"V" is for Volume

How loud is loud enough? How soft is too soft?

If a high-end manufacturer such as, let's say, McIntosh (no shade on Mac; I just needed a name) produced a preamplifier with always-on tone controls that lacked any markings, center-detent, tone-defeat button or other zeroing facility, and instructed listeners to simply "set them where it sounds right," the audiophile outcry would echo from here to Zanzibar and back again.

But in effect that's precisely what Mac — and every single other manufacturer of preamps, receivers, and integrated amplifiers — does with a far more important control: the one with a "V" on it.

If queried, most audio types, this one included, will tell you, "I like to play back music at its original level, for the most realistic experience." But how loud is that? Even if we knew the "original level" perceived at — for the sake of argument — row 23 in Carnegie Hall on Beethoven's 9th under Fürtwangler on November 18th 1951, the number of variables interposed between the microphones in the hall and the loudspeakers in your listening room make nailing that exact level unlikely: the biggest one being that the recordings we receive have no reliable reference point, and certainly no reference tones from which we can scale dynamics in our individual rooms and individual systems, with their individual amplifier-gain-structures, loudspeaker-sensitivities, and acoustics.

It was even worse in the analog age, when the "0 dB" reference point might be just about anywhere. In the case of acoustical (classical, jazz, folk, etc.), music with an effective dynamic range of perhaps 85 dB had to be squished onto a medium, the LP record, capable of, optimistically, real-world dynamic range of perhaps 55 dB on its best day. Worse, the set-point of the requisite compression, that is, the balancing point between the turned-upness of softer passages and the turned-downness of louder ones, was an artistic decision made (nearly always) by the mastering and cutting engineers, and so might be pretty much anywhere. And it varied, a lot, with frequency, since the LP medium was challenged at higher frequencies, while vinyl's lowest ones had to be compressed even further, often well beyond the RIAA curve specs, to keep groove-width from eating up playing time. So even if a reference tone had been included on every LP, which most certainly was not the case, there'd still have been a net variance of perhaps ±10 dB in subjective "realistic loudness" from record to record.

The compact disc and subsequent digital media have given us 20 dB, at least (a sheet-ton), of additional useful dynamic range, and a reliable fixed reference point: 0 dBFS — "full-scale," or the maximum level the digital medium can accommodate without the train-wreck of digital clipping, so that a 0 dB tone from one disc on one CD player, connected digitally, should produce exactly the same analog level, whether from line or speaker outputs, as any other disc and any other player. (In practice, engineers always cheat in a dB or two or three of headroom, so even here the 0 dB point is not reliably fixed, but the range of variance is a lot smaller.

In the filmsound sphere, of course, we now have a fighting chance thanks to the legacy of the original THX, and then Home THX programs instituted by then-Lucasfilm stalwart (and now Apple scientist) Tom Holman. One of the first cornerstones of THX to be laid was its mandate to standardize reference levels in THX theaters, so that an audience — at least, those in the sweet spot around, I suppose, row 15-of-25 — would hear the same levels as did those who mixed the movie in the first place. This was incorporated (with a reduction for domestic speaker-proximity) in the Home program. With Home THX, for the first time, the loop got "closed," since with digital media (well, in truth, analog — Laserdisc— to begin with) — first DVD and now Bluray — a THX-certified preamp or receiver's setup noise could now be calibrated to produce a certain level (75 dBC-weighted SPL) at a certain volume setting, usually shown as "zero," "Ref," or "75." This was accomplished, first, manually, using a thumb on remote-control up/down keys and a handheld SPL meter (we're gonna miss you, Radio Shack!), and in the modern era automatically by the robotic room/speaker-EQ/setup-adjustor built into nearly every AV receiver. In either case this took amplifier- and speaker-sensitivities, and listener position, out of the variables list. Though largely no longer so labeled, "THX reference" has become a de facto standard: I've tested scores of AVRs over the past decade or so, and with a few exceptions, whether THX-certified or otherwise, they get main-channel levels, both relative and absolute, right within a dB or so. As a result, if you sit in the calibrated seat and dial the master volume up to the cal point, you have a very good chance of experiencing filmsound reasonably close to the way it was intended.

And that's important: not only do effects and ambience depend mightily on absolute volume at the ear to "work" as the sound-designers intended, but both the intelligibility and nuance of dialog, as well as timbral coloring and dynamic impact of music, are just as keyed to a "correct" absolute volume level.

Unfortunately, this reference does not necessarily translate to music. For all the reasons already mentioned, there's no reliable equivalent reference level for music recordings, and every one is different. If you play the latest R&B sensation's hit — likely "louded-up" with compression to an effective dynamic range of perhaps 6 dB — you may well get club-like levels with your master-vol well below reference level. But cue up a DG disc of Webern's op. 10 "Five Pieces for Orchestra," and you may have to dial up 6 dB beyond reference just to hear the damned thing.

Which is right? We just don't know. When you take a seat at Carnegie Hall or the AMC Burbank 16, on there's no volume control on the seatback in front of you: loud is loud and soft is soft, and you don't give it a second thought.

I'm not suggesting we should ditch the preamp and plug our source components directly into the power amp (though it's a thought!). But we have the technology to provide a volume reference for music recordings. It's about time the RIAA, or somebody, stepped up and started doing so.

COMMENTS
eugovector's picture

You start this article by pointing fingers at the hardware manufacturers, but isn't this clearly a problem in the production process? Until the source is produced at a known reference level, everything downstream is working with unknowns.

supamark's picture

Pro studios/engineers always print tones on the tape (1/2" 2 track analog tape is still quite common, and the tones are required for best playback). Everything in a professional studio is done to a reference level (usually +4 dBu at 1 kHz).

In reality, very few people have audio systems capable of realistic dynamics and volume levels. Most preamps aren't even capable of realistic dynamics (let alone the amplifiers), and few non-professional speakers can really ouput realistic volume levels.

Slardybardfast's picture

Thank for the excellent article on the lack of any way to ascertain the intended volume level for playback listening. Buried in the 4th from last paragraph is a very important little gem: the 2 words “timbral coloring”. I believe, for music listening, we use the volume control to get the proper timbral coloring from our playback systems.
Part of what we are doing when we raise the volume level to our preferred level is too bring the bass into auditability. This is what I think Daniel Kumin means in that paragraph. To put it more directly, the volume control is also a “timbral coloring” control. I believe more people use the volume control this way than those that use it to bring the SPLs up to “concert” levels.

X
-->