Sound Experiment

[Don Herman rev2]

Derek,

Thanks, hope it helps! I can provide a few references if you need them (books at home and work and not at hand at the moment). The curves are in most acoustics texts, and reference books like my old Reference Data for Radio Engineers by ITT.


On other topics: Higher frequencies are generally attenuated more than lower ones, but because our hearing is more sensitive we don't really notice (in fact, the opposite is usual). A player with a "brighter", "brassier" sound will generally project (carry) and sound better, at least to us. Anybody at the demo (forgot the name, durn it -- I think TH's own Steve T may have been one of the players?) at the 2004 ITG Conference can attest to that.

I'm not sure I buy the constructive/destructive interference model as the explanation. I can (and have, many times) map out peaks and nulls to optimize microphone placement, but in fact the freqency span is so great and actual reverberations so complex in the real world that it's pretty much a meaningless exercise. It may help explain "sweet spots" in rooms but generally increasing the reverberation time (to a point -- look at C80 and similar metrics) helps everybody out by providing a large number of (acoustically) reflective points and making interference patterns all but intractable to analysis. If you've decent stereo, move forward and back and side to side a bit and notice how the sound changes. In a room designed for music, there will often be less change.

Aside; Places like multipurpose auditoriums and churches are notoriously difficult to design. The long reverb times and complex "live" reflection patterns that make music sound good are bad for speaking, and the short/minimal reverb that makes spoken voices clear, makes music sound dull and lifeless. Another of life's little compromises...

There are some nice (and pricey) computer programs for modeling room acoustics that do a decent job, but even they tend to break down if the room is moderately complex.

Onwards - Don
_________________
"After silence, that which best expresses the inexpressible, is music" - Aldous Huxley

[swthiel]

[iambrassman]

I'd like to see the results. Some of the other science geeks already weighed in, but consider this possibly surprising result lurking in the wings.

Yes, Kalijah is right, sound will reduce in direct correlation to the square of the distance from the source. Different pitches, however, decay at different rates. High frequency pitches decay very rapidly in air, whereas very low pitches can carry enormous distances. I'm not an accoustics type engineer, but my understanding is that is because of the lower pitches' longer wavelength.

I would assume that a player / horn combination that produces the greatest number of UNDERTONES will carry the best. The undertone series will dictate what percent of volume remains at a distance.

For once the overtones that make us sound brilliant, or sizzle-ey, won't do us as much good at 300+ feet away.
_________________
Jim "IAmBrassman" Utley
Callet Jazz 0.470" /
Greg Black made John Blount Personal Mpc

[Don Herman rev2]

If you're talkin' about me, I'm not sure if both Kalijah and I have been complimented, or insulted...

Attenuation is generally caused by surfaces the sound reacts with, not air (as a transmission medium). The longer wavelengths require "bigger" things to absorb, which makes them harder to attenuate. This is well-known in acoustic design; it's easy to dampen high frequency peaks in a room, but tough to deal with low ones since they may be around the wavelength of the room dimensions.

The difference in our hearing response at low and high (for trumpet) frequencies is HUGE and is (imo) a primary reason higher, not lower, tones (harmonic or not) "carry" further.

In case anybody cares, here are some wavelength numbers (using 1100 ft/second for velocity). No guarantee I got all the numbers right! The equation is wavelength = velocity / frequency. Also, note that a quarter wavelength is a "magic" number -- it will cause sound to cancel for a round-trip path as it comes back out of phase.

20 Hz: 55 feet (16.76 m)
100 Hz: 11 feet
440 Hz (A concert): 2.5 feet
1000 Hz (around concert high C for us, two octaves over middle C on a piano): 1.1 feet
3000 Hz: 0.367 feet (4.4 inches)
10,000 Hz: 0.110 feet (1.32 inches)

Hopefully this helps explain why high frequencies tend to be more attenuated in the real world.

HTH - Don
_________________
"After silence, that which best expresses the inexpressible, is music" - Aldous Huxley

[Derek Reaban]

Don,

I did a little research based on your recommendation. I looked up the Fletcher-Munson Loudness Curve. Very interesting stuff!

Based on the Sprectral Analysis that was presented in the "Visualizing Resonance" topic, I decided to use that diagram as a starting point. I penciled in my impression of what I learned from the F-M Loudness Curve on my diagram and showed this "high point" at 30-50 feet from the player on stage. This high point represents the sensitivity of human hearing between 2-5 kHz.



In my digram, Point A is the sound just as it comes out the bell. The fundamental of the sound is the most present here. As the sound decays over distance, all of the frequencies decay equally. However, the human ear perceives frequencies between 2-5 kHz MUCH better than the fundamental.

So, Point B shows that the overtones that exist in a players sound (say between 2-3 kHz) will be heard more prominently by a listener. If a player has a mature sound, with lots of higher frequencies present in the sound, this sound is what will "project" to the audience.

This would clearly explain why a player with a "less complex" sound would be heard clearly in a small radius around the player (mostly the lower frequency fundamental), but since the higher frequencies are "less strong", the sound is not perceived well at a distance. The player with the "complex" sound, can play the fundamental at a softer perceived dynamic in a small radius around the player, and even though this sound decays more quickly than the other player, the energy in the higher overtones gets picked up at a distance.

Comments?

I'd love to hear them, because this is just my impression so far, and I haven't yet down my experiment (had to cancel my lesson today due to last minute birthday preparation for my son's big day tomorrow).
_________________
Derek Reaban
Tempe, Arizona
Tempe Symphonic Wind Ensemble / Symphony of the Southwest

[PowerSpectrum]

that makes a lot more sense

[PowerSpectrum]

I was a bit reluctant to do it since I was concerned by the result.
But I did it:

High D. World class soloist during Haydn concerto cadenza...


High Bb. Myself during my practice...


There are no words....

[jpetrocelli]

What software is that?
_________________
Joe Petrocelli
www.joepetrocelli.com

[Don Herman rev2]

I think it's becoming clear that more overtones lead to better sound with more carrying power... Of course, we need to be able to control the overtones to color our sound for the music.

The ITG link Steve posted goes to a seminar which proved that pretty decisively for those in the audience. Probably the biggest thing we have to face as players is to realize that the sound in front of the bell is very different from that behind. A horn which sounds bright to us may not to the audience, and vice-versa.

And of course, the world is filled with band directors and conductors who think a trumpet should sound like a low-octave flute...

PowerSpectrum: what recording setup did you use? It's possible (likely) that your mic rolled off the high end pretty substantially. You are likely nowhere near as low in overtones as that plot shows! For those sort of tests, I use a (pricey, about $1600) Earthworks measurement grade (calibrated) mic and of course you need a great preamp and recorder as well. My mic is likely better than my current recording setup; I use it with a special preamp into a big HP (Agilent, now) audio spectrum analyzer. Can't do it too often -- people at work get nervous when I start hauling a $50k piece of equipment around!

Derek -- I need to digest a bit. The work you do, and the time and thought you put into it, is simply amazing! Sometime I hope to have time to do and contribute more; you set a high standard! For now, do you have Mathcad or Matlab available? You could make some nice plots and play around with things a bit. At any rate, I think you are on the right track -- that matches my experience and the basic theory (the complicated theory is, well, complicated! - and I don't like multidimensional wave integrals any more, not that I ever did...) Attenuation and other factors aside, I think the basic idea that higher overtones (for trumpets) lead to greater (perceived) power due to the ear's much greater sensitvity to them is, err, ahh, "sound"! - Don
_________________
"After silence, that which best expresses the inexpressible, is music" - Aldous Huxley

[_dcstep]

[PowerSpectrum]

[irishrevolution32]

I don't know a whole lot about this matter, but here:
http://www.yamaha.com/artistmodelseries/index2.html
if you go to the 'video' tab in the lower left hand corner and watch "David Krauss - Your Bigest Asset is you Sound" He talkes alot about how one opera singer can make his or her voice carry all the way to the back of the hall w/o a mic. and over an orchestra, and he puts it all to the effects of resonance. I don't know if this is completly off base, but I hope it can help in some way.
_________________
~defy gravity.

[trptStudent]

[iambrassman]

[Derek Reaban]

I now have a cord to transfer information from my mini disc to my computer. I spent a little time last night trying to figure out how to bring the information into the Audacity program, and after a little fiddling, I was able to bring in several sample recordings!

I feel much better about this important aspect of this project now that I'm able to get the information uploaded onto the computer.

Whew!

2 more days until I gather some data!
_________________
Derek Reaban
Tempe, Arizona
Tempe Symphonic Wind Ensemble / Symphony of the Southwest

[mcamilleri]

Well, I don't pack a slider rule but I own one and know how to use it! It is even a circular slide rule (if that makes any sense).

I remember one of my college physics professors talking about the good old days before computers and high precision desktop calculators how they built an 8 foot long slide rule so that they could do highly accurate calculations without having to use the massive high accuracy log tables of the time (which were the size of an encyclopedia). Slide rules are quick but the handheld ones are only about as accurate as a 3-4 digit calculator.

There was a time when you could earn a masters degree in physics by inverting a large matrix or doing a single fourier transform, as they had to be done by hand calculation. Now you can do all that work on a PC in a few seconds.

Good luck with your experiment. Somehow I think you are not going to find anything that is obviously different in the sounds. The subtleties of pitch stability and the Q-factor of each harmonic (width and height) may be very important, and that is not so easy to measure with a spectrum analyser.

I will be very interested in your results. Please do you best to fully document each sample you do - the details of player, horn, mouthpiece, setup etc are important.

Michael
_________________
Alpha Angles
Besson Loyalist

[Derek Reaban]

Well, I managed to make the sound experiment project happen this weekend. I’ll have to say, this is one of the biggest projects that I have undertaken in quite a while! It’s been really nice to have the Olympics on while I’ve been doing this “analysis” work. I would have been watching them anyway, and instead of sitting on the couch, I’m tapping away on the computer with the TV on. Not too bad!

Friday I ended up taking a vacation day because my Wife’s Parents were able to come up from Tucson and they had volunteered to help us with our citrus trees. This is the time of year that we have to get all of the fruit off them and do our big pruning project (before they bloom). We have 8 trees (2 grapefruit, 2, lemon, and 4 orange). We ended up getting about 40 bags of fruit, and lined the street with citrus branches! Whew, what a way to spend my vacation.

It took us about 5 and a half-hours to finish and we were done by about 3 PM. I had plenty of time to get cleaned up, gather everything that I needed for the project, and get a short warmup before I drove to Benny’s house for my 4:30 PM lesson.

After arriving at Benny’s house, he had to find his microphone and charger cord for his mini disc player. While he did this, I played a few notes and then got my pencil and tape measure out to make some measurements of the room:

Here’s a picture of the room set up:


_________________
Derek Reaban
Tempe, Arizona
Tempe Symphonic Wind Ensemble / Symphony of the Southwest

[Derek Reaban]

After some technical difficulties in figuring out how to record on Benny’s MD player, we were just about ready to start. I spent some time to give Benny an overview of the idea of my “sound experiment”. I showed him my picture and told him my premise about what the microphone would be “hearing” at our ears versus the microphone at the other end of the room.

Then I gave him an overview of what we were going to be playing. I’ve been thinking about this for the past week, and I decided that the best way to be consistent would be to play some standard exercises from the James Thompson Buzzing Basics (BB) book (#5, #11, and #12) while listening to the recorded tracks on my iPod and earphones. I also had scheduled a standard orchestral excerpt (opening to Pictures at an Exhibition).

The one last detail before we started was to write down some details about the project:

Horns and Mouthpieces used for Buzzing Basics exercises:

Benny:

Mount Vernon Model 37 Bb Stradivarius (S/N 26,XXX)
Park-Hagstrom 24 Throat, Schmidt Backbore

Derek:

Elkart Model 37 Bb Stradivarius (S/N 154,XXX)
Monette B1-1 (I think with an 18 throat)

For the excerpt (since my Monette is in the shop for the next 6 weeks), we used his new Yamaha Chicago Model C trumpet. What a great instrument!

Last details: It was 64F for the high temperature, it was cloudy all day, we started recording about 5:10 PM, and the lesson was in Tempe, Arizona.
I played first. The close microphone was just to my right side (a little more than an arms length away) and the microphone across the room was exactly 26 feet from my bell. I played very well and feel that this is a very representative snapshot of my playing. Benny sounded great as usual, but this is the first time that he had read through the Thompson exercises while hooked up to an iPod. He was a VERY good sport!

A brief description of the exercises is probably worthwhile for those of you that aren’t familiar with the book. BB#5 begins on a low C half note, slurs up to a second line G half note, and then up to a whole note 3rd space C, and then 4 beats of rest. This pattern continues down chromatically to Low F. The dynamic begins at p on the low C and crescendo to a mf on the 3rd space C.

BB#11 is in eighth notes with the last note being 3 and a half counts long. I’ll try to include a sound clip here if I can figure out how the putfile.com site works. The pattern is all open harmonics, starts on 2nd line G, goes up to 4th space E, down to low C, and then up to the top of the staff G. It continues down chromatically to a starting note on low C.

BB#12 is in eighth notes as well, with the last note being a half note. All open, low C, 2nd line G, C-G up an octave 3 times, and then down slurring down to a low G and ending on a low C. Then continues up by half steps to a starting note on a 3rd space Bb.
_________________
Derek Reaban
Tempe, Arizona
Tempe Symphonic Wind Ensemble / Symphony of the Southwest

[Derek Reaban]

With that description out of the way, I looked through all of the data and decided that I would show a picture of what the High C (from BB#11) looked like. This considers a picture of the note at the “close mic” and the “far mic” for me and Benny.

Click here to listen to Derek playing High C (close microphone)




Click here to listen to Derek playing High C (far microphone)





Benny playing High C (close microphone):




Benny playing High C (far microphone):





While those pictures are very interesting (especially if you’ve had a chance to play with the Audacity program), a little analysis is in order to really understand what you’re seeing. Once I had the Frequency Analysis, I exported the data to a text file. This included the frequencies with the corresponding dB values. I quickly discovered that the frequencies were always the same in the text files, so I had to pick out the frequency of the fundamental pitch (like A440 Hz), and the figure out what each of the overtone frequencies were for the notes. You will understand this when you look at the next picture:




You’ll notice right away that I was playing my C at 931 Hz and Benny was at 941 Hz. 931 Hz was not a value that was specifically called out in the exported text file, so I just pulled in the frequencies that were closest to 931 and 941 Hz. I show the Decibel levels at the close mic and the far mic for both me and Benny. My fundamental was quite a bit louder than Benny’s (we didn’t have a way to standardize our dynamic – which didn’t really matter). You will notice that I was a little louder at the close mic but quite a bit louder at the far mic. In the Freq Diff (Frequency Difference) column, I show Benny’s sound magnitude minus my sound magnitude. This simple means that when there is a minus in the Freq Diff column I was louder than Benny. When the value in this column is positive, he was louder than me.

To make it very simple, I color-coded the results. When you see BLUE, I was louder for that specific frequency range and when you see YELLOW Benny was louder. It is very clear from this analysis that Benny is ringing all of the overtones above the High C MUCH stronger than I am. Does this surprise anyone?

Even though I don’t show the individual Frequency Analysis snapshots (you wouldn’t believe how long it takes to compile all of this data!!!!), I want to show the spreadsheet analysis that I did for several other notes for comparison:






I pride myself for my sound, and I’ve worked very hard to cultivate a vibrant sound model in my daily practice sessions. This snapshot shows that compared with a fantastic symphony player, I occasionally can get a note off that is REALLY GOOD. If we had a college or high school player in the mix for comparison, I think the difference in vibrant, resonant tone quality would be MUCH more apparent.

I’m still amazed that a note which can sound as loud at my ears or louder than another player with great carrying power (as confirmed by the microphone) but fail to get out to the hall. I’ve known this now for quite a few years, and this experiment simply verifies my sound impression from behind the bell.
_________________
Derek Reaban
Tempe, Arizona
Tempe Symphonic Wind Ensemble / Symphony of the Southwest

[Derek Reaban]

We chose to play Pictures at an Exhibition on Benny’s Yamaha Chicago. This is where Benny shined! It’s no surprise to me that his Pictures would look different than mine on the Spectral Analysis. But, HOLY COW! His plot was like artwork for every note! The notes were all so well defined, and look at the “size” of his sound compared to mine on the Far Microphone!


Click here to listen to Derek playing Pictures (1st 10 Seconds) (close microphone)




Click here to listen to Derek playing Pictures (1st 10 Seconds) (far microphone)




Benny playing Pictures (2nd D) (close microphone):





Benny playing Pictures (2nd D) (far microphone):




And here is my analysis of that 4th line D:




Aaaahhhh. It’s 1:40 AM. I’ll finish this tomorrow and try to get those sound clips figured out. Then you can get a feel where I fall between Hack and Herseth on the Pictures excerpt.
_________________
Derek Reaban
Tempe, Arizona
Tempe Symphonic Wind Ensemble / Symphony of the Southwest