58 replies to this topic

[Anders Buen]

Claudias questions reposted in a new thread, hopefully making it possible to focus better on the subject. Put the last square bracet back on the end of quote mark to get Claudia as the right to quote in any answers.
[quote name='Claudia Fritz' timestamp='1340657795' post='547426']
Here are a few questions that I'd love to see discussed on this forum as that could be very constructive and fruitful:
* What is projection for you as a maker? as a player? as a listener?
* What is the influence of the player on the projection of an instrument?
* What is the influence of the venue on the projection of an instrument?
* What is the influence on what is played?
* What are the valid conditions to evaluate the projection of an instrument (for instance, what kind of accompaniment?) from the player's point of view? From the listener's point of view?
* Do you think listeners would agree on projection when hearing different violins played by the same player?
* Do you think the listeners' evaluation would agree with what the player thinks/estimates from what he hears under his ear?
* What are the elements to make a violin project? (for instance brightness?)
* Do you think that instruments that are quiet under the ear can indeed project well?
And one more question which does no relate to projection:
* Do you think you can recognize when a violinist plays on his own violin compared to other violins, behind a screen of course (so no body language!)?
Thank you so much in advance for your responses, and sorry for these two very long posts!
[/quote]
[quote name='Claudia Fritz' timestamp='1340657795' post='547426']
Here are a few questions that I'd love to see discussed on this forum as that could be very constructive and fruitful:
* What is projection for you as a maker? as a player? as a listener?
* What is the influence of the player on the projection of an instrument?
* What is the influence of the venue on the projection of an instrument?
* What is the influence on what is played?
* What are the valid conditions to evaluate the projection of an instrument (for instance, what kind of accompaniment?) from the player's point of view? From the listener's point of view?
* Do you think listeners would agree on projection when hearing different violins played by the same player?
* Do you think the listeners' evaluation would agree with what the player thinks/estimates from what he hears under his ear?
* What are the elements to make a violin project? (for instance brightness?)
* Do you think that instruments that are quiet under the ear can indeed project well?
And one more question which does no relate to projection:
* Do you think you can recognize when a violinist plays on his own violin compared to other violins, behind a screen of course (so no body language!)?
Thank you so much in advance for your responses, and sorry for these two very long posts!
[/quote

[lyndon]

if we were to eliminate the contribution of reflected sound and deal just with direct sound, the difference between projection and under ear sound would be very easy. projected sound would be the amount of sound coming of the right side of the violin directed at the audience, and under ear sound would be largely the sound coming off the top of the violin directed at the player, these two can be closely linked, or sometimes not linked well at all.

at given frequencies a violin that puts out in different directions a mostly in phase signal is going to be louder than a violin that puts out a mix of in and out of phase signals(all other factors being equal). i theorize projection has a lot to do with the relative phases of the output signals(a function of the arching, shapes, sizes and graduations of the violin), as well as the shear amount of wood vibration, hence displacement of air that we hear as sound

the idea that projection is largely a factor of increased highs seems to be an oversimplification, quite likely a brighter violin is brighter at a distance, but that doesnt make it better, there are plenty of not so bright violins that project very well, if the brightness hypothesis were true, every violin could get better projection by putting on brighter strings, like evah pirazzi

to do a real scientific study of projection vs under ear volume, a simple test would work, two microphones, one near the players ear(actually one for each ear or binaural mikes might be better), and one well back in the hall, then test multiples of violins with good and bad projection, the compare the relative volumes of the two mikes signals,

i think a study like this would produce hard scientific evidence that under ear volume is not directly proportional to projection, whereas a subjective study like claudia is proposing couldnt give hard proof, only fallable subjective indications. this doesnt mean that other parts of her study wouldnt be very useful, its just if i want a quantifiably measure of a violins volume, i think a microphone is more reliable than a bunch of listeners opinions. on the other hand if i were testing the subjective quality of a violins sounds, listeners can tell me much more than a microphone

[Michael_Molnar]

lyndon,

Are you suggesting an anechoic chamber? Eliminating reflections is an interesting idea. I just wonder whether that is feasible. But then again is that desirable? I don't know.

Mike

[DGerald StephenR]

if we were to eliminate the contribution of reflected sound and deal just with direct sound, the difference between projection and under ear sound would be very easy. projected sound would be the amount of sound coming of the right side of the violin directed at the audience, and under ear sound would be largely the sound coming off the top of the violin directed at the player, these two can be closely linked, or sometimes not linked well at all.

at given frequencies a violin that puts out in different directions a mostly in phase signal is going to be louder than a violin that puts out a mix of in and out of phase signals(all other factors being equal). i theorize projection has a lot to do with the relative phases of the output signals(a function of the arching, shapes, sizes and graduations of the violin), as well as the shear amount of wood vibration, hence displacement of air that we hear as sound

the idea that projection is largely a factor of increased highs seems to be an oversimplification, quite likely a brighter violin is brighter at a distance, but that doesnt make it better, there are plenty of not so bright violins that project very well, if the brightness hypothesis were true, every violin could get better projection by putting on brighter strings, like evah pirazzi

to do a real scientific study of projection vs under ear volume, a simple test would work, two microphones, one near the players ear(actually one for each ear or binaural mikes might be better), and one well back in the hall, then test multiples of violins with good and bad projection, the compare the relative volumes of the two mikes signals,

i think a study like this would produce hard scientific evidence that under ear volume is not directly proportional to projection, whereas a subjective study like claudia is proposing couldnt give hard proof, only fallable subjective indications. this doesnt mean that other parts of her study wouldnt be very useful, its just if i want a quantifiably measure of a violins volume, i think a microphone is more reliable than a bunch of listeners opinions. on the other hand if i were testing the subjective quality of a violins sounds, listeners can tell me much more than a microphone

I made, what seemed to me the obvious suggestion, of using the unbiased microphone for testing "projection" in the first thread but it was pretty heavily shot down. Something to do with a 50yr old test and ghost sounds or psychotics or something. It is apparently much better to use humans that could all hear differently produced sounds coming from the same instrument at the same time in order to do a scientific study
I don't see why using microphones, at very least in conjunction with human listeners, isn't a standard for this sort of test. If microphones are so bad (or is it so good?) then why are they used at all in tests such as mode frequency strengths and the like?
I would hypothesize that microphones pick up every sound, even these "psychotic ghost" sounds and essentially hear "better" than us humans. To me, that sort of reliable data that could be documented from something as unbiased and technically good at retrieving sound, as a microphone, would be a logical and most reliable starting point for sound "projection" analysis. Which is apparently what this test is about.
In my initial suggestion of microphones, David made the joking comment "Am I going to tell a customer, "Five famous microphones really liked this fiddle?"". But in the alternative, is it that "five famous listeners really liked my fiddle?" Sorry, but it will absolutely not be the most defined of listeners hearing these instruments on a daily basis. It will be the musician first and foremost but then, it will be concert goers; the old blue-hairs who have had a love of classical music their entire lives. It will be the wealthy, whom are there for the "scene" of it all. It will be the young children who are starting their appreciation of classical music. It will be the nervous couple out on an awkward first date, trying to impress each other. It will be a whole myriad of listeners from young to old.
If microphones were used, this sort of study could also get much larger in its statistical trial pool. The recordings could be sent out to music colleges and universities to have students of music partake in it. I recall hearing tests as a child where random tones and frequencies would be generated at different volumes into each ear via headphones and I would need to distinguish where they were coming from, if at all audible. Something like this could take place with the recorded violin sounds. When we're talking "projection", I believe this is possible. It could be a fun little trial for performance attendees to partake in during intermission. The larger the statistical pool, the more accurate the results get. Limiting this kind of test to a hand-full of people (with all their foils as human listeners) and then trying to state fact, is just not science IMHO.
But then again, I guess microphones are a bad thing.

[lyndon]

michael i was talking theoretically, though your dead right, the relationship of direct sound output in different directions could be realistically tested in an anechoic chamber, but that doesnt mean it means much in the real world where reflected sound is always a factor.

anyone who wants a simple volume test comparing violins can be more accurately and consistently tested by microphones than people, more subjective qualities need people, though an "opinion" from a microphone may help as well, for instance recording tonal balance or frequency response

there are two distinct meanings to projection, one is perceived or measured volume at a distance, two is perceived or measured volume at a distance divided by under ear volume, the ratio of how loud the sound at a distance is compared to the sound up close

for instance if a quiet violin still sounds decent at a distance, we say it carries well/projects well for its volume, if a really loud violin under ear doesnt sound really loud at a distance, we say its loud up close but doesnt project well

i tend to say definition two is more correct, good projection is the ability to project more of the sound that you started with(under ear volume) than is average

im proposing two standards, projection being the ratio of volume at a distance/volume up close and a second term????? often called projection, being the actual volume of a violin at a distance, which is obviously going to favour violins that are both loud up close and project well also.

so i leave it to the experts to decide, what is the true definition of projection, the ability of a violin to spread what little or great sound that it starts with in the direction of the audience, or simply which violin spreads the most sound in the direction of the audience.

for instance if you switch your violins strings from really quiet ones to really loud ones, youve increased your violins volume both up close and at a distance, but has its projection improved, i think not.....

[Don Noon]

I'll just pick and choose a few from the list:

* What are the elements to make a violin project? (for instance brightness?)
Short answer: broad frequency response in the range to which the ear is most sensitive.
The ear is most sensitive to the range normally considered "bright", so yes, good projection should favor bright instruments.
Broadness of the response is important too, IMHO. There is only limited power in the vibrating string, and it is distributed in a fairly smooth way across a very wide frequency range. It would seem the most effective way to use the power would be to use ALL of it... i.e. a wide frequency band. In the ear, there also might be benefit for the wide band approach.

I have some semi-measured support for this concept. Fan Tao's violin was reported to have the best projection in a "vacuum cleaner" test at Oberlin some years ago (I wasn't there)... where a vacuum cleaner was used to create background noise. The spectrum I measured on his violin is very unusual: relatively flat response from 1200 to 3500 Hz, without the usual dip in the 1200 - 2000 Hz range. In other words: very wideband response in the range where the ear is most sensitive.


* Do you think that instruments that are quiet under the ear can indeed project well?
A qualified "yes", and related to my first answer. A broadband response might not sound terribly loud, but be recognizable as the intended note. An instrument with some extremely strong response peaks and deep dips can sound annoyingly loud, but not in a well-defined way.
There may also be some relationship to what areas of the violin body are vibrating and how: If upper and lower bouts are producing out-of-phase sounds, it can be very loud at the player's ear, but cancel out in the far field. Also any tones from the lower bout will sound louder under the ear, whereas those from the upper bout (of equal far-field volume) will sound less loud to the player.
However, everything else being equal, louder here should be louder there.

Somewhat related thought, but doesn't answer any specific question.
Projecion might also be related to what the player can do with the instrument, in terms of bow force and speed to get maximum power into the string. I haven't fully formed the thoughts on this, but perhaps a stiff instrument (which might not sound particularly loud under normal circumstances) allows the player to use more bow force, closer to the bridge, and therefore ultimately allow more total power to be produced. It will project well, when asked to.

[Anders Buen]

Brightness is a timbre descriptor and it can be achieved by both removing the lows or enhancing the highs. Comparing two similar fiddles, one with muted lows and one without any muting, the latter will win the projection contest under my understanding of projection. The muted lows fiddle will sound brighter, but will not project better.

[Anders Buen]

There is an interesting theorem by Lord Rayleigh on the sound radiation from closed shells. It is saying something like the radiated power of that shell structure can be estimated from its internal sound level, as the surfaces will radiate equally much to both sides of the plate or shell.

So it is possible that the sound power level of a fiddle can be estimated from what comes out of the f-holes. These are close to the players left ears so maybe a fiddler does hear the overall projection capability from what comes out of the f-holes?

[lyndon]

once again i hear people confusing volume(up close) with projection, volume is how much sound a violin produces(total), projection is simply how much of that sound is directed towards the audience, and under ear volume is how much of that sound is directed right at the player

don makes a very good point, if the top and back are out of phase, the sound directed sideways towards the audience is going to cancel out, whereas the sound, mostly from the top, directed at the player is still going to sound loud

on the other hand if the top and back are in phase, the sound directed sideways at the audience is going to be twice as loud (6db) and the sound directed at the player not much louder than the out of phase example

in both cases the player hears the same signal, but the audience hears either a loud or a quiet signal, explaining why under ear volume can be so deceptive in the real world

[David Burgess]

Since the eventual goal will probably be to integrate listener impressions with measurements, it would be nice to have both listeners and microphones present.

Without good data on how the two relate, and since the standard now and thoughout history is sound assessment from live playing and listening, it would be premature just to use microphones.

[Marty Kasprzyk]

I'm having my viola tested for projection tomorrow. Should I specify overhand or underhand throws or should I just let the players decide what to do?

[Don Noon]

Marty,
Whatever is required to get it to project to the back of a large hall would be good.
At least you would win for volume.

[robertdo]

I'm having my viola tested for projection tomorrow. Should I specify overhand or underhand throws or should I just let the players decide what to do?

I thought you were talking about baseball so I read this. that should help you to decide...

[carl stross]

There is an interesting theorem by Lord Rayleigh on the sound radiation from closed shells. It is saying something like the radiated power of that shell structure can be estimated from its internal sound level,

Reading the good book ? Respect !

[Anders Buen]

Reading the good book ? Respect !

No, but have read one who might havd read the good book. It is for sale from ASA, and it is on my wish list. He was just great and touched a huge set of pheomena. He is said to be a very good scietific writer too. To the point all the time, and no BS! :-)

[lyndon]

Since the eventual goal will probably be to integrate listener impressions with measurements, it would be nice to have both listeners and microphones present.

Without good data on how the two relate, and since the standard now and thoughout history is sound assessment from live playing and listening, it would be premature just to use microphones.

i never said use mikes for everything, i just suggested for tests measuring volume, mikes are much more reliable than peoples subjective perception, especially if there is any element of time between samples, although even on volume there are spectral differences that mikes might have trouble picking up on, that might make a difference to a listener

[carl stross]

No, but have read one who might havd read the good book. It is for sale from ASA, and it is on my wish list. He was just great and touched a huge set of pheomena. He is said to be a very good scietific writer too. To the point all the time, and no BS! :-)

Give us a title.

[Anders Buen]

Give us a title.

Do you mean Rayleighs book or the source I have this from?

There are several sources. One is Colin Gough in person at Oberlin or in Cambridge. But he has also written about it in one of his early papers on measuring acoustic scatter cross sections of violins based on Raylighs theory. (The scattered wave energy will be equal to the incident wave cross section on the violin in some way. I do not remember the point whith this exactly, but the spehrically radiated modes will tend to give an uniform scattering). I think it is a SMAC (Stockholm Musical Acoustics Conference) paper from 1983 (Acoustic Resonant scattering- a possible new technique for studying musical instruments. C.E.Gough. Proc. International Musical Acoustics Conference (SMAC83) pp 197-206 (Royal Swedish Academy of Music, Stokholm, 1985)). I might scan it when I have more time for that. Guettler, Jansson and Morset tried it on some violins, I think, and the idea has been used in loudspeaker tests. (The reference used in Guettler, Jansson and Morsets article has the theory originally from Raylighs book, I think)

There are simple theories for how the soundfield inside a room and the sound escaping from an open door or an opening relates given that the soundfield inside the room is diffuse (equal sound levels throughout the room). Not true for the lowest mode frequencies of a violin but more so for the highs I think. Maybe the acoustic source strength can be measured just knowing what the sound levels are in the f-holes along with some information about the area of the f-holes.

[carl stross]

Do you mean Rayleighs book or the source I have this from?

There are several sources. One is Colin Gough in person at Oberlin or in Cambridge. But he has also written about it in one of his early papers on measuring acoustic scatter cross sections of violins based on Raylighs theory. (The scattered wave energy will be equal to the incident wave cross section on the violin in some way. I do not remember the point whith this exactly, but the spehrically radiated modes will tend to give an uniform scattering). I think it is a SMAC (Stockholm Musical Acoustics Conference) paper from 1983. I might scan it when I have more time for that. Guettler, Jansson and Morset tried it on some violins, I think, and the idea has been used in loudspeaker tests. (The reference used in Guettler, Jansson and Morsets article has the theory originally from Raylighs book, I think)

There are simple theories for how the soundfield inside a room and the sound escaping from an open door or an opening relates given that the soundfield inside the room is diffuse (equal sound levels throughout the room). Not true for the lowest mode frequencies of a violin but more so for the highs I think. Maybe the acoustic source strength can be measured just knowing what the sound levels are in the f-holes along with some information about the area of the f-holes.

Rayleighs "Theory of Sound" is a free d/load from Internet Archive. I'll give these some thought - there might be something here. Thanks.

[Marty Kasprzyk]

Jascha Heiftz's 1940 testing of new and old violins led Saunders (1) to make this statement: "…it seemed obvious…that the artists wanted two qualities in a violin: first, great power; second and even distribution of strength among all ranges of frequency, the lowest octave being of special importance."

Fast forward 72 years and John Soloninka (2) said this in Claudia's test regarding his favorite picked from new and old violins:

"It feels like it would really project at the back of a hall. It’s a good complex quality of sound. No dead spots… Big powerful G. You don't drop of a few dB as you go across the strings."

Players seem to like power and evenness.

I think we should simply measure the individual played note loudness with a sound level meter across the entire range (Saunders loudness test) and see how that correlates with good "projection" and "complexity".


1, F.A. Saunders, "The Mechanical Action of Instruments of the Violin Family", 1946, The Journal of the Acoustical Society of America, Volume 17, No. 3, kindly provided by Anders Buen

2, John Solonika, Exclusive: "how I blind-tested old violins against new", January 6, 2012 By Norman Lebrecht in Slipped Disc, link kindly provided by Roberto in his post #383 in Followup to the Indianapolis blind study, June 28