Posted by Rick Denney on April 16, 2003 at 17:28:07:
Based on the discussion several of us have been having about the effects of lacquer below, I conducted a stupid little experiment. (I'm saying it's stupid up front, so if you read through all of the below and you think it is stupid, I don't need to hear it: I already know it. Scientific refutation that would expose my flawed thinking, on the other hand, is welcome.)
It is said that all models are false, but some are useful. I hope this fits in that category.
I just searched around my office for a metal object and found a sheet-metal bookend. I hung it off the end of a pen and rapped it with a fingernail to ring it. It rang with a typical metallic clang. I recorded that clang in CoolEdit, and analyzed the sound. The sound showed significant resonance peaks at 16, 55, 59, 117, 155, 186, 230, 300, 444, 620, 790, 1080, 1300, 1600, 1850, 2400, 2530, 3000, and 3800 Hz. The highest peak was the one at 1080, which was about 25 dB higher than the lowest, which were those at the highest frequencies. Of course, none of these resonances are harmonic, which is why it's a clang and not a ring, but I could clearly hear many of these frequencies.
Then, I applied a coat of "lacquer", in the form of a complete covering of one or two layers of adhesive tape. I think we can all agree that adhesive tape will damp vibration to a much greater extent than lacquer. After all, I can peel the tape with my thumbnail, but lacquer is much harder and resists such treatment. Also, the tape is thicker, but I don't know by how much. Lacquer is probably the thickness of the adhesive alone and not the plastic that is stuck to it.
I then recorded a clang of the "lacquered" "bell". I found resonant peaks at 16, 55, 59, 155, 169, 186, 300, 444, 620, 790, 1080, 1300, 1600, 1850, and 2530 Hz.
With all noise, there is a stochastic component, and many of these peaks are fuzzy at best. But there is some ring in the clang.
Interestingly, some low frequencies were affected. 118 Hz was damped by about 10 dB, dropping it into the noise, as was its harmonic at about 230 Hz. 169 was present in the taped-up ring but not in the unfettered ring. The highest frequencies were damped by the taped, as one would expect. Because those highest frequencies were strong in the sound (as they are in nearly all percussion instruments), the articulation of the sound changed a bit. Also, the sound attenuated more rapidly.
So, a thick layer of scotch tape can rearrange the lower frequencies slightly and damp out the upper frequencies (above 2600 Hz) from the clang of a piece of steel. The difference in the hearable clang was noticeable but not profound.
To make our false model work, we have to think about why it is false. 1.) Scotch tape is much thicker and softer than lacquer, so it would damp a broader range of frequencies, both lower and higher, and to a greater extent. 2.) A piece of unbraced sheet metal hanging from a hard hook and rapped by something hard will clang much more than a well-braced tubular brass instrument that is sitting on a very gooey lap and hugged by very gooey arms. If I take my bookend and rest it on my leg (by its own weight), the clang is attenuated by at least ten times that of the tape. If I lay a finger on the side, the clang is attenuated by a factor of 100 or 1000 to a dull thud, because doing both braces both the extended plates of steel. Even if I rest it on a hard surface and brace it with a hard object, the clang is still attenuated to a dull thud.
Our next step in applying our false model is in trying to understand the role of the resonance of the metal in the first place.
By far, the bulk of the sound produced by a tuba comes from the vibration of the air within it. By bulk, I mean probably 99.999999%. I suspect that the ringing of the most resonant hunk of brass while being held by a gooey tuba player would be unheard at all and probably only felt by the player (we can feel vibrations up to frequencies of perhaps 400-600 Hz). Any effect the resonance of the brass has on the sound does so by its coloration of the overtones. Metal resonates at a variety of frequencies, as revealed in my stupid experiment. Some of those frequencies might be out of phase with some of the upper harmonics of the air vibration and damp them, or they might be in phase and reinforce them. These effects are small, and many players believe they are harmful in most cases. They certainly would not affect all notes equally, and what effects they did have on certain frequencies would be tiny.
It seems to me that our bodies damp the brass so much more profoundly than lacquer could, and that such damping itself has a small effect on the sound produced by the instrument, that the effects of lacquer, but largely above the tuba's acoustic range anyway, would be a tiny piece of a tiny piece of a tiny piece of our sound.
I don't know how to square this conclusion with the passionately held experiences of those who insist they hear a difference, but much as I would like to verify that experience (given that I have two unlacquered tubas), I don't see any evidence for it. I have to conclude, therefore, that it's in the same category as 5-gauge speaker wire, compliant wire-spoked bicycle wheels, and other mythical effects widely believed but defying any possible physical explanation.
(Trumpets are different. 1, they have musical content into high frequencies--probably as high as 5000-8000 Hz. My tape attenuate frequences higher than about 2500 Hz. 2, they are less braced and not held in a gooey lap with a large surface area contact. But even with trumpets, hard experiments show that the effects of lacquer are subtle at best.)
Rick "who wonders if some archeologist will someday wonder why a metal bookend has plastic tape on it for no apparent reason" Denney