Viewpoints: To measure or not...
Do test and analysis tools help ­ or hurt ­ the process?

By Doug Jones

“It’s why an MRI operator doesn’t interpret the scan; rather, a physician with much more education and experience does this.”

Measurement: good God y’all, what is it good for? Absolute-ly nothin’ ­ sing it again... Is this going to be our new anthem? To the “non-boomers” out there, I apologize for this archaic musical reference. But it seems to me that we need once again to separate out the process of measurement from the process of analysis so we can try to fix what is broke in our thinking.

Fixing what is broke is the whole point. Recently we set out to install a vintage console at our new facility at the college. In condensed form, this process illustrates the line of reasoning. The power supply was connected to the console, we turned it on... And all sorts of warning lights activated. Here was our first bit of data.

The console system was reporting that a power supply had failed. Something was not the way it should have been. The very first thing I did was to try to verify that the console was telling the truth. Before we “killed” the supply (taking no more than a few seconds), I verified that the logic systems of the console were totally messed up, consistent with an eight-volt supply failure.

TWO POWERFUL WORDS

We had collected data, now we needed to do some analysis. In other words, how can these data help us get the console up and running? Well, we knew that the power supply was down, or at least it was not getting to the console. Where to go from here? To figure out what happened, and why. Again: What and Why. Two very powerful words, indeed.

“If what you really want to do is mix, don’t worry about learning how to run TEF.”

The question “what?” can often be answered by measurement. We disconnected the supply from the console and through measurement confirmed that the supply was not delivering eight volts. We opened the supply and discovered through measurement that the 60-amp fuse for the supply was blown. (We also discovered in the three large binders of schematics from the manufacturer that there was nothing about the power supply other than the pin-outs for the connector ­ but that’s a different story!) At this point, readers, I’m sure we divide into two camps. Some will say, “A 60-amp fuse was blown ­ how did that happen?” The others will say, “so?”

The difference between the two groups is experience. I know that a 60-amp fuse blowing is significant because of my 30 years in pro audio, because my dad taught me how to fix cars at an early age, and because Ray Rayburn taught me how to fix Ampex 300 tape machines. And also, because I’ve had thousands of experiences with troubleshooting “stuff” over the years.

Until this point, I’d never seen the insides of this particular power supply, but I knew right away that it was significant. So did Dave Dennis, Brett Johnson and Ray Morales, all of whom were working with me. Together, we represent 90-plus years of experience in fixing stuff.

O.K., so measurement told us that a power supply was dead and that a 60-amp fuse was fried. But analysis and experience told us far more.

WORK OF ART

The specific problem was that the 8-volt sense line was inadvertently connected to the 48-volt phantom power rail, causing the “crow bar” on the supply to clamp down, blowing the fuse in addition to taking out a small signal transistor on the regulator card. At the same time, all 16 of the 3055s were spared. Whoever designed that supply has my undying admiration. It’s a work of art. The console is now up and running.

The measurement device used was a Fluke DVM. We could have used a Simpson 260 analog meter; we could have used a Radio Shack meter. Or we could have used a Heathkit VTVM if a working unit could still be found. The point is that we did not need fancy digital gizmos. An ohm meter setting served to check continuity and test the suspect transistors. A volt meter setting checked voltage. The rest was analysis.

“We need to know what to do with the data we collect.”

Making the measurements with the right test equipment and using our experience were both necessary to get the thing fixed. If we had no test gear, the supply likely could not have been fixed. At least by us. Lets bring this back to reinforcement. In our power supply story, we pulled out the test gear because we needed information we could not get any other way. (Well, I guess one of us could have put our tongue across the rails to see if it could deliver 60 amps, but I believe there are laws prohibiting this.)

Here is the first principle: use test gear to get information you can’t get any other way. Does your sound system sound good? If yes, fine. But what if it sounds lousy? If you know why the system sounds bad, and there’s something that can be done about it, you’ll probably do it. (Duh.)

The tricky part comes when the system isn’t sounding right and you don’t know why. Then, measurement makes some sense.

The next principle is using the right test gear for the job. If you’re trying to measure distance, don’t use a thermometer. If you need DVM, don’t use TEF. If you need TEF, don’t use RTA. If you don’t know what to use, you probably don’t have enough experience to make the measurement in the first place.

The final principle (for now) is making sure that measurements make sense. In the class I teach on acoustic measurement, I see this principle in action all the time. Students set up the analyzer totally inappropriately for a given measurement, then make the measurement and present the data ­ without ever stopping to think about what is presented.

PROBLEM WITH YOUR CHOPS?

One of my favorite torture tests in the old days was to set up a TEF 12 using the cosine output instead of the test output. The sweep sounds the same, and data is collected. But the data makes no sense.

One wouldn’t realize this, however, unless it’s looked at carefully and there is understanding of what is being measured, and why. If TEF says that there is no high end in the system, yet you feel like 8 kHz is drilling a hole through your forehead, there might be a problem with your measurement chops.

“Students set up the analyzer totally inappropriately for a given measurement.”

I agree with Fred that we need training. However, back to the power supply story. Training in the use of a DVM would not have enabled one of my students to fix the supply. I’m not trying to discourage folks from getting adequate training on sophisticated systems, and for example, Gold Line offers very good TEF classes while Syn-Aud-Con seminars are also a very good training ground. But what’s really needed is much deeper than simply learning how to operate an instrument. We need to know what to do with the data we collect, and this is much more difficult to teach and learn. It’s why an MRI operator doesn’t interpret the scan; rather, a physician with much more education and experience does this.

I’m not suggesting that we need to go to a model where technicians measure and an elite class interprets. If what you really want to do is mix, don’t worry about learning how to run TEF, much less learning how to interpret the data.

However, if you want to set up systems and make them sound good, chances are that at some point you’ll need sophisticated analysis to get you out of a bind. More than a user manual is required to pull this off.

Doug Jones is the Chair of the Sound Department at Columbia College. Reach him at djones@livesoundint.com

October 2003 Live Sound International