Thursday, August 13, 2009

EQ (or, how to turn down the "suck" knob)

Warning: this blog post contains a science experiment.

So your sweet guitar tone sounded like crap when you got to the concert. Or maybe band practice was awesome... but things sounded really bad at the gig. Why does my stereo sound so bad when I turn it up really loud? Or down really quiet?

The answer: it's your ears.

No, I'm not saying "it's a personal problem." You are normal. But, your ears have a built-in equalizer... and the EQ settings are different depending on how loud the noise is. Scientists have studied this, and the result is called the "Equal Loudness Curve:"[1]

If you're like me, you're wondering, "huh?" Each line is an "equal loudness line." The bottom line is the threshold of hearing — stuff we can just barely hear. The top line is dangerously painful sound levels. In the middle (around 1000 Hz, about a high C) we hear sounds the easiest. In comparison, the bass sounds have to work harder and move the air more harshly. Likewise with some of the higher frequencies.

But even more interesting is that the curves are shaped differently depending on how loud it is. This means that if you change the volume, the balance between hi/lo/mid will change.

Ok, time for a science experiment. Get your MP3 player or iPod and sit down with a "large" song (like Steppenwolf's "Born to be Wild"), and do this:[2]
  1. Turn it down until you can barely hear it. This is the bottom line on the chart. What do you hear? You can hear mostly the cymbols. A little bit of guitar and vocals. Probably can't hear the bass guitar or any other drums.
  2. Turn it up a little so that you can hear it clearly, but still quiet (about 20%). What do you hear? You can probably hear the vocals, guitars, keyboards, cymbals, snare, some of the drums all clearly. Hardly any bass guitar... or tone.
  3. Now turn it up so that it sounds "full" (probably about 50%). You can probably hear everything clearly... but the stuff in the middle (like the cymbals) are starting to get annoyingly loud.
  4. Turn it up more... so that it's "a little loud." Chances are, this is the first time you heard a decent bass guitar sound. Notice how all the other volumes have changed dramatically... but not the bass?
  5. Turn it up really, really loud. Now, you probably can't stand it because the drums, guitar, and keyboards are overdriven.
So, as you changed the volume... you got a big response on the mid and hi frequencies... but the low frequencies were slow to respond. This is exactly what the chart above is showing us.

This has a couple implications::
  • Loudness is subjective. That's why the developed the A-weighting for decibel meters, because this reflects the way our ears work. In contrast, the C-weighting has a flat frequency response and is useful for scientific measurements.[3]
  • You have to change your EQ settings when you change the volume. You can't set the tone knobs in your room and then turn it up for the gig. It will sound different.
Here's another experiment to try:
  1. Get some kind of media player with an equalizer (5 or more bands, e.g. WinAmp or Amarok).
  2. Set all your EQ's to 0.
  3. Set the volume to about 50% and get a feel for how it sounds (even if it's not good).
  4. Now set the volume to be about half as loud. Change the EQ settings until it sounds about the same as it did at 50%. If you succeed, chances are that your EQ looks like the chart above. You had to push up the bass... but the rest didn't need much adjustment.
  5. Now set it even lower and try again. You probably had to turn down the mids and trebles to make it sound like the bass was louder. (Kind of like the curve above.)
  6. Now turn it up pretty loud (75%) and see what you have to do. For me, it ended up looking a little like a line that gently sloped downward toward the high frequencies.
What's the takeaway? For recording, this means that your mix will sound different at different volumes. If you're doing it yourself, you need to be careful that you're always mixing at about the same volume. In live music, people want to be moved by the lower frequencies: the bass, the kick, the rhythm guitar. When you turn it up, you will have to turn down the higher frequencies (lead guitar, keyboards, vocals, snare, cymbals) to keep them in balance.


[1] Source: http://en.wikipedia.org/wiki/Equal-loudness_contour
[2] If you can, set all your EQ settings to 0.
[3] See http://en.wikipedia.org/wiki/A-weighting
[4] See also the last post, which discussed loudness.

Crank it up, man!

Rock'n'roll is supposed to be loud. Is not just a childish attempt to rebel or a substitute for real art. There is something that sounds different when the music is loud. When the kick and the bass are physically shaking your body. There is something inspiring about how a guitar sounds when the amp is turned up to 11. It is not just a fad.

But sometimes you wonder, "is it too loud? Am I going to damage my hearing? How loud is too loud?"

The simple answer is that musicians are typically exposing themselves to levels that are OK for an audience... but too loud for the musicians. Let me explain...

There has been quite a bit of research put into how much volume our ears can take before it damages them. The standard benchmark is the OSHA guidelines:

PERMISSIBLE NOISE EXPOSURES [1]
______________________________________________________________
|
Duration per day, hours | Sound level dBA slow response
___________________________|_________________________________
|
8..........................| 90
6..........................| 92
4..........................| 95
3..........................| 97
2..........................| 100
1 1/2 .....................| 102
1..........................| 105
1/2 .......................| 110
1/4 or less................| 115
___________________________|________________________________

So, if you're going to be exposed to noise for 8 hours a day... you need to make sure the noise is less than 90 dBA to avoid damaging your hearing. On the other hand, if you want to crank it at a rock concert... you need to keep it below 95 dBA.[2]

That's great... but what the heck is a "dBA slow response?" First, here is a table with several relative loudnesses to help give you a feel:

LOUDNESS OF TYPICAL EVENTS [3]
______________________________________________
|
dBA | Equivalent event
________|_____________________________________
|
0 | threshold of hearing
10 | rustle of leaves, a quiet whisper
20 | average whisper
20-50 | quiet conversation
40-45 | hotel, theater between performances
50-65 | loud conversation
65-70 | traffic on a busy street
65-90 | train
75-80 | factory noise (light/medium work)
90 | heavy traffic
90-100 | thunder
110-140 | jet aircraft at takeoff
130 | threshold of pain
140-190 | space rocket on takeoff
________|_____________________________________

"Cool... but what's a dBA?" See below.[4]

So, for a music performance (1-4 hours), the audience can take 95-105 dBA without damaging their hearing.

But for the musician, regular exposure 95-105 dBA will result in long-term hearing loss.

This is a real problem, because:
  1. Musicians are taught to "practice to perform." That is, practice as if it was really the performance.
  2. Balance, levels, and dynamics have to be set at performance volumes. Things literally sound different when you change the volume.[5]
  3. Musicians love the way it sounds at performance volumes. Can't get enough.
Here's some tips for musicians to protect their hearing:
  1. Ear plugs. Not the 35 dB instrustrial kind. Get some light 5-10 dB earplugs.
  2. Turn it down for practice... especially if it is not on a performance day.
  3. Turn down your MP3 player. Typically 50% or less. See this.
Be a man! Use earplugs.

Next post: Why does it sound different when I change the volume?

[1] 29 CFR 1910.95(B)(2), Table G-16 (link)
[2] For any of my friends at New Hope... don't freak. We've been measuring dBC. With our mix, 105 dBC is probably around 90 dBA.
[3] I found this list here, which attributes the Encarta 2005 encyclopedia.
[4] To be more specific, a dBA is the decibel reading that you get from a decibel (dB) meter that is set to the A-weighting and set to a slow response. A decibel (dB) is measures the air pressure that is stirred up by the noises. The A-weighting applies the principles of the "equal loudness curve", and it considers the fact that our ears don't pick up very low and very high frequencies very well. (More on that in a future blog post.) Most decibel meters that you buy will have the A-weighting available.
[5] This, too, is the equal-loudness curve... see [4].

Monday, August 10, 2009

Over and Under

Properly coiling your audio cables is the most important thing you can do to care for them. Most of you probably already know how to do the "half-twist" when coiling the cable. However, recently my friend Jeff showed me an even better way to coil cables.

To coil with the standard half-twist method, you gently twist the cable while making each coil. This removes some of the twisting tension in the cable. That tension would normally make the cable want to twist up and tangle. However, since you removed the tension, it is as if you had coiled the cable on a spool. This is excellent practice and provides years of reliable use.

But, if you have done this for many years, you have noticed some problems with this. First, to keep the cable from coiling like a phone cord, you have to diligently unwind the cable -- putting back your half twist with every coil. Further, these coils do sometimes get tangled without much effort. The cable also takes on a permanent coil and does not lay perfectly flat or straight.

This is where the over-and-under method comes in. This method will literally allow you to throw a 100-ft cable and it will not get tangled:


[1]

What's cool is that it works for more than just cables: garden hoses, ropes, and... erm... that's all I can think of.

You can read more about it on Wikipedia.

[1] Video from Chris Babbie Location Sound