Don’t Fear Your Meters: How to Pick the Right Reference Levels for Your Mixes

View Single Page

Image by Nowy szeryf. Licensed under Creative CommonsIf there’s a topic that will get rabid audio fans to their feet, it has to be metering and reference levels, am I right?

I will acknowledge that it’s not as glamorous or sensational a subject as some, but there’s a reason that meters are still with us after all these years.

In the most basic sense, meters are there because we can’t physically see or count the electrons traveling through the console or compressor or mic pre, so we can’t know how far above or below the optimal signal level we are with any certainty.

But despite their ubiquity, I feel like the true usefulness of a good audio meter is often overlooked. And not all meters were created equal.

I would wager that most musicians, producers, mixers and engineers don’t think much about their metering, and about the correlation between what a meter is trying to tell them and the reference level at which they’re working.

I also suspect that I could safely place another wager that most of them really don’t care what reference level they’re working at, which is fair.

But just because most people don’t think much about their metering or reference levels doesn’t negate the fact that there is much to gain by thinking just a little bit more about yours.

Why Bother Talking About Meters, Anyway?

After your ears, your meters can tell you more about your mix than any other tool. And different types of meters can help lead you to different decisions.

When I started developing my skills in audio way back in 1987
VU (Volume Unit) meters were standard on tape machines and on the main outs of the console.

Back then, the LED ladders and bar graph meters, which are now so common today, were usually limited to “newer” compressors and effects devices.

The same was true in the live sound world: The input meters were often LED ladders, or in simpler arrangements, a “signal” and “peak” indicator which let you know that each channel was passing signal and was not clipping at the input stage. The trusty console output VUs, however, were there to tell you what was really happening, as LEDs weren’t (and often still aren’t) a very good indicator of overall volume or RMS levels.

As consoles got smaller and cheaper, the need for compact metering became essential. Soon, only the most expensive, large-format consoles would include a pair of VUs on the main stereo buss.

The first time I ran a small show on a 16-channel mixer with nothing but LED metering, I found that it was not easy to know exactly how hard I was really driving the system on average, aside from listening for how loud it was, of course. Having grown accustomed to using VU meters, I was a bit out of sorts.

In my years of doing live sound, I learned early on that knowing a sound system’s optimal level and absolute limitations are key to getting great results out of the PA.

If a sound system has been properly aligned and gain-staged, you can keep your average levels around 0VU and you will usually be in good shape throughout the entire signal path. By doing this, you could generally ensure that you would have minimal noise and more-than-adequate headroom to keep things clean—even with large transient changes.

Could I just use my ears? Sure, why not. Ears are a remarkable at discerning tiny changes and differences in a way that meters don’t always bear out. But our ears also tend to prefer “louder” and “brighter” (especially when directly comparing options in the short term) so they are not the most objective instruments. Meters, on the other hand, are completely objective and absolute.

If you use meters deliberately to help make better decisions about how to make things sound good then you’re much more likely to consistently deliver excellent results.

In the end, using meters isn’t about trying to stay strictly within the boundaries. Rather, it’s about clearly defining where those boundaries are so you can comfortably brush up against them.

Gain Structure in Consoles

The best way to understand how metering works in the digital domain is to first understand how it works in the analog world.

Recording consoles—and all other analog audio devices for that matter—are designed so that when you use them properly, you can maximize the signal-to-noise ratio while keeping distortion low or non-existent.

Console manufacturers try their best to provide a roadmap by clearly labeling the faders and pots, and calibrating the meters to specific standards.

All audio circuits have a “noise floor”—the operating noise of the electronics even when no signal is passing through them— that serves as the lower boundary for their dynamic range. They also have a “maximum output level”—which is often qualified with a distortion percentage, expressed as a percentage of “Total Harmonic Distortion (or THD) within a certain set of parameters.

Faders generally have a “unity gain” marking to give you a starting point for a kind of operational sweet spot. By setting the fader at unity gain, you can then dial in the amount of input gain needed for a particular source. When setting the input gain, if I set my average levels to be around 0VU, then unless an extreme peak comes through, I should have a mostly clean, robust, low-noise, distortion-free track.

On a professional analog console or piece of gear, a signal at 0VU normally leaves about 20dB of headroom before audible clipping. It should also leave the noise floor far below—usually between -72dB and -90dB, depending on the quality of the piece of gear, the number of channels engaged and so on.

20dB above 0VU is actually a ton of headroom for a single input. But when you consider this headroom across a full mix and the transient nature of most music, that 20dB can quickly disappear, or at least seem closer than it appears in your rear-view mirror.

To add to the precariousness of balancing and policing levels in a dense mix, an analog console’s summing amps are located before the master fader. If you happen to be slamming the output meters and think you can remedy the problem by bringing the master fader down? Sorry, you’re out of luck. Anything that exceeds the maximum output level is likely clipping the summing amps as well. The only way to fix this problem is to bring all of the faders down a proportionate amount.

This differs from Pro Tools today, where bringing a master fader down actually lowers the operating level of the corresponding bus, thus eliminating the headroom problem. Since modern DAWs have an almost unlimited amount of effective internal headroom—especially in today’s world of 32-bit floating point processing—the concept of “running out of headroom” in a digital mix may seem peculiar or foreign to newer mixers, but I assure you it exists!

Tape is the Same…But Different

The behavior of an analog tape machine is similar to any other analog circuit, except at the extremes.

The noise floor of good, professional, virgin tape is around -68dB, though print-through can make the practical figure more like -56dB (as per ATR Tape).

At the top end of the scale however, the output level at 3% THD (Total Harmonic Distortion) is +12dB.

This means you have about half as much headroom above 0VU with tape as you do on a good analog console. That would seem like a problem on its face…because why would you want to lose headroom? Dynamics are important in music, right?

But tape, being the sneaky bugger that is, is not a linear medium, unlike so many audio circuits. Tape compression, which occurs as you exceed 0VU is very gentle and pleasant sounding. In fact, analog tape is so good at setting compression thresholds and ratios that we hardly notice it. This is because tape responds gradually and dynamically to changing levels in a sonically pleasing way.

Pages: 1 2 3 4Next Page ❯View Single Page