hours ago, Johnmasters said:

AIR IS A DISPERSIVE MEDIUM FOR ACOUSTICAL WAVES ALSO.  

From wikipedia on "speed of sound":

 In dry air, the speed of sound increases by about 0.1 m/s as the frequency rises from 10 Hz to 100 Hz. For audible frequencies above 100 Hz it is relatively constant.

Notion that air would significantly spread out frequencies didn't seem credible to me... just consider a distant fireworks explosion, which can take a few seconds to arrive.  If it was dispersed, you'd hear a strange sound ramped in frequency, rather than the normal bang.

Certainly the higher frequencies experience more damping, and are filtered out thru loss, but they don't arrive at a different time.

I can't answer about electronics, but comparing sound waves (mechanical pressure waves) to light (individual photons of electromagnetic energy, each with a specific frequency) does not seem all that applicable.

[Marty Kasprzyk]

A light case makes a violin carry better.

I've been making really light violins with the hope that they would carry better but I found this is a waste of effort if the case isn't also light weight.

Of course you can carry this to the extreme case.

[Ken_N]

Of course there is Pete Townshend who sang of a Gibson, without a case.  But he didn't have to worry about a bow too. 

[David Burgess]

4 minutes ago, Don Noon said:

Notion that air would significantly spread out frequencies didn't seem credible to me... just consider a distant fireworks explosion, which can take a few seconds to arrive.  If it was dispersed, you'd hear a strange sound ramped in frequency, rather than the normal bang.
 

Have you not noticed the difference?

[Johnmasters]

7 minutes ago, Don Noon said:

From wikipedia on "speed of sound":

 In dry air, the speed of sound increases by about 0.1 m/s as the frequency rises from 10 Hz to 100 Hz. For audible frequencies above 100 Hz it is relatively constant.

Notion that air would significantly spread out frequencies didn't seem credible to me... just consider a distant fireworks explosion, which can take a few seconds to arrive.  If it was dispersed, you'd hear a strange sound ramped in frequency, rather than the normal bang.

Certainly the higher frequencies experience more damping, and are filtered out thru loss, but they don't arrive at a different time.

I can't answer about electronics, but comparing sound waves (mechanical pressure waves) to light (individual photons of electromagnetic energy, each with a specific frequency) does not seem all that applicable.

But you cannot really hear the late-arriving overtones.   Your ears may saturate.  But more importantly,  I am not sure if a simple fourrier consideration would be right for a shock wave.  Fourrier series suggest a linear superposition of components. This would not apply for a shock wave which compresses the working medium.

Considering a Fourrier integral for a linear superposition is pretty basic,  and I do not see how you can find it to not be relevant unless you reject the notion of phase velocity varying much in the listening situation.   It WILL occurr, the question is by how much.  And if you do not see any dispersion at all in a wave packet,  then why don't all violins sound the same at a distance vs close up?  

[Johnmasters]

1 hour ago, David Burgess said:

Have you not noticed the difference?

And I think that the leading edge (in time) of the wave packet would be more affected.  The initial bang should not see a "ramped" volume because these faster contributions will have passed and been  damped out or at least much weaker than the peak of the pulse.

But yes,  I see your point and will do a calculation with your numbers for a non-shock wave such as a violin note.

100 Hz is certainly a low frequency, and I don't know what to say.  I have no idea of what constant would be.  At 30 meters,  how far ahead of a 3000Hz wave would be a 7000 Hz wave?  And would their separation then be at the level of a half wave of so?  I choose 3000Hz as it is approximately the upper level for useful music hearing (for me at least)

Do you have any ideas involving acousical dispersion ,   or do you think it is not something to consider?

Also, your example of an explosion still does not seem to be a good example using your numbers.

 

[Don Noon]

Even if you use the extreme case of .1m/s difference in speed (which is not true) for a 3000 Hz tone, at a distance of 30 meters, that calculates out to 26 microseconds, which would be 8% of a wavelength of the 3000 Hz tone.

This is a complete red herring.

4 hours ago, Johnmasters said:

And if you do not see any dispersion at all in a wave packet,  then why don't all violins sound the same at a distance vs close up?  

Um... room acoustics would seem to be a biggie.

Also they sound louder when you're closer.

[Don Noon]

4 hours ago, David Burgess said:

Have you not noticed the difference?

Of course there is a difference... the farther away you are from a boom, the more the high frequency content gets filtered out.  It would sound very strange if the low and high frequencies arrived at different times.

[Violadamore]

5 hours ago, Don Noon said:

Even if you use the extreme case of .1m/s difference in speed (which is not true) for a 3000 Hz tone, at a distance of 30 meters, that calculates out to 26 microseconds, which would be 8% of a wavelength of the 3000 Hz tone.

This is a complete red herring.

Um... room acoustics would seem to be a biggie.

Also they sound louder when you're closer.

IMHO, the better-attested cases of "projection" occasionally described on MN fit the pattern for interference effects.  Some of what's observed probably involves beam-forming.

That said, given the immense quantity of variables unique to any single performance, with regard to the instrument, the bow, the strings used, the performer, the enclosure, and the atmospheric conditions, I currently despair of being able to create a rigorous treatment of the dynamic pressure fields involved, capable of producing a tightly-meshed 3D plot of the fields (summed over all audible frequencies in the generated spectrum) throughout a performance, of say, a classical sonata.

I feel that anything short of that will unfortunately fail to settle these matters. 

[HoGo]

OMG, I thought this is violin MAKING forum....

Back to OP.... Lets not forget different sensitivity of human ears at various frequencies and ability of brain (especially trained one) to make up bunch of the lost components just like with speach in noisy environment. I believe there were studies that pointed to that.

 

[David Burgess]

8 hours ago, Don Noon said:

Of course there is a difference... the farther away you are from a boom, the more the high frequency content gets filtered out.  It would sound very strange if the low and high frequencies arrived at different times.

One thing I've noticed with lightning is that up close, the sound is of very short duration, while from far away, the sound is spread over a longer time interval. I guess I've never tried to pay attention to frequency  content variations over these longer intervals, which would be expected if the longer duration was due to high and low frequencies arriving at different times.

[HoGo]

54 minutes ago, David Burgess said:

One thing I've noticed with lightning is that up close, the sound is of very short duration, while from far away, the sound is spread over a longer time interval. I guess I've never tried to pay attention to frequency  content variations over these longer intervals, which would be expected if the longer duration was due to high and low frequencies arriving at different times.

But the longer duration can be due to echoes from terrain between you and the lightning...

[Don Noon]

5 hours ago, David Burgess said:

One thing I've noticed with lightning is that up close, the sound is of very short duration, while from far away, the sound is spread over a longer time interval.

Lightning is usually a quite long (length) electrical discharge that happens essentially instantly.  But sound is slow, so sound from the far part of the bolt can reach you many seconds later than the near part.  If the whole thing is relatively far away, then you'll get the rolling thunder.   If it hits close, you're so deafened by the near part that you might not notice the softer rolling part from the far end, or ascribe it to echoes.

For the violin question about carrying power, it is my opinion that fullness of the response spectrum, probably in the range of 1.2-4 kHz, has a lot to do with it.  The more overtones that you have, the easier it is for the brain to assemble a recognizable note from that particular instrument, and fullness of that part of the spectrum means that all of the notes played can be identified, not just a few of the notes that happen to line up with peaks in the response.  Of course, overall dB level can help too.

[David Beard]

20 hours ago, Marty Kasprzyk said:

A light case makes a violin carry better.

I've been making really light violins with the hope that they would carry better but I found this is a waste of effort if the case isn't also light weight.

Of course you can carry this to the extreme case.

!!!

[David Beard]

7 hours ago, David Burgess said:

One thing I've noticed with lightning is that up close, the sound is of very short duration, while from far away, the sound is spread over a longer time interval. I guess I've never tried to pay attention to frequency  content variations over these longer intervals, which would be expected if the longer duration was due to high and low frequencies arriving at different times.

There can be some time dispersion of overall signal due to sound reaching you through somewhat different paths.  But the key point is that audible sound dispersion is not significantly frequency dependent.

 

[Okawbow]

I had a good performer play 4 of my best sounding violins on stage in a noisy, medium size hall. He was accompanied by guitars, piano, and banjos. My Guarneri copy was so loud under my ear that it made my ears ring, but could barely be heard across the room in that hall. My Stainer copy was very soft and quiet under my ear, but was very clear and projected to every corner of the hall. It almost over powered the other instruments. The other two Strad copies were a little better than the Guarneri, but not nearly as good as the Stainer.

 

Is is that the effect you are discussing?

[David Burgess]

2 hours ago, Okawbow said:

I had a good performer play 4 of my best sounding violins on stage in a noisy, medium size hall. He was accompanied by guitars, piano, and banjos. My Guarneri copy was so loud under my ear that it made my ears ring, but could barely be heard across the room in that hall.

Do you think the violin was actually loud under your ear, or might it have only been "harsh"?

[Okawbow]

1 hour ago, David Burgess said:

Do you think the violin was actually loud under your ear, or might it have only been "harsh"?

When a good player plays the Guarneri copy, it sounds powerful and very clear. It seems very loud under my ear, but not really harsh. It definitely has more “punch” than the Stainer. 

[DBCooper]

I think David touches on a good point.  Often, I think that this "phenomenon" of great violins sounding softer under the ear is just miscommunication and misunderstanding.  

 

What the player is hearing in these cases is less harshness and more quality, which they may equate to "softer" because they are used to a harsh under-ear sound that doesn't carry all that well.  I've played many violins that sound brash and loud under ear, but the frequencies they produce do not come together properly to actually project well.

[Johnmasters]

54 minutes ago, DBCooper said:

I think David touches on a good point.  Often, I think that this "phenomenon" of great violins sounding softer under the ear is just miscommunication and misunderstanding.  

 

What the player is hearing in these cases is less harshness and more quality, which they may equate to "softer" because they are used to a harsh under-ear sound that doesn't carry all that well.  I've played many violins that sound brash and loud under ear, but the frequencies they produce do not come together properly to actually project well.

That is true,  and I was thinking of the high-frequency noise that you speak about.  These are what I suggested had a higher phase velocity.   

There is another possible thing to think about...   damping of acoustical waves depends on frequency...  It would be true even if these higher frequency components did not travel faster than the entire tone (wave packet.)   I found a differential equaition  here.

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

It is the second equation,   but it would take me a month to plow through it..........  Maybe Don can at least tell us if attenuation increases with frequency.   That would also explain your example.

[Don Noon]

4 hours ago, Johnmasters said:

Maybe Don can at least tell us if attenuation increases with frequency.

I thought that was common knowledge. 

[Anders Buen]

In general damping of plate structures, at least in building acoustics applications, will usually drop with frequency.

Example giving a gypsum wall will have an loss factor eta = eta_internal + 0,4 or 0,8/(frequency)^(1/2)

0,4 is for empty walls 0,8 is for walls with mineral wool inside. I have measured loss factors in a concert hall stage floor once and this relation fitted the data pretty well for the 0,8 version as it was glass wool under it. The wood was 22mm merbau, tongue and groove single planks 60 mm x 300 mm, 600 mm or 900 mm. Supports every CC 300mm.

[Don Noon]

I thought we were talking about damping of the air, rather than structures.