AUDIOCLINIC (Aug. 1994)

But It Looked Okay on the Meter

Q. I was doing a charity recording of a "live" rock-club date, featuring musicians who seldom are in a position to play together. The recording was a unique, never-to-be-repeated event. I decided to use my Tascam 38 half-inch eight-channel open-reel recorder. I meticulously aligned it for 3M 996 tape.

At the session I bridged into the vocal mike circuits, premixed stereo vocals to two tracks, and premixed stereo drums to two other tracks. The remaining four channels were used for two lead guitars, bass guitar, and, on the last channel, harmonica and acoustic guitar solos. This was intended to give me a chance to optimize the balance when the session was done.

At the sound check, all levels were adjusted for the proper mix and recording level was checked, using the VU meters on the Tascam.

The rest of the night proceeded without a hitch. Three sets and seven reels of tape later, I came away with a certain euphoria from having captured this rare event for history.

A day later, I set up the deck for playback and mounted reel 3. Reel 3 was the first half of the second set-the best of the evening.

I found that nothing was on the tape! A quick review of the other six reels confirmed my worst nightmare. Nothing on any of them either.

I immediately descended into analysis mode, fighting off anger as I did so. The deck must have broken down, I told myself This excuse quickly evaporated as I confirmed with scratch tape and an FM tuner that the deck was fine and recorded perfectly.

I went over each detail of mixer and mike preamplifier connections, yielding nothing obvious. I went into hypothesis mode: Per haps the reels of tape were wound such that the backing was against the heads. Nothing doing! It finally dawned on me. Perhaps, in a surge of overconfidence, I had plugged the mixer into the deck's outputs. That couldn't be! How could I have gotten such perfect levels even when the deck was in "reproduce"? I quickly pulled out the maintenance manual and reviewed the wiring. The VU meter circuits in my deck are resistively coupled directly to the output jacks, with no intervening buffer amplifier! The mixer drove the meters directly, making for the perfect illusion of a "perfect recording." Needless to say, I have burned into my mind a valuable and painfully learned les son, which I share with any of you who may own similarly configured equipment.

-T. Burkhard, New York, N.Y.

A. I've owned at least two other makes of recorder wired like that Tascam Magnecord and Ampex. The only reason I never made the same mistake was that the connectors for the output were different from those used at the inputs.

If it makes you feel any better, Mr. Burkhard, any of us who have been in the sound recording field for more than 40 years can tell his own horror stories. Here is one of my modern highlights of recording horror: I own a portable audio/video recorder that has the interesting ability to handle six (count 'em, six) stereo track pairs on a single tape. This is true only if the machine is switched to the "audio" mode.

Like you, I was making a recording that could not be made again. Sound was fine in my headphones. But when I got home, the playback was erratic. Worse, I lost my new recording as well as the other recordings on that tape. Leaving the machine in the "video" mode not only results in a recording made with no sync but also in the destruction of data on all of the stereo channels. Welcome to the club!

Load vs. Tube and Solid-State Power

Q. Many solid-state amplifiers' power output ratings almost double as the impedance of the load decreases by half Yet tube amplifiers, even the best ones, seem to be rated at a more or less constant power regardless of the impedance connected to them. Please explain this difference and any impact it has on matching amplifiers to loudspeakers.

-Tom MacGregor, Barre, Vt.

A. The power output of a solid-state amplifier would not increase when the impedance of its load decreased, if we truly matched impedances between the amplifier and its load. But the solid-state amplifier's output impedance is far lower than the load impedance (the amplifier's damping factor is the ratio between these two impedances), so the lower the load impedance becomes, the closer it approaches an impedance match with the amp's output.

We can never obtain a true impedance match, and we don't even want to. If we managed to match these impedances, the output stage would be destroyed as it tried to supply a huge amount of current to the load. This is much like attempting to match the impedance of a home appliance to that of the power line-simply a practical impossibility.

Tube equipment behaves differently, be cause the load is not connected directly to the output devices as it is in solid-state equipment. Most tube amps have output transformers whose primary winding is connected to the tube plates. The secondary winding is tapped in various places, and the speaker is connected between one end of the secondary and one of the inter mediate taps. The tap chosen depends upon the impedance of the load. The highest load impedance (typically 16 ohms) is usually connected to both ends of the secondary winding, with 8- and 4-ohm taps at intermediate points.

This maintains a constant load on the output tubes regardless of the load impedance being driven. Because the match to the tubes does not change, the amount of power delivered to the load will be more or less constant, regardless of the load. If the secondary of the output transformer was not tapped, all loads would be connected across the full winding. We would then ob serve a significant decrease in output for lower load impedances.

Building and Equipping a Small Recording Studio

Q. What equipment would you recommend for use in a small recording studio intended mainly for recording vocals? The dimensions of the area I have to work with are about 5 feet x 5 feet. I need to eliminate any undesired sounds. What kinds of materials would be required to build this type of room?

-George Marrow, Jr., Washington, D.C.

A., is difficult for a performer to work in a room as small as the one you are planning. I have to hope that the singers won't be sharing space with the recording equipment and the engineer. If nothing else, the equipment is bound to add some undesired noise that will be picked up by the microphones.

Although I understand that this room will be used for vocals, what about the mu sic behind the singers? How is that to be generated? If it is supplied by synthesizers, MIDI sequencers, or multitrack tape over dubs, fine.

At the least, you must use very heavy carpeting on the floor and really good acoustic tile on the ceiling and probably the walls. Don't forget to line the tile.

External sound often enters a room be cause the walls are set to vibrating. If this is your problem, you will have to make these walls as massive as you can. The stiffer they are, the less they will vibrate.

I have obtained excellent results by lining all walls and the ceiling with thick, dense fiberglass, held in place with chicken wire. (No, it is not pretty to look at, and some people feel very closed in when working in this environment. But if this can be overcome, it works well.) The boominess often associated with small spaces is very much reduced. You may find a need to add more highs because of the lack of reflection from room surfaces. You will also want to add reverb, because the sound is very dead.

In order to determine the equipment you will need, you must list what work you plan to do. Obviously you will need some kind of a mixer. It doesn't need to be an elaborate one, with "track solo" and "monitor sends" and the like. You may well want the opportunity to place some special-effects processors between each mixer position and the mixer bus for that stereo channel. Chances are that you only need two outputs.

The recorder can be a very good cassette machine, although I would choose an open-reel, VHS Hi-Fi, or DAT recorder.

Chances are that you will want to copy your masters onto cassettes, so the cassette recorder still has a place in your studio.

If you only plan to use a sequencer, you won't need a multitrack recorder. If, how ever, you need to sweeten the music with instruments not available in your sound modules, then you will need a multitrack tape recorder of some kind-digital or open-reel.

The choice of microphones is a very subjective one. Many sound technicians use dynamic mikes. They're great when a singer really shouts it out and runs the risk of overloading a condenser mike or over loading the input the mike is feeding. I personally use some ribbon mikes made just for vocals, because I like a natural-sounding voice, free from peaks. But many find such peaks exciting and enhancing to the performance.

Negative Feedback, Pro and Con

Q. Negative feedback is considered to improve amplifier performance. Why, then, do some high-end audio equipment makers avoid using it?

-Ray Segura, New Orleans, La.

A. Negative feedback definitely lowers distortion, but too much of it reduces sound quality. Years ago, a leading maker of audio gear introduced a high-quality line of preamps that used a tremendous amount of negative feedback, not only around the circuit as a whole but also around individual stages. To me, the preamps sounded mushy, as if the signal had to be pushed through them. When I removed some of the feedback from one of these units its sound improved remarkably.

However, preamps and power amps have different properties. Many power amplifiers, for example, employ Class-B output stages because Class-B circuits have good electrical efficiency and run cooler than other biasing arrangements would allow. This means lower costs for power supplies and cooling facilities.

However, Class-B circuits have inherently high distortion--especially at low power output levels, where they're operating in a more nonlinear portion of their input/out put curve. Adding negative feedback to such a stage dramatically reduces the distortion. Class-B or even Class-AB circuits would be virtually unusable for high-fidelity applications without negative feedback.

A high-end manufacturer, less concerned with cost, can lower distortion in other ways. He can, for example, use an output stage with inherently lower distortion (such as one using triodes instead of pentodes) and can also bias it for Class-A operation, despite much lower electrical efficiency than Class B. Even at low power levels, the input/output curve is always linear.

Measures like these allow a manufacturer to eliminate feedback or drastically re duce it. (For example, some amplifiers have no overall loop feedback but do use local feedback on early stages.) Yet I can't help wondering if the sonic performance of amplifiers that used no negative feedback at all couldn't be improved by using at least a bit of it.

(Audio magazine, JOSEPH GIOVANELLI, Aug. 1994)

= = = =