With digital recorders and editing systems so universal these days, why talk about analog audio? Because there are still analog stages in most of the work we do, whether it’s getting the sound from your microphone into your recorder, or sometimes from the recorder to a digital workstation.
In order to take best advantage of digital encoding, the audio signal should be delivered according to good analog practices. You are already at a disadvantage if the audio is damaged before it reaches your digital processors. Whatever device you are using to make the analog to digital conversion, make sure to feed it the proper level. Sending a microphone signal (or too low of a line level signal) into a line-level input will result in noisy, coarse sound, encoded at effectively a lower bit-depth. Sending a line level input into a mic level input will usually create distortion and clipping.
Confusingly, the most common audio connectors are used for cables that transmit several different signal strengths, and some are even used for both digital and analog connections. So it is important to pay close attention to the type of signal that is being sent, just because the connectors look the same does not mean that they are compatible. The most common types are 1/8-inch (“mini”) 1/4-inch, RCA and XLR cables.
Three-connector “balanced” cables, either XLR or 1/4-inch TRS (tip-ring-sleeve) provide better shielding from noise, and so are generally preferable for pro uses, especially if the cables are very long. Make sure all cables are wired the same way, there is oddly no standard for wiring balanced cables and one can encounter cables wired with pin 2 or pin 3 as the “hot” signal. These wiring differences could lead to sound that is “out of phase” where the peaks and troughs of an audio waveform are electrically reversed between channels. In the extreme, this can cause mono signals to cancel out completely, but more commonly it can weaken the sound, giving it a hollow, sometimes metallic sound with poor imaging.
Although balanced cables use three conductors to reduce noise, headphone outputs use 3-conductor jacks to transmit stereo sound on one cable. Outside of small, portable recorders and some computer soundcards, there are virtually no inputs that accept stereo signals this way, so if one is connecting a headphone output, (or a line-out that is carried on a stereo mini cable) to another device, it usually must be adapted to terminate in separate left and right cables. These adapter cables are commonly available for connecting portable audio devices to one’s home stereo.
There are a myriad of different signal levels one can encounter in analog audio, but there are three ranges one is most likely to deal with: Microphone level, -10dB line-level and +4 dB line
There is the added complication of impedance, which is the amount of resistance in an alternating-current circuit. Low impedance signals are better for long cable runs, and so professional equipment usually uses low impedance, balanced signals. Consumer gear (including microphones) that uses 1/4″ or mini connectors is usually high impedance, which is fine as long as cable length does not become too great. The rule of thumb is to connect devices with similar impedances, but low impedance sources can usually be fed to higher impedance inputs without any problem (in fact it is desirable for the source to have a lower impedance than the input.) High impedance sources generally put out a stronger signal, so if, for example, one is not getting enough volume from a low-impedance microphone to drive the high-impedance input of a minidisc, using a transformer to raise the impedance of the microphone a bit can give some level boost.
Because digital audio transmits at much higher frequencies, impedance matching is more important, so be sure to use cables with proper impedance for digital signals.
The tricky part is that very different signal strengths can use the same types of connectors, or the same signal strengths can use different connectors. One can get distortion by plugging an XLR +4dB line level signal into an XLR microphone input, or insufficient volume by plugging a 1/4″ microphone cable into a line-level input. Adapters can help in keeping outputs plugged-into the correct inputs, but simply using an adapter to change one end of a cable to a different type of connector does not necessarily make the two signals compatible, the type of connector does not indicate signal strength.
Portable minidisc recorders, DAT recorders, mixers, and some computers’ basic soundcards often have mic-level inputs. Microphone level is usually somewhere in the range of 40-60dB lower than line level, and those signals need to be amplified before going to tape, or a mixer channel, or digital converters. “Mini” plugs are often used for mic-level inputs on small, portable devices, such as minidiscs or the Sony D-7 DAT recorder. Many computer soundcards use these connectors as well. The most common connector for professional microphones is three-pin XLR.
Most “consumer” tape recorders and minidisc machines output their signals on RCA cables or mini connectors, and operate at Ð10dB line level. Similarly, soundcards with RCA or unbalanced 1/4″ inputs and outputs usually operate at that level.
Pro machines and interfaces using balanced (three-conductor) XLR or Tip-Ring-Sleeve (TRS) 1/4″ connectors usually operate at +4dB levels. One could use adapters to connect audio from an unbalanced Ð10 source to a balanced XLR input, but the levels will not match properly. It may successfully transmit sound, but often with poor signal quality: hums buzzes or thin, brittle sound. For better quality, a leveling transformer should be used, or one should simply be careful to only connect compatible level signals. Only plug mics into mic inputs, -10 sources to Ð10 inputs, etc.
Headphone outputs are another flavor of line-level, with variable amplification, that can get louder than the line level of a fixed output. Headphone outputs are notoriously noisy, and should only be used as a last-resort for anything other than driving headphones, Some small minidisc recorders only provide a headphone output, and so, one needs to use that connection to simulate a line-level signal. This can actually be an advantage in some circumstances, as one can use the headphone volume control to send the optimum level into the computer’s input.
Maintaining or manipulating proper signal level comes into play at several steps along the recording and mixing process. The final volume control, whether it is on a mixer, a tape recorder or a virtual channel in a digital workstation is only one factor in the equation.
Use your meters, but always listen while recording your sound, whether on headphones in the field, or on studio monitors in a control room. Low levels can be raised later in the game, but the audio won’t sound as good as tracks that were recorded properly to start. Levels that are recorded too loud may have distortion that is impossible to remove.
Remove obstacles to the airwaves.
Help Transom get new work and voices to public radio by donating now.
Many audio inputs, such as those on mixer channels, digital converters, and mic preamps, have “trim” controls, often a small knob near the input jack. That knob regulates the sensitivity of the input to that channel, the amount of pre-amplification or attenuation that is applied to the incoming signal. The range may be fairly small, to adjust for different microphone outputs, or be large enough to accommodate everything from quiet microphones to +4 dB line level. The typical procedure is to introduce a signal to the channel, and only that channel, with any later level controls set to “unity” (no boost or cut, often marked on a mixer’s channel fader with a zero, a detent, or a line.) Then the input trim is adjusted to produce the desired signal strength at the output.
That optimum signal strength will vary depending on your project, and could range from aligning peaks to full-scale on a digital meter if working with mono or stereo material, to several dB lower if setting-up for a multitrack mixdown. A one-thousand kilohertz (1Khz) sinewave test tone is often used to set gain levels. By adjusting the trim so that this pure tone registers at somewhere between Ð18 and Ð12 dB, a more typical real-world signal will usually have a proper signal strength, with peaks near zero. In general, one wants to set this gain stage so that each signal is maximized when the channel is set at unity gain, and all later controls, such as the faders on a mixer or the output control on a preamp, would be used to add only small amounts of boost or cut to the signal. This is similar to the process of “normalizing” a digital signal, getting each element to maximum volume, then using mix controls to drop the individual levels as needed.
In order to maintain best signal quality, it’s important to record as loudly as possible, without distorting. Your ears are very good tools for this job, but understanding your recorder’s meters is crucial as well. Old-style VU meters with bouncing needles are very good at displaying average signal level, but not as helpful for displaying peaks. Digital bar-display meters can sometimes be set to display either average or peak readings, but most often show peak levels. Both modes are informative as long as you understand what each is indicating. Peak indications are especially useful when recording to a digital format, because allowing even short peaks to hit zero on these meters usually results in unpleasant distortion. But peak-reading meters can imply that average levels are louder than they actually are. “Saturating” analog tape by allowing peaks to exceed the maximum level that the tape can physically encode sometimes results in a pleasing sound, indeed some music recording studios still record to analog tape in order to obtain this effect. But in most cases, the safe policy is to avoid distortion, especially when recording voices, by adjusting the recording level to stay below maximum level.
But no matter how good your meters, also use your ears. Some phenomena, such as very brief peaks, might not be fully indicated on your meters, but your ears can tell you if something is distorting if you use good headphones or accurate monitor speakers.
At the time of the final mix or mastering, the issue of peak vs. average levels becomes important. Even if your meters are registering signal at the maximum level, the audio may have more or less “loudness” depending on the average level (and some inherent qualities of the sound.) The higher the average level of your audio is, the “louder” and more present the sound will be. One can have an audio track that peaks at the maximum, but if the average level is low, it will sound quieter, and less present. And the converse is true as well, if average levels are too high, usually as a result of excessive compression or limiting, the sound will feel too loud, have no life, and sound unnatural and squashed.
Most current recorders or editors will record two channels, which can be two separate, discrete signals (such as one mic for the interviewer and one for the interviewee) or a stereo signal, which is made up of two related, synchronous signals (such as the output of a stereo mic, or a CD player.) Not everything that has two channels is stereo. While in the digital world a stereo signal is carried on a single cable, in the analog realm one generally needs one cable per channel (with the exception of stereo headphone cables).