Viewpoints: Ears vs. Measurement (Part 1)
Welcome to New York: A Full Frequency Response to a Measured World

By Jack Alexander

Editors Note: We often insert the phrase “Performance Audio”. Jack is prefers the simpler term “sound”. Fortunately, the editors get the last read.

This is our AES issue, so it seemed appropriate to tweak those guys a bit. Originally, Doug suggested we discuss cabling (as in, all cables don’t sound the same), which would have drawn adequate blood for our purpose. I’m going to expand on that a bit, using the cabling and related discussions as a launching point.

SUBJECTIVE DIFFERENCES

Many among us accept subjective (audible, and unmeasureable or difficult to measure) differences between cables. It wasn’t always like that. In the early days those of us who’d noticed this stuff took lots of hits in the pro community.

The notion that cabling characteristics affected sonic quality originated among the audiophiles. The “golden ears” were formed into armed camps on this issue. Harry Pearson et al at “The Absolute Sound” heard the difference, and Peter Aczel, did not. Pure tech guys (that’s the AES, is it not?) viewed and continue to view this topic as pure rank voodoo.

I tried some early tweako cable on my home rig (circa 1983) and heard something or other change. I have experimented with cabling and other tweaks since, at home and in the field. I am utterly unqualified (few are) to discuss the known physics of cable design and optimization. I have experimented with cabling in home and professional systems since 1984 and have noted significant changes as well as no change.

But that isn’t my primary point. There are those whose world view would be sullied by general acceptance of subjective criteria by analysis of sound. These are physical science types, who measure with their eyes using test equipment and the language of science. They have applied their paradigm to every aspect of the audio business and have made many dollars doing so.

Such types have also provided us with much of our hardware, for which they should be commended. It’s a safe bet they support the cable status quo. These same business people have also provided us some horrible permanent installations based on measurement driven analysis/specification programs that meet the requirements of architects and proof of performance, but often sound utterly wrong.

Of course they are well represented in the AES. (It’s their club, right?) By the way, I have no problem with the AES, as long as they acknowledge (don’t hold your breath here) that theirs is not the only way.

DOING THE DOUBLEBLIND

These business people reveal their amusicality when confronted with the notion that subjective analysis has at least as much if not more validity than their prized objective criteria. In such situations they immediately call for a doubleblind A/B test, wherein they swap from one item to another fast and the tests are used to identify differences in cables, CD players, amps, whatever.

Often, the statistical data derived from double blind tests yields the 50-50 equivalent of getting heads or tails on a coin flip meaning, from the AES point of view, that there is no difference between cables or amps or whatever, as proven by their “scientific method”.

This method ignores one component utterly critical to the live engineer. Live engineers process the shifts in the room sound as the day progresses and backline is cranked and mics are opened up in the rig, and the crowd comes in and further adjustments are made under load.

We work additively, not in bursts. For this discussion, I’ll call this the time/complexity gradient. A normal show has a noon or earlier in, 16:30 soundcheck, 18:30 doors, support at 19:30 and headliner at 21:15. The sound (not the audio folks, and yeah, I think I’m going to start a Sound Engineering Society) in that venue changes constantly from start to finish.

Our job is to look for buildups at 160Hz as more vocal mics are opened, and squeaks at 4K if the sides are too close to the mains. Decisions are constantly made about the system EQ, channel EQ, FX and many more sound issues as the day/show progresses.

A single short blind test is utterly irrelevant to how we hear or work. Such a test is merely an attempt to apply limited criteria to an almost unanalyzable event with such a broad time line. Much like measuring the PH balance of the aged balsamic at Charlie Trotter’s (a landmark Chicago restaurant), making judgments based on that measurement as to whether the salad sucked or not.

FUNNY IS AS FUNNY DOES

As we all know, attempts to apply AES style measurement to tweaking and God forbid mix/adjusting live show systems usually produces hilarious outcomes. I will never forget such an occasion on a huge tour where I was doing speaker sales meetings in each city (We had sold the sound company their kit) and spent a fair amount of time near the mix position watching the show.

Act B had an active measurement system and a programmer who had snapshots of each song in his computer that he quickly tried to apply to the system for each using a well known hardware/software package.

These guys had their own console and outboard, and a very well known engineer was in attendance radiating attitude. (I will be merciful and not mention names) They were, of course, associated with a speaker brand other than what we had at the site, and various discouraging words were heard about the superiority of same, blah blah blah.

This is not about the speaker wars, as I’m kind of partial myself to the stuff they liked. It is about that silly-assed application of measurement to an ongoing mix. So they do their set and it sounds OK, but I wish I had a video of the engineer emeritus breathing down the neck of the poor slob doing the active adjustments.

The engineer with act A had been sitting next to me during this display and mentioned that the rig sounded a little muddy with all that stuff being applied to it. He was being nice.

To me it sounded like a great sound system had been made generic and inoffensive with no dynamics and no emotional optimization of the mix, because: These idiots probably didn’t know how, and They spent all their time servicing their measurement/EQ gizmo.

Soon after, act A engineer, Jeff Dunn, then with the Black Crowes, walked up to his desk, futzed the 160x’s a bit, and flattened the preceding act. He was in service to nothing except the mix.

His concept was rock and roll, and the thing breathed and had life and changed in response to music/show (not measurement) criteria. Even better, he was fast and loud and it didn’t feed back.

The Act B crew sounded most lame (and yeah, you know who you are) and I had the best laugh I had in years as I observed looks of shock and envy over by the act B desk. Measure this, lads, I thought.

THE EPHEMERAL ART

Performance Audio is ephemeral. We don’t have a rewind button (yet anyway). We must constantly quality control what is happening and fix errors fast, within a few bars of a song. Then you must move onto the next problem.

There is no way to quantify this multilevel/multitasking process. (Where’s that buildup, house or stage? Is it 200 or 160? Should I recut the bandwidth on the bottom band of EQ on the bass DI to minimize this mess?

What about panning the bass a little house right to change the audible impact of the standing wave? Let’s punch up kick and bass at the same time and recut both bottom bandwidths to spring them loose from each other. Maybe that will help.

Screw it. I’ll call the stage and have that S.O.B. turn down, because I now realize it ain’t feedback. Instead, it’s that silly Acoustic 371 (bass amp) dinosaur: modulating the deck, and creating a huge bump in the house.

Some folks require a generic world. They need to package their experiences within predictable structures, i.e., scientific method. The total anarchy of live performance flips them out, as they know how hard it is to manage the sound of a show.

They probably know they aren’t up to it, so they impose a set of rules with which they are comfortable so that they can participate. The fact that their rules answer few, if any, useful questions and usually lead to crummy sound is irrelevant.

Such rules and scientific method provide solace to the unwashed, less sophisticated end user corporate types and church committees. These folks need a piece of paper or LCD readout to tell them the sound is OK, because they are fundamentally incapable of making their own judgments and won’t spend the time to learn how.

BEYOND QUANTIFICATION

So much of what we encounter in Performance Audio can’t be measured or predicted anyway. How can you measure the rightness of the sound of a console, for example?

A Yamaha PM4K measures pretty flat, though it sounds a bit harsh (hard 1.6 ­ 3.15). A Midas would never measure that flat. Depending on vintage, there is a dependable Midas bump in the upper/bass low mids, that sounds warmer or less harsh. Is there a warmth-o-meter, or do we just have to depend on our ears to evaluate the sound of a desk?

And what about amps. I challenge anyone to take a BGW 500 GTA, which has a real power supply the size of a bowling ball, and shoot it out against any switching amp.

Assess both for tone, loudness, bass extension and the BGW will vaporize amps with three times its RMS power rating, because our current measurement/spec arrangements can’t predict the sound quality of amplifiers in any really useful way.

Theoretically, Power Distro (PD) is totally quantifiable in its effect on sound, right? An amp at rated output pulls x amperes single phase and that’s it, or so we are led to believe.

I was on the road doing a speaker sales presentation, and directed by the customer to a rehearsal hall on the side of a hill in Seattle. Some huge act or other was down the hall rehearsing, so they stuck us in a large room where there was to be a rave later that evening.

They had many 60amp circuits, so I was a pig and gave every amp 60amps and did the same with the processing. I had played that demo system for clients across North America, and it never remotely sounded like that, period.

What were we (not) hearing? Can’t say, but we sure as hell had removed PD from the sound equation and the difference was not subtle. The AES method to PD is to establish what the amp needs at rated and let it go at that ­ as long as it doesn’t pop the breaker at that point then things are jake.

Well if you look at it that way, you will never get whatever it was I got that night in Seattle, and believe me, you want it, assuming that you can glom enough PD.

REMEMBER WHEN CDs CAME OUT?

Our AES friends and their acolytes chanted “perfect sound forever”, but it didn’t work out that way and now people have realized how crummy CD really sounds and there is a huge market for tube and class “A” stuff to impose some humanity on the antiseptic sound of digital signal chains.

Engineers who went with their gut, not gadgets, realized that something was amiss and started slapping tubes on output stages and inserts. They could have cared less how the stuff measured. They just knew that something was wrong and hacked together a fix.

It is up to the engineer to be man (woman/whatever) enough to sort out how to qualify what is good sound, and how to deliver it to their clientele, without requiring vindication for their decisions from outsiders who have their own (limited) agenda (re: control and power).

There is a famous audiophile quote, (not mine, but I wish it was) that succinctly speaks to these issues:

“If it sounds bad and measures good, it is bad. If it sounds good and measures bad, then you are measuring the wrong thing.”

Richard Heyser, whom Doug will discuss at some length, gave much to sound/audio, including TEF, which works quite well in many static situations in the hands of a good practitioner, although it has little to lend to Performance Audio.

Heyser also understood the importance of subjective method. Even though he is considered one of the greatest (maybe the greatest) audio physicists, he knew that the test equipment wasn’t getting the whole story.

Heyser argued for convergence, which is not the message that his intellectual heirs at AES push. It’s the same tired story with these AES guys. Their agenda is most evident in their educational outreach, to which I was once subjected.

So, I’m in the back of a lecture hall and have been threatened with physical violence by my faculty colleagues should I be disrespectful to the AES representative during his sermon. Many of the students are aware of my views on this, and as I scan the room I see their looks of anticipation along with Doug’s warning glare.

The message was recycled BS: All amps sound the same, digital is perfect and cabling differences are the rankest voodoo, a mere figment of human perception. (This guy needed to be reminded that it is human perception that mixes the shows).

I lean back staring at the ceiling dreaming of having this pretentious wanker in the front row during my lecture wherein the role and importance of the live engineer in the audio food chain is thrown down, as it were.

IT GOES LIKE THIS

Some of you may be aware of the story of the Battle of Midway during World War II. The U.S. forces had four aircraft carriers and duked it out with four enemy carriers and managed to sink all four of them with torpedo planes, while only losing one carrier.

Someone had to design, build, and test those carriers and planes. And the Admirals had to get the fleet to the right spot at the right time to engage the enemy. And someone had to maintain the torpedo planes, and others had to peer into the cloud cover and try to figure out where the enemy ships were hiding.

Still, certain other individuals got into those flimsy plane and flew off into the void. These brave souls found those ships, made their runs in the face of immense defensive fire, and sank the damned ships.

They had control of the show and did what was necessary to bring home the victory. Sure, all these other folks mattered, but you are either the one flying the plane, or you aren’t.

It is no different live. If you can’t take full responsibility for the ongoing tweak and mix of the show, please go back to your suburban basement, turn on Rush Limbaugh and play with your fluke or something, while the adults handle the heavy lifting out in the world.

Finally, if you feel offended by all this attitude, please come out with me and do me up a hundred-channel, six zone symphony mix with a five minute soundcheck. I’m confident the crowd will enjoy the double blind test.

Reference Note: Check out some the papers Bruce Brisson (designed the original Monster Cable, and then formed his own company M.I.T.) has written, along with multiple reviews in Absolute Sound and Stereophile during the last fifteen years.

Jack Alexander is a man with many hats. Among those he instructs on topics allied to Performance Audio at Columbia College (Chicago) and can be reached via e-mail at jalexander@popmail.colum.edu

September/October 2001 Live Sound International