Manufacturer's Specifications:
Element: Back plate electret, permanently polarized.
Polar Pattern: Unidirectional (cardioid).
Frequency Response: 20 to 20,000 Hertz.
Sensitivity: 55 dB (0 dB =1 mW/10 dynes/cm^2).
Open-Circuit Sensitivity: 0.28 mV (71 dB) re: 1 v/dyne/cm^2.
EIA Sensitivity: -149 dB.
Impedance: 600 ohms nominal, matches 150 to 1,000 ohm inputs.
Maximum Input Sound Level: 125 dB.
Signal-to-Noise Ratio: Greater than 50 dB at 1 kHz, 1 microbar.
Battery Type: UM3 (AA) (or Mallory MN1500 Alkaline).
Battery Current: 200 microamps.
Battery Life: Up to 5,000 hours.
Weight: 6.5 oz. (185 grams).
Dimensions: 8 in. (204 mm) long; 2 in. (51 mm) head diameter, 13/16 in. (21 mm) body diameter.
Cable: Model AT8302, 16'/ ft. (5.03 m) two-cond. shielded, vinyl jacket, with Switchcraft A3F connector at microphone end, 1/4 in. phone at output end.
Accessories Furnished: Model AT8201 Slip-in Stand Clamp for 5A27 threaded stands, carrying case, battery.
Price: $110.00
The Audio-Technica Model AT813 is a carctioid unidirectional electret microphone powered by an internal battery. It was designed for audiophile or professional recording applications, or sound reinforcement. The large metal screen on the head end encloses an integral wind/pop screen. The handle includes a professional-type three-pin output connector and a power on-off switch. The output is balanced and low impedance. The cable supplied with the mike has a 1/4-inch phone plug on the output end, for use with many models of cassette and open-reel tape recorders.
A slip-in stand clamp as well as a padded carrying case are supplied with the AT813, and many other accessories are available. The Line Matching Transformer (AT-8201) will match the low impedance of the microphone to recorders having high-impedance inputs, and the accessory cables permit connection to three-wire balanced inputs. A snap-in stand clamp and a shock mount are available as alternatives to the fitting supplied with the microphone.
The Audio-Technica data sheet is unusually complete, as it includes a detailed schematic with component values and very clear illustrations on wiring of the output connectors.
The text explains the virtues of the so-called "back-electret" design which, in theory, should yield performance similar to the more expensive air condenser (externally polarized) microphones.
The data sheet indicates that the electret element is internally shock-mounted, and the accessory shock mount may be used also for additional isolation with stand or boom mounting. Most electret elements are extremely resistant to shock and vibration, an advantage for microphones which may be handled (or dropped) by performers. The handle and head of the AT813 are all metal and sturdily constructed, and the low-reflectance gray finish is highly scuff-resistant. The power on-off switch is recessed so that it is not easy to operate accidentally. The battery is easily replaced by unscrewing the head from the handle and is the readily available alkaline penlight type. I noted that the rated battery life (5,000 hours) is longer than that of a recently reviewed low-cost electret microphone which has no on-off switch. Many audiophiles will obtain shelf lifetime from the cell (two to five years) and ought to remove the cell from the microphone if it's not to be used for a week or more.
Therefore, the switch could be omitted, with a possible improvement in reliability and a cost saving. Any power switch which can be finger-actuated by the performer may occasionally be misused as a "push to talk" switch, so that AT813 should not be just handed to the user without also giving them a briefing. One alternative idea would be to change the switch to a type which must be actuated with a pointed tool.
The data sheet states that the AT813 is designed for very high input sound levels, but the rated SPL is only 125 dB. This is 5 to 10 dB lower than the ratings of other electrets I have reviewed, but some of those are internally powered by 6-V batteries or externally by 9- to 48-V power sources.
Naturally, the peak-to-peak output voltage swing cannot exceed the supply voltage. The instructions do not indicate that higher voltage batteries may be used to obtain higher input SPL/output voltage capability. (I would think that most FETs in microphones could withstand 6 volts, which could result in up to 12 dB higher SPL rating.) The frequency-response proximity effect (bass boost with close talking, which is inherent with most types of cardioid microphones) is cited as an advantage: If the microphone is used with a flat-response audio amplifier, the effect adds "warmth" which may be desired by the performer. If an external bass roll-off equalizer is used, the voice sounds natural, and distant low-frequency noise, such as room "rumble," is attenuated. My choice would be to include such a roll-off filter in the microphone which could be actuated by a third position on a power switch. An internal equalizer and "voice-music" switch is desirable on all directional microphones exhibiting proximity effects, unless they are used only in recording studios where equalizers are available on every microphone channel. The curves shown in the data sheet show substantial proximity effects, because the response is shown to be flat for sources at 12 inches or farther. Our measurements show that the response at 12 inches and farther is rolled off so that proximity-effect bass boost is less drastic than the data sheet indicates.
Measurements
Fig. 1--Impedance vs. frequency.
Fig. 2--Frequency response loss due to loading of our microphone preamp.
Fig. 3--Plane wave frequency responses for two AT813 samples.
The impedance-versus-frequency curve (Fig. 1) shows a nominal 600-ohm impedance throughout the midrange. The impedance rises to about 1,400 ohms at 70 Hz, which appears to be a resonance between the microphone transformer inductance and the coupling capacitor. As with other microphones I've tested, I "mismatch" the 600-ohm impedance to the 150-ohm output of the broadcast-type preamp in my test rack. This is an "unloaded" input with a center-tapped transformer. The actual input impedance varies from 1,000 ohms at the extreme low and high frequencies, to several thousand ohms at mid-band. Since our frequency response graphs are corrected to open-circuit conditions, it is necessary to show loading effects on a separate chart (Fig. 2). With most 150- or 200-ohm mikes, loading effects are negligible, but with 600-ohm mikes, some nonlinear responses are encountered. The loading of our preamp on the AT813 caused a 3-dB "valley" centered at 70 Hz. Loading could be more severe with other equipment rated at 150/200 ohms; however, the loading generally acts to reduce bass response and compensate somewhat for proximity effect. "Low-impedance" inputs of tape recorders vary from about 200 to 2,000 ohms, the rating based on the microphone impedance. The AT813 will work satisfactorily with most of these recorders; however, if the user is bothered by loss of bass with a 200-ohm input, or high noise (usually hiss) with a 2,000-ohm input, a line matching transformer may correct the problem. The Audio-Technica AT8201 line transformer may only be used with high-impedance inputs (approximately 50,000-ohms rating) though lower-ratio transformers are available from other manufacturers for line matching in the 200-2,000 ohm range.
We would also recommend that, to prevent pickup of radio stations or power line noise, a balanced circuit be used for applications requiring more than the 16 1/2-ft. cable supplied with the microphone. This will require cables with A3 connectors (such as AT8303), plus a suitable line transformer located near the recorder.
The measured frequency responses (Fig. 3) of two AT 813 microphones show smooth and flat characteristics from 100 to 5,000 Hz, which includes the entire range of voice frequencies. This is a desirable response characteristic for many voice and music pickup applications, and the bass rolloff (for plane waves or distant sources) is useful to attenuate room noises. The dips at 6,500 and 10,000 Hertz were initially thought to indicate a defective microphone (unit #1). A second microphone (unit #2) was furnished by Audio-Technica with a custom-drawn frequency response strip chart (Fig. 4). The vertical scale of the chart was not marked, but as a 50-dB potentiometer is supplied with B & K recorders, we assume that the chart is 50 dB wide. This is more compressed than our scale, but the response may be correlated with our curve (Unit #2 in Fig. 3). Such response variations at extreme high frequencies were not found in electret microphones previously reviewed. (See Audio, August 1980 for an example.) We suspect that they may be a function of geometry, that is, of the large head of the AT813, which includes pop noise blast filters. They will not affect voice reproduction, but could add undesirable coloration to very high frequency musical sounds. Thus, the designers of the AT813 may have traded some music pickup quality to obtain pop-free vocals.
The curves of the two microphones (Fig. 3) fall within a ±2 dB (4 dB wide) envelope. If these units represent extremes of manufacturing tolerance, then the variations conform to acceptable limits for professional microphones.
However, the audiophile is advised not to buy a pair of AT813s off the shelf, assuming they are matched well enough for, say, orchestral recording, because the mismatch between our two units would be easily heard. We do not usually test two microphones, but our review of the Shure SM-81 (Audio, August 1980) shows an example of frequency response curves of two units. Those curves match exactly except for 1or 2-dB differences below 100 Hz. The AT813 is a much less expensive microphone, but since electret elements are manufactured by automatic machinery (at a cost of a few dollars each), we would expect to find close-tolerance elements in microphones that sell for more than $100.00. The low-frequency response of the AT813, similar to other cardioids, varies with distance, more or less in accordance with theory (Fig. 5). However, none of our curves shows an essentially flat response down to 50 Hz, as the data sheet and the A-T strip chart do. We think that the observed bass rolloff (plane wave) is proper for a microphone which has no integral voice-music feature, because this results in flat response at 6 to 12 inches on voice. It also aids in reduction of room noise.
The directional frequency-response curves (Fig. 6) show a pattern which tends toward hypercardioid. A hypercardioid has more rejection than a cardioid at 90° (greater than 6 dB) but less at 180° (less than 15 dB). The greatest rejection for a hypercardioid is at approximately 135°. We are very impressed by the 8-dB rejection at 100 Hz and 90°. Rejection at 90° is important in sound reinforcement applications because speakers are frequently placed in the 90° plane. Hypercardioids, in these circumstances, will allow higher system gain before feedback than cardioids will. The directional characteristics are uniform and well-behaved over the entire audio range, and the peaks and dips at high frequencies simply reflect the response curves at 0°. The measured spectrum of microphone noise (Fig. 7) shows a pronounced hum peak at 60 Hz which could be extraneous noise pickup in the circuitry, and is perhaps related to the high microphone impedance at 60 Hz. This had relatively little effect on the overall "A" weighted noise level, which is 20 dB equivalent SPL. The AT813 can, therefore, be used in very quiet studios without introducing noise into the audio.
The peak output voltage (clipping level) on speech is ±0.2 volts. The equivalent rms input sound pressure level is 127 dB. This is adequate for most applications, but on extremely close and loud vocals, some clipping may occur.
Also, the AT813 may clip if placed extremely close to high-level musical instruments or instrument speakers.
The total dynamic range or maximum S/N of the microphone is 127 20 or 107 dB. The specifications refer to S/N relative to one microbar, accounting for the large difference between rated and measured values. Based on one microbar, rated noise is 74 50 or 24 dB, and total rated dynamic range is 125 24 or 101 dB. The microphone tested, therefore has 6 dB more dynamic range than specified. This is adequate for both analog and digital audio recording.
Last, but not least, the phasing test showed pin #2 positive, as per the specifications.
Fig. 4--Frequency response curve supplied by manufacturer for sample #2.
Fig. 5--Frequency response vs. distance from source (sample #1).
Fig. 6--Frequency response vs. angle (sample #1).
Fig. 7--Microphone noise spectrum (one-third octave bands) of AT813.
Use and Listening Tests
For comparison purposes, we used a Nakamichi CM-700 electret microphone with a cardioid capsule as a reference.
The reasons for the choice of the reference microphone have been stated in previous reviews.
The AT813, with distant music and speech sources, sounded very similar to the reference mike, but the AT813 audio was, pleasingly, freer of disturbing low frequency room noise (air-conditioner sound). The sound quality of the AT813 does not vary significantly on or off axis, same as the reference mike's. When used for close-up speech, the AT813 sounds much "heavier" than the reference, with the latter in Lo-cut mode. The overall sound quality with close talking is duller than the reference microphone's.
Magnetic hum pickup is very low, about 10 dB less than the reference. Pop or breath blast sound with the AT813 was about 5 dB higher than the reference. In this test, the reference mike was on Lo-cut with the accessory windscreen. We conclude that the integral pop filter may be doing a good job, but a bass-rolloff equalizer or voice filter is required to achieve higher pop noise reduction.
Vibration and handling noise is much less (approximately 20 dB) than the reference, so the AT813 is well designed in this regard.
Except for the previously mentioned potential problem with the accessible power switch, I find that the AT813 has excellent feel and balance for hand-held use. I think that most performers would find the microphone to be attractively styled.
As a final test, I listened to the AT813 with one of two test preamps connected for unbalanced input (one line grounded) and compared the sound to the other preamp which is connected for a balanced, center-tap grounded input.
There was no difference in sound quality, audio levels, or noise.
Conclusions
I think that the AT813 is a good choice for a wide variety of speech and music recording and sound reinforcement applications. It is an excellent microphone choice where feedback immunity is required. For close-up speech or vocals, a bass-rolloff equalizer may be desired by many users. Also, for sound system use, additional treble boost may be desired because of the flatness of the AT813's frequency response. The AT813 is a good choice for close-miking in popular music recording if the SPL limits are adhered to, though it is not our top choice for highly critical classical music recording using two microphones in stereo. The dynamic range of the microphone is excellent.
-Jon R. Sank
(adapted from Audio magazine, Sept. 1982)
Audix SCX-ONE Studio Mike System (Dec. 1992)
Brüel & Kjaer Type 4003 and Type 4007 Studio Condenser Microphones (Nov. 1984)
Microphone SENSITIVITY Ratings (Dec. 1976)
The Compleat Microphone Evaluation (April 1977)
The Compleat Microphone Evaluation--An Update (Sept. 1978)
Miking the PRO Way (Nov. 1977)
A Guide to COINCIDENT MIKES (Nov. 1978)
Build A Microphone Preamp (Feb. 1979)
= = = =