Questions for the Tech Folk: Power Versus Volume in Sound?

[David Beard]

I'm wondering how the power behind a sound and the perceived loudness differ, or if they differ??

 

Somehow it seems like there is some distinction and qualitative difference between a wimpy quiet sound, and a quiet seeming tone with power behind it.  But I'm unclear on how this can be, or if it can be in a quantitative sense.

 

And how does carrying strength of a tone relate to this.   It seems as if two tones can have a comparable sense of loudness, but one can have greater power behind it and carry better??

 

 

 

Same question in different guise:   If I play back the same recording of a favorite string quartet on two very different quality stereo systems, one very good quality and high power, the other less on both counts, but I play back on both at seemingly equal loudness levels, how will the playback sounds actually differ?

 

 

Would love to understand this better.

[palousian]

I think that there are at least three related components here.  The most obvious one is sound pressure level--the actual volume of the vibrations.  Connected with this, however, is the timbre of the sound--WHICH frequencies are being amplified?  If the sound is amplified, some parts of the spectrum might be terrifically strong, but if it is above the threshold of human hearing, for example, they won't be audible.  Though they can mess things up if you are trying to record.  The human ear is not neutral, and we respond to different frequencies differently, so if, say, the area around 2K is amplified, we will probably perceive more "loudness" than if the area around 10K is boosted by the same amount.  The relationship of overall volume to which frequencies are being amplified gives most of the impression of "loudness," but there is one other component--envelope.  That is, the responsiveness (or lack of it) in the amplification.  So if the amplifier (electronic or musical instrument or medium through which sound is traveling) is slow to respond, and the attack of the sound is very fast, the amplifier will act like a compressor and effectively mute that attack.  For that reason, if you play a recording on a system with poor-quality amplification, some of the muted/deadened effect will be because crisp attacks and sudden shifts in timbre are being lost through the slower response time of the amplification.  Often this sort of effect is also frequency-specific, so the amplifier might respond quicker to some frequencies than others, and that will contribute to a perception of distortion.  These differences can introduce phase cancellation as well--the same basic technique that an equalizer uses to boost or cut frequencies, so this can begin to change the timbre of the source as well.   

[Don Noon]

I have my own concepts of what power/loudness/carrying power are, and I expect others may have different ideas.

 

We have to start with the realization that a violin note is not a single frequency, but a whole bunch of them.  On an open G, for example, I counted over 50 overtones on a bowed note response plot, although probably only the first two dozen (up to 5 kHz) are important.

 

I think the perceived power may be close to the actual power, i.e. the sum of all the energy in all of the overtones and fundamental.  Since there may be a practical limit to how much amplitude you can get out of a single frequency, the power may also be related to how even and dip-free your instrument response is.

 

Strong output in the signature mode range (below 700 Hz) I'd call "big", or "full", or "round"... and element of power, but not the only part.

 

A strong output around 1 kHz (plus or minus a few hundred) is what I feel is LOUD, but if it is lacking other overtones, I don't think it carries very well.

 

The higher frequencies (~1300 and up) is where I think most of the carrying power, clarity, and projection reside.  Again, if there are dropouts, it won't carry so well.  And above 4-5 kHz, I think you get some crackle, grit, and harshness; some is good, though.

 

A couple of Strads I have played seemed a bit wimpy and weak under the ear, yet when played by someone else, the notes were very clean and clear.  I have convinced myself that the commonly-observed modest (or even weak) response of Strads around 1 kHz and strong response in the higher frequencies leads to that phenomenon.

[Dwight Brown]

https://en.wikipedia.org/wiki/Equal-loudness_contour

 

Perhaps this may be of some application.  There can be a great difference between the way an instrument feels and sounds under the ear and how it sounds in a hall.  I know I have been told and read that some great instruments that sound wonderful are not all that easy to play.  The hall, of course makes a big difference.  There have been some that have been re-done 3 or 4 times (Philharmonic/Avery Fisher/ David Geffen Hall)  I imagine there are as may variables in violin design and building as there are in concert hall acoustic design) 

 

From the standpoint of scientific method I guess it would be important to eliminate as many variables as possible to objectively test instruments.  A mechanical bow substitute, standard anechoic chamber, same strings, temperature, barometric pressure, humidity, etc.  I think it would be quite difficult, but not impossible.  I do know it would take someone much smarter and better educated than the likes of me.

 

Experimental design would be the place to start along with selecting examples of fine instruments that are universally considered to be among the very best as played by world class players.

 

dlb

 

well I tried.........

[ctanzio]

Sound power is the energy the violin is pumping into the air per unit time. So think of it as vibrations that are resisted by the surrounding air. If a note cause the top plate to vibrate strongly, then the violin will sound that note with a lot of power. OTOH, a large mass, like the fingerboard or the chin rest, can suck up a lot of energy and move very little air. These can kill the power transfer to the air but give the false impression that the violin is "powerful".

 

The raw sound intensity is the power per unit area that hits a listener's ear. There are three factors to consider:

 

1. How the sound waves spread out from the violin,

 

2. How the energy of the sound waved is damped (converted into another form of energy, mostly heat) by the air, and

 

3. The sensitivity of the ear to different frequencies.

 

1. A violin projects sound approximately in a spherical shape.  So if P is the total power being pumped into a note, the undamped intensity decreases as 1/(distance from the violin squared), i.e., inverse to the area of a sphere. So a hall that expands out from the stage in a more spherical shape shares the power of the violin (and the orchestra) in a more uniform manner. A hall that is very deep will have problems pleasing the people in the cheap seats way in the back without driving those in the front orchestra-level seats deaf.

 

2. Air dampens sound intensity inversely proportional to frequency squared. So if high frequencies are associated with violins that project, it is because they are more efficient at overcoming this damping effect.

 

3. Once a sound wave hits the ear, the perceived loudness is a complex function of raw sound intensity and frequency. Luckily, the human ear is very sensitive to frequencies in the 1k to 5k range, so the damping due to effect 2, above, is partially offset by ear sensitivity.

[David Beard]

So what I think I'm hearing is that lower frequency tones and partials can carry increased energy without sounding particularly louder, but that comparatively smaller changes in the range of 1k to 5k have a stronger effect on our sense of loudness?

 

I guess means there is some room for the player to make a more full spectrum tone, with lots of additional power in the lower parts of the tone, with particularly sounding 'louder'.

 

 

So are there more aspects to explaining the difference between a ppp that floats out and fills the room, versus a pp that gets lost before it barely leaves the stage??

[DGV]

This thread reminds me of a comment I heard regarding Domingo's voice. "He is not loud but his voice carries all the way to the back of the hall." It probably means he has more energy in the 3k-5k range. 

[Johnmasters]

If one plays a recording on a stereo and turns the volume way down in an ff passage, you can still tell that it is "loud." Even if you do not know the piece.

Likewise, turning up the volume in a pp passage. Something is there other than amplitude of air pressure at the eardrum. Is it overtone sequences or what?

[ctanzio]

We are starting to cross over into the realm of psychological perception of tone as opposed the basic physics of power, intensity and loudness. I would think these are questions better posed to a professional soloist.

 

For example, if one wants a passage to sound piano, yet still fill a large hall, clearly something more must be done than simply playing fast and light towards the board or slowing the bow stroke and pressure near the bridge. This will please the people in the front or in the back of the hall, but not both. It would be especially challenging for notes higher up on the E string.

 

Maybe the width and tempo of vibrato can be used to convey a soft tone even when playing loudly? Maybe playing with as little attack as possible can do it? Maybe a light left hand touch so certain partials are suppressed, ala playing harmonics? Don't really know.

 

YouTube concert recordings can be deceptive since they are frequently closely mic'd and subject to professional editing.

 

There is no loss of players who can project expressively in recital rooms and recording studios. But I suspect that one of the things that makes a great soloist great is the ability to "fool" an audience in a concert hall into experiencing a wide range of dynamics and articulations with technique rather than basic physics.

[David Burgess]

Many years ago, a well-know person in the field of violin acoustics told me that the greatest volume was produced not by digging in, but by using very fast and lighter bow strokes. That didn't sound right to me, so I tested it, and found that it was true for measured volume. But not for perceived volume or the intensity of the sound. Measured loudness and perceived loudness can be very different things.

[Dwight Brown]

Perhaps the Timbre and the wave form (shape) make a big difference?

 

 

dlb

[DGV]

Perhaps the Timbre and the wave form (shape) make a big difference?

 

 

dlb

 

Yes.  A saw wave always sounds louder than a sine wave at the same amplitude.

[ctanzio]

I've never conducted an experiment to quantify this, but I can think of an example to illustrate David's point about the difference between measured and perceived loudness.

 

When playing close to the bridge, the tone is more brilliant than when playing the same note half-way between bridge and board. Playing near the bridge seems to excite the higher fundamentals where the ear would be more sensitive.

 

I "perceive" a note to be louder at the bridge than when playing half way between bridge and board, even when playing with a lot less bow action at the bridge than at the mid point.

[David Beard]

Many years ago, a well-know person in the field of violin acoustics told me that the greatest volume was produced not by digging in, but by using very fast and lighter bow strokes. That didn't sound right to me, so I tested it, and found that it was true for measured volume. But not for perceived volume or the intensity of the sound. Measured loudness and perceived loudness can be very different things.

 

An element of flowing the bow a bit more, whatever basic stroke, weight, or tone you're using, is perhaps part of loading more power into the lower partials of the notes.   This is perhaps then part of the recipe for a more open and carrying tone for a given volume.

 

However, from other things said, it's probably also necessary to put enough power into the 1k to 5k range to keep the sound articulate.

 

?????

[Johnmasters]

Perhaps the Timbre and the wave form (shape) make a big difference?

 

 

dlb

I think that is the issue. Timbre = wave form (shape). A Fourier breakdown of the waveform gives the overtone structure.

It seems as if the others are saying that a perceived loud (played back soft) really has more overtones. A sawtooth would have more overtones than a sine wave because more high frequency components are necessary to construct the "points."