SOAS University of London

Sound Recording: Microphones

David Nathan, ELAR

For language documentation and archiving, it is important to record the best possible sound quality for the following reasons:

  • your recordings may form the only documentation of the language or specific performances in it
  • people will want to use your recordings to study or learn the language in detail, including pronunciation and intonation
  • it is likely that products will be made from your recordings that make unexpected juxtapositions of sounds (e.g. a speaking dictionary or concordance)
  • unlike film, there is no contextualisation, such as a view out of a window establishing that there is a busy street outside. Such contextualisation orients the film's viewer to the environment, so that environmental sounds do not detract from the audio. This is not usually possible when recording sound only
  • you or someone else may have to listen intently to the recordings for many hours to transcribe them. This will be tiring and uncomfortable if the sound is not well recorded.

The microphone is the most important item of your recording equipment. For most purposes, the main microphone should cost at least 50% of the cost of the recorder, and at least US$100. (Note that in recent years digital recorders have dropped in price, while the cost of quality microphones remains much the same, so the relative cost of microphones has increased).

Different microphones have different capabilities, strengths and weaknesses, and a fieldworker should have a (small) variety of microphones for handling different situations and choose appropriately.

In any case, remember that the usefulness of your recordings will depend overwhelmingly, and equally, on two factors:

  • quality; and
  • consistency

Consistency is important because psychological factors are so great in the perception of sound, and because the eventual usages of your sound resources may result in arbitrary segments played in juxtaposition.

Factors in choosing microphones

Choose microphones in regard to factors including:

  • directionality. Choose a directional (also known as unidirectional or cardioid) microphone, or an omnidirectional microphone, depending on the circumstances.
    • Directional microphones are typically better for linguistic work, because you can reduce sound pickup from sources other than the speaker. For stronger directionality, use a shotgun or hypercardioid microphone. All of these will give good results but require more work, as they must be well-positioned and aimed. Microphones with greater directionality do not have a pickup zone that extends further, they simply have a more narrowly restricted zone of pickup. Directional microphones are more sensitive to wind noise and popped aspirations, so you may need a foam windshield or equivalent.
      Omnidirectional microphones pick up sound from all directions with more or less the same sensitivity. In many cases this is not what you want, although lapel microphones are often omnidirectional.
  • condenser or dynamic . Condenser microphones are also called capacitor microphones; some types of condenser microphone are called electret.
    • Condenser microphones are more accurate and provide more output but require a power source. Condenser microphones are nearly always used in language documentation.
    • Dynamic microphones are less accurate and provide less output, but they do not need power, and are more robust in adverse physical and climatic conditions such as high humidity.
  • impedance. Low impedance microphones are more accurate and pick up less interference, but tend to be more expensive and more fragile. The microphone's impedance must match the recorder. Some recorders have sensitivity settings to handle microphones of different impedances but you should always check compatibility before buying.
  • power requirements. You need to consider the availability of mains and battery power for your microphones, recorders, and rechargers.
    • Condenser microphones require batteries, which may need to be replaced, or use "phantom power" from the recorder itself, which adds a power burden to the recorder if it is running on batteries.
    • Dynamic microphones don't require power but are not as accurate. Nevertheless a good dynamic microphone may be suitable and can provide a good second or backup microphone.
  • size/form. Microphones come in several shapes and sizes. Shotgun microphones are often long and thin because they are normally attached to a boom. Lapel microphones, also known as lavalieres, may be useful but risk the dangers of handling or movement noise as the speaker rustles his/her clothes or moves about, or moves the cable, which can destroy the recording. Alternatives are to attach the microphone clip to a speaker's hat or glasses. Use of lapel microphones may require "wiring up" the speaker and/or having a radio transmitter; they are perhaps better used in controlled studio situations where the speakers can be instructed and are sitting still.
  • stereo/mono. It is necessary to use stereo to make the authentic, high quality, multipurpose and easily-listenable recordings appropriate for today's language documentation. Stereo is achieved by using a stereo microphone (actually a housing with two microphone capsules in it), or two mono microphones that you can place independently, or set up in a special configuration, such as ORTF. Mono recording was previously considered acceptable for linguistic work, and has the advantage of halving the amount of data to be stored (in most digital recorders). In either case, make sure that you are aware of the equipment you are connecting; for example, some computer microphone input jacks and some microphone plugs are mono miniplug. In some combinations, this can result in recording only one channel of a stereo signal, which will result in a poor recording. (On the other hand, a mono microphone with a stereo plug will split its signal and there should be no problem). For higher quality equipment, XLR (or "Canon") connectors are often used; you should make sure that you have and test the right combinations and extensions etc.
  • use of stands/booms. It is necessary for any high-quality recording to get the microphone as close as possible to the speaker's head (without being so close as to suffer from plosive pops).
    • The ideal is a hand-held boom or extension-piece enabling someone to continuously locate the microphone within 30cm of the speaker's mouth and pointing at the speaker's forehead. However, this is not always possible, and requires more people.
    • Simple handheld mode is a good way to manage a microphone if you have limited equipment and crew, but can be tiring for the person holding the microphone after about ten minutes. Using a handheld microphone gives the most control over its position and direction. The person holding the microphone should take special care in handling it. Some microphones (and their cables) are very susceptible to handling noise, such as caused by the hand tensing or moving around the microphone. Never, of course, cover the meshed end or any holes around it. If you ask a speaker to hold the microphone, show them where and how to hold it (it may help to record some good and bad examples to listen to).
    • Table stands (or tripods) for microphones are effective if they allow the microphone to be located close enough to the speaker’s mouth. If the table is too far from the speaker, background noise levels will be higher. Take care not to disturb the table: vibrations from tapping on the table, kicking the table leg, or walking on the floor, will be transmitted through the stand to the microphone. A thick insulating mat between the table and stand will reduce this type of noise transmission. The presence of a table may encourage noise from shuffling papers, which will damage the recording.
  • balanced or unbalanced. Some professional equipment uses balanced connections, where the voltages in the recorder, cables and microphone are arranged so that cables do not pick up extraneous interference from the electrical/magnetic environment. For many situations, especially where connecting cables are not long, this is not necessary, but beware of incompatibilities and that, at the worst, unsuitable combinations can damage components.
  • automatic level control (ALC). An ALC function is built into some recorders and cameras. ALC will try to raise the input level when the input volume is low. There will naturally be low input volume when the speaker speaks very quietly or is silent between utterances, and ALC will cause the recording of annoying changes in background sound at the beginning and end of utterances, and unexpected levels of background sounds. ALC should always be turned off; equipment with ALC that cannot be turned off should be avoided.
  • using a video camera. If using video, don't forget about sound recording. Substituting video for audio without attention to sound quality may leave you without any good record of the language sound at all. Avoid using the video recorder's internal microphone. However, a video recorder with a good, compatible, well-located, external microphone can make excellent recordings, and can serve as a second or backup sound-recording system. Take care with the microphone cable and avoid any mechanical noise made by the camera itself.

Microphones should be selected to be compatible with recording equipment. You will need to take into account the recorder's price, quality, connections, power availability, manufacturer's recommendations, and compatibility advice from vendors, professional users, and others.

Preparing and practising

The fieldworker should spend at least half a day thoroughly practising and testing equipment and recording methods by recording in a variety of situations. For example, check how the equipment performs when recording at various distances from a speaker, how it handles background noise, its response to physical handling (e.g. moving hands on the microphone or its stand, boom etc). To evaluate recordings, you must use closed headphones (so that you are experiencing and focusing on just the recorded sound); mid-priced headphones (US$25-$50) are quite sufficient.

The fieldworker should have an idea of the sound properties (such as echo) of various environments and surfaces, and understand that sound perception is strongly psychological and different to the "reality" that will actually be recorded. Although your mind can focus on a particular sound source to the exclusion of others, or ignore a sound coming from a direction you are not interested in, a microphone cannot: "what you hear is not what you get" (Rose 1999: 94). For example, microphones pick up more echoes, reverberation, and distant sources than you would normally consciously hear.

Adapting to the fieldwork situation

Prepare strategies to deal with situations such as playing children, refrigerators, traffic noise, wind, fans, and electrical interference from cabling, computers etc. When at the recording site, survey the expected recording locations in regard to their acoustics – not only within the building or immediate area but also the direction of noise sources such as roads. Some simple steps can help: if indoors, move the speaker away from walls, and, if hard reflective walls cannot be avoided (for example by working near curtains, bookshelves, furniture or other textured surfaces), then try to face the speaker at an angle (i.e. not square-on) to walls in order to lessen sound reflections. The sound environment changes throughout the day, so if possible, acquaint yourself with the daily sound rhythms of the location (e.g. of traffic, household activities, homecomings, sounds of birds and animals) and plan accordingly. If using a radio microphone, it is especially important to test it at the actual location and time of recording because this technology is the most vulnerable to interference, loss of signal, flat batteries. Although a radio-transmitted lapel microphone might seem like a good choice for field situations, Rose (1999) recommends that radio-transmission is only reliable in controlled studio situations. (It may also not be legal in some countries.)

Turn off all mobile phones (cellphones); not only can they ring, but they regularly emit signals to page their base station which may seriously interfere with your recording.

For outdoor use (and for indoor use of shotgun microphones that are sensitive to pops and aspiration), you may need to protect the microphone against wind noise. Again, familiarity with the equipment and testing and monitoring are important. Typically, foam windshields are used, although in an emergency you can wrap some open weave or other low-density fabric around the head of the microphone.

You should also take into account the people you expect to record; whether there will be one or more people, whether they are likely to perform simultaneously (e.g. singing), their mobility, acceptance of guidance, comfort with technology, possible sensitivity of content, and the privacy of the location. Alert people you work with that touching the microphone, its stand, cables, or the table on which it is sitting, can damage the recording. Take special care with paper – shuffling papers can easily make parts of your recording useless.

If recording in multiple sessions, keep a record of the microphone(s) used, location, orientation etc, so that you can achieve consistency where necessary, such as when recording for a talking dictionary, where arbitrary sounds can get sequenced when the final product is in usage.

Listening and monitoring

You will need some ways to listen to the sounds you record at various stages. Have at least one pair of headphones. Use closed headphones to check a sample of your initial recordings to check sound level and quality, paying special attention to background sounds, echoes and reverberation. You might use open headphones during recording to monitor the input from time to time. In addition, it is worth having a small pair of speakers to share and review the recordings with participants.


Quality brands include (in alphabetical order) AKG, Audio Technica, Beyerdynamic, RØDE, Sennheiser, Shure, and Sony.

References and useful information

Rose, Jay. 1999. Producing Great Sound for Digital Video. San Francisco: Miller Freeman
Ladefoged, P. 2003. Phonetic Data Analysis. Oxford: Blackwell.

This document

Author: David Nathan

Version: Second revision

Date: 15 May 09

Acknowledgements: Thanks to the following people for helpful input: Rob Kennedy (Language Centre, SOAS), Jerry Glasgow (AV, SOAS), and Gary Hammond (AV, SOAS).