Noise, part 2B. Or not 2B. Imagine for a moment that you are in a library reading room. Prominently displayed on the wall are signs calling for silence. After a few minutes, your ears adjust to the sound. Through the window, you hear a steady, complex sound of traffic. On the wall, a clock ticks softly. You hear people moving, turning pages, and talking softly under the "silence" sign. You hear the click of keys on a laptop, and softer but more frequent sound of cell phones, tablets and phablets. A reader unwraps a piece of candy which creates an astonishingly loud sound. In the corner, a radiator emits a steady hiss. Overhead, fluorescent lights, one of the great uglies of civilization, pretend to burn. If you turned all of these off, you would then hear transformers, or "wall warts" plugged in to the wall to charge and power gizmos, all emitting a steady hum. All of these make noise, and you can measure the noise in terms of loudness, say 12 dB, 30 dB. But all of these sound different, and some are much more annoying than others. And most of these sounds are like microphones. A microphone is a noise producer. It is like the radiator or the wall wart, emitting a steady sound. Or perhaps it has a crackly sound. But whatever sound it has, you don't want it on your recording. A microphone records a sound, and then sends it to a preamp. And this preamp again has its own sound. Plug a mic into a preamp in a silent room, turn up the gain, and, voila, hiss, hum, noise, maybe some turbulence or crackle. Just like the radiator in the corner, it hisses away. But it is always there. And on lute recordings, it often is seriously, way too much there. And, obviously, that's because the lute is soft. Same with the clavichord. And the lute and the clavichord are two of the most difficult instruments to record. If you put a tuba next to the radiator, you won't hear the radiator. You will just hear nice, warm tuba. But the hiss is there, just in a smaller proportion to the overall signal. In the old days, in order to get rid of all the noise in the signal chain, one had to buy very, very expensive recording equipment and mics to basically turn off the radiator part of the sound. Now, the freeway sound, the sound of clothing, hum in the room, and the skritch skritch of the lute, that is all "signal" the mic does not know whether it is you playing the lute, or your audience member slowly, painfully unwrapping a cough drop.
In the previous post we talked about how a lot of the noise comes from the player. And also that if the string is struck properly, you will get twice as much sound, more or less, and so your signal to noise ratio is much better. NB: In other words, a well struck pair of strings reduces noise better than the most expensive microphone, because the mic will amplify everything in the room. When you buy a microphone, it has a number for the amount of noise it makes. And that figure is usually between 10dB and 20dB, A weighted. A weighted means "the way we hear" weighted. If it does not say A weighted, it is a cheapo mic with fake specs. Lower is better, so 10 is way better than 20, A weighted. Unless, and this is a big unless, the noise is annoying. Remember the wrapper? If the self noise is irregular, it will be intrusive. If it is smooth, it will more or less disappear. Also, if it has a high, biting insect sound it will be annoying, even at the same volume. So that means, when you bring that mic into the room, or recording studio, you are bringing in let's say a tiny teapot that makes a soft, steamy noise. A noise that is always there. So if you play a really soft note on the lute, you will hear half teapot, half lute. For lute, ideally, you want that number to be 10-12dB, or lower. With exceptions noted below. What does 20dB sound like? Well, it depends on how close you are, but think average computer fan. But inside the mic, you can't walk away and have the noise go down. The noise in the mic follows the mic. The other "gear" noise is from the box that powers and records the microphone, sometimes this is the recorder itself like the Fostex FR2LE; sometimes you buy a separate box that just powers the mic. And, yes, this box makes some noise of its own as well. Unlike the mic which is always sputtering or hissing away in the corner, the amplifier may or may not make more noise at certain volume levels. There may, or may not, be a sweet spot where you get more amplification without cranking up the noise as well. So this box has a noise number as well. Here again, lower is better, but these numbers are really confusing, depending on the maker. The theoretical limit is around -129.5dB ein. What is "ein" besides the word for one? Well, it stands for equivalent input noise. So it isn't really the noise, it is kind of a guess at the noise. Basically, as far as this number goes, you should take it with a grain of salt, a lot of cheap gear claims to have really low noise specs, when it doesn't. But if is sports a figure of -126dB, well, that's probably not good enough for lute. Tuba, no problem. So since it isn't the good old days, we can solve the gear end of the noise problem very easily and for not too much money. We can buy the Fostex FR2LE, and a pair of Sennheiser microphones, either the MKH20 (omni pattern) or MKH40 (cardioid pattern) and problem is solved. You have two quiet mics, a quiet audio recorder. So this is very cool. You can also buy the Tascam DR680 and have a multitrack recorder for using more than two microphones. You can also buy a Grace Audio, or Audio Upgrades mic preamp, and it will be just as quiet, maybe a tiny bit smoother. It used to be there was more of a difference. As it turns out, these Sennheiser mics are not only quiet in the "specs", but they actually sound quiet. And the sound good on the lute. After all, what does it matter how quiet the recording is if the lute does not sound good? There are also some mics, like the DPA 4006, than happen to sound really great on lute, and even though on paper they are a lot noisier, the noise does not seem as loud. But if you want it really quiet, buy a quiet mic. And one of the reasons for this is that certain mics really amplify finger noise. That's because they are designed for pop music, where the goal is to "punch through" a mix. If I hear a lute recording with a lot of finger noise, I find it annoying, but this is also a matter of personal preference. The Schoeps MK2 and MK2H are also very good for lute. But if you want it quiet, and you want it to sound good, it is hard to compete with Sennheiser. Newer models of Sennheiser? Well, the MKH800 is a good mic, but very expensive and not as good for lute as the 20 and the 40. Newer models, like the 8040, to me are not as good for lute, and are also more cheaply made. Others seem to like these. Newer mics are now sometimes mass produced. Budget mics that are quiet and do not amplify lute noise? The Studio Projects B1, as well as the Oktava 012 are decent. And I must note here that people often ask what mic to buy, and I go through the whole deal, and then quite often they say well, what about the "Fresno Bull Wrecker" or some other famous brand. And I patiently explain that the person at the store, who doesn't record much classical music, will insist that it sounds like the Sennheiser, but at a fraction of the price, and they may not even make a commission on the Bull Wrecker, but it won't sound good. And I know in my heart of hearts that's the one they will buy. Or something similar. And so it goes. And the Zoom, the potato chip of the recording industry? Not that quiet, but not that noisy either. They usually are forgiving on the lute related noise, which is one reason that they are popular. Lastly, you will notice that mics come in all sizes. And if you look inside, you will see a little round disc of thin material that is the mic diaphragm. It will shimmer if you speak; it captures the sound. Now, the bigger the disc, the more sound can be picked upawith limitsaand so some of these really big mics have very low noise figures. But it also turns out that these really large designs color the sound, and also do not capture the fast moving harmonics of the lute very well. So the smallerathink range from cigar to magic markeramics are better for lute. Some people swear by the big ones, and, if you are just interested in the background noise of the mic, or you want a certain color for lute songs, they can be an acceptable choice, but you may find the coloration of the mic a bummer in the long run. Really tiny mic capsules, like the Earthworks mics, can be very accurate, and resist the sort of electric, bitter sound of larger designs, but they produce too much noise unless you put them really close to the lute. DPA mics have special cones that you can use to place them close to the lute. So here is the 64 dollar question, if these mics and preamps are readily available, and not prohibitively expensive, how come there are so many noisy lute recordings? And the answer is, I have no idea. But background noise, and lute noise are certainly part of the answer, and we will address these in the next part, "My way and the highway" -- To get on or off this list see list information at http://www.cs.dartmouth.edu/~wbc/lute-admin/index.html
