Studio Insider #169 April, 2012
Tips for mastering and for maximizing audio level (making it really LOUD)
Before launching into this month’s topic, I must confess I’m still feeling the glow of WinterGrass, the annual eclectic music festival in Bellevue, WA, near Seattle. It’s a great infusion of festival vibe in the middle of winter, and the folks who run WinterGrass, along with the hosts at the Hyatt, are dedicated to making the festival a fun and engaging experience for all. I gave an afternoon workshop in recording acoustic instruments in the home studio, and also led a much more comprehensive day-long seminar with the same subject matter. We ran a ProTools audio recording session, and the participants and I recorded a bluegrass version of “Red River Valley” to demonstrate the various techniques we’d been discussing. One issue that came up at the WinterGrass workshop, and which appears frequently at my recording workshops elsewhere, is a tricky one that concerns volume maximizing.
Joe, have you used ABC Audio’s new ultrasuperspecialgainultramaximizer?
Perhaps you’ve heard the phrase “the volume wars,” or seen names like “ultramaximizer,” “ultrasuperspecialgainultramaximizer,” or others like that. The subject of the “volume wars” is the increasing competition among producers of digital audio to get their productions to sound louder than the rest of the world. I’ll give you a bit of background, and then explain what’s going on today in recording studios and radio studios. If you know what’s really happening with today’s volume wars, you can take steps to make your recordings sound better.
Many years ago, and well before the advent of digital audio recording, the Federal Communications Commission (FCC) was charged with keeping an eye and an ear on how broadcasters used their powerful transmitters. Part of their charge was to enforce fairness between competing stations and networks, so that one company’s broadcast wouldn’t exceed a particular maximum signal level and thereby infringe on other stations’ signals. In its efforts to enforce their charge, one factor that the FCC monitored was the “peak level” of the broadcasters’ signals. If they sent out a signal that went beyond the legally allowable level, then the FCC would catch them, fine them, and threaten to pull their license to use the airwaves if they didn’t stop. It was a big deal.
It didn’t take smart engineers very long to figure out that if they kept most of their signal level (its volume) very close to the allowable maximum, and virtually never let it drop much below that, then their station would seem louder than others. However, that process also made the audio sound really bad – in fact, we call that sound “AM radio,” since most AM radio stations do just that. Many FM stations took pride in broadcasting high quality recordings, and would resist the temptation to maximize its volume like their “lesser quality” AM brethren. But as soon as the ads came on (just like on TV), there would be a very noticeable jump in volume. That still happens today. Many times I’ve had a TV in a hotel room turned to a comfortable and non-disturbing level for the program I’m watching, only to have it jump hugely in volume for every commercial break. This is absurd. And it can disturb the guest in the next room!
So what were those darned engineers doing to the audio, anyway? Were they circumventing the FCC’s ironclad enforcement of maximum allowable volume? Is this at all like today’s volume wars with digital audio? The answer to the last question is a definite “Yes!” Let me explain.
Peaks and valleys
The loudest sounds in a recording are called “peaks.” Think of them like mountains, towering over the relative flat or slightly hilly areas below, which represent the average listening level. Sometimes a recording is very quiet (for example, in a whispered intro of soft female voice and acoustic guitar). Sometimes there are moments when it’s really loud, for example when a drum crashes, or when a quartet of singers hits the apex of the chorus on an emotional gospel tune.
In digital audio, we call the loudest note possible “0.” (Zero). Everything else is measured in decibels (dB) below that. For example, -5 dB (minus five decibels) means five decibels below the maximum of 0. The quietest that CD audio can portray is -96 dB, or 96 decibels below the max. So CDs and equivalents can give us a dynamic range (the distance between loudest and softest) of 96 dB. Compared to anything possible before digital audio, that is mind-boggling. Most of the music you heard on LPs exhibited a dynamic range of approximately 15 to 20 dB.
If you make a recording of crickets singing gently in the night, and want it to be heard well by home listeners, you’ll raise the level of the crickets on the digital recording so that their loudest note is somewhere close to that max of 0. But if half – way through your 3 minutes of cricket songs, the neighbor’s dog tips over your metal trash cans and rolls them down your concrete driveway, you will have several peaks (loud sounds) that are much, much louder than those crickets. And if you don’t modify your digital recording somehow to turn down that banging, here’s what happens: The peaks from the bangs of the garbage cans might be 30 dB louder than those mellifluous crickets. If you turn up the volume of your recording so that the garbage can peaks are at 0 (the max), then your cricket songs will be barely audible, since they are 30 dB lower.
So what can you do to make your cricket songs appear as loud as the rest of your friends’ CDs, even if your digital recording of them portrays them as 30 dB below those thundering garbage cans? You need a super-duper-ultra-extra-maximizer-ultralizer from the ABC plug-ins company. This plug-in examines your recording (now called your “file”) and then finds all the “peaks,” or loud parts. The software engineers at ABC are really bright, and they invented a plug-in that knows how to turn those garbage can bangs (the “peaks”) down and quite close to the surrounding volume level (the crickets), without making it sound like they’re modifying your recording. So the plug-in turns the garbage can bangs down from 0 (the max) to -25, and we can hardly tell they’ve been messed with. Now, remember that the crickets are at -30 dB, and the engineers at ABC made a great plug-in that doesn’t mess with the crickets. Since the peaks (the garbage cans) have just been lowered to 25 dB below max, we can turn up the entire recording, bringing the garbage can peaks to 0 and the crickets to -5 dB.
This process is very much like what “maximizers” do in digital audio. The only difference is that the peaks usually aren’t garbage cans (I know, some may disagree) and the average level of the rest of the file isn’t 25 dB lower. But what these plug-ins do is to find the peaks, lower them imperceptibly, and then bring up the entire file to a hotter level. The result is a louder recording.
So what’s wrong with all of that? Remember that when the software is turning down the peaks and bringing up the rest of the file, it is essentially raising the volume of the low-level sounds. These include breaths, tongue and lip sounds, air whistling through teeth, strings rattling on frets, a guitar pick hitting a string, a banjo fingerpick hitting the head, a mandolin pick hitting the fingerboard, a fiddle bow scraping a string, etc. To some degree, bringing up these sounds to make them audible adds a sense of intimacy to the recording. But if it’s done too much, the character of the voices and instruments is changed significantly, and might not reflect the sound that the musicians or the recording engineer intended. And if it’s done too much, the entire recording sounds like it’s coming from a tiny AM radio.
So when you master your next recording, or send it to a mastering engineer, have a short talk with her/him about level maximization, and compare the maximized version to another, untreated one.
More next month!
Joe Weed records acoustic music at his Highland Studios near Los Gatos, California. He has released six albums of his own, produced many projects for independent artists and labels, and does sound tracks for film, TV and museums. He recently produced “Pa’s Fiddle,” a collection of 19th-century American music played by “Pa” Charles Ingalls, father of Laura Ingalls Wilder, the author of the “Little House on the Prairie” book series. Reach Joe by calling (408) 353-3353, by email at [email protected], or by visiting joeweed.com.