This introduction to location sound recording assumes no prior knowledge of film or video sound. It covers both solo shooters and those who have a sound recordist. Although it is aimed at beginners, those with more experience may also benefit from it. As well as those involved in the sound department it should also help others understand and best work with sound recordists, which is to everybody’s benefit.
- Having a separate skilled sound recordist is part of the real solution to getting great sound. This guide will help where one is not available but not having one is more risky and may well degrade the quality of your film.
- Don’t skimp on kit; get a decent tie clip and shotgun microphone. A good stereo mic for wildtrack/ambient is also a great asset. Aim to spend at least 25% the amount on microphones as you have on your camera.
- Microphone placement is crucial, there is no point is having a great microphone unless you know how to use it.
- The microphones built into Camcorders should generally never be used as they tend to be of lowish quality and can pick up camera noise. Even separate kit microphones are often fairly low quality.
Bad sound Can't be fixed!
There is a misconception that bad sound can be fixed, this may be one of the reasons sound can take second place to picture. The truth is that generally bad sound cannot be fixed. If someone who is not a sound recordist notices a problem, then the sound is probably already too bad to fix. Given a lot of effort, time and skill it may be possible to make things a bit better, but unless you are very lucky you will still have unusable sound.
A while back I was given footage of an interview with a nasty hum in the background. I tried to remove the hum by filtering out frequencies and this did reduce it. However, the frequency of the hum overlapped the frequencies of the human voice so when the hum frequencies were removed the voice no longer sounded full and natural. I may have removed a lot of the hum but the end result was worse, so the hum was left in. This is the kind of thing that often happens when sound is not properly recorded. The unwanted and wanted sound often overlaps, making the unwanted portion of an audio track impossible to totally remove. It is cheaper, safer and easier to ensure you get clean sound in the first place.
The ideal setup is to have a sound recordist who has a Portable Field Mixer (see right). The mic plugs into the Field Mixer. A lead from the Field Mixer runs to the camera or separate sound recorder. Recording separate sound allows the sound recordist to record when the camera is not running. The sound circuitry in separate recorders is normally better than the circuitry inside a non-pro video camera. This does mean syncing up all the audio and video later which takes time, but it is often worth the effort.
Full instructions on using a Field Mixer are beyond the scope of this guide but it is similar to using a mixing desk. Levels are constantly monitored using headphones and level meters, adjusting levels manually. For more details see B&H Photo Portable Field Mixers Guide.
Unless you have a proper budget you will probably not have a Field Mixer (they are expensive). The alternative setup is to have the mics plugged straight into the camera (or into a portable recorder), bypassing the mixer. If you are recording straight onto the camera, changing levels during a shot is probably not possible. Your options are:
- set the audio to auto which may give you reasonable results, especially if the camera has a decent limiter (check your camera manual and turn it on). The limiter will automatically reduce the level if the signal gets too loud but can do so abruptly, which is not ideal.
- set the levels manually. This is preferable but requires more experience and skill. You need to set the level high enough to get a good signal but never let it go off the top of the scale. If the audio is too loud and ‘peaks’ (going off the top of the scale), the sound will not be usable. This can happen for many reasons, e.g. the contributor may get excited and start talking louder than during the soundcheck.
Auto is probably the best option when you begin as most modern camcorders handle this well, but for best results levels should be set manually. Whichever approach you take, make sure you do a sound check ensuring levels don’t peak and there are no unwanted background noises.
It should be noted here that if you are shooting with a DSLR rather than a traditional camcorder, it is not possible to set the levels independently. There are several other good reasons for not employing the onboard audio recording in a DSLR; for more details see the Rough Guide to DSLR Video.
Camera Mounted Mic
The least desirable setup is to mount the microphone on the camera. If you are shooting on the hoof/run-and-gun without a sound recordist this may be the only way to go, but is generally less than ideal. The microphone will often end up further away than if the sound recordist were positioning it, and the sound quality will be degraded. In this setup it is also possible for camera noise to be picked up. If you use the mic that comes with the Sony PD150 mounted to the mic clamp on the camera you will probably pick up camera noise. The solution is to either use a better, more directional mic, or mount the microphone further from the camera. This can be done by getting a mic clip/holder that attaches to the mic shoe rather than using the built in clamp. See Microphone Accessories below and choose a clip which positions the mic as far from the camera as possible. Lastly using internal camera mics is not desirable as they are generally low quality and often pick up camera noise.
Setting Sound Levels (Clipping/Peaking)
A couple of times in this section we talked about not letting sound levels get too high. When the audio gets loud it can peak off the top of the scale. If this happens the loud part of the audio will be lost; this is referred to as clipping. This is very important because if sound clips it is unusable. Your camera should have a level meter and generally you will want to keep the average audio level in the top 1/3rd of the meter (when contributor is talking/situation is at its loudest). As you turn the audio down (when in manual audio) the audio will go less far up the level meter. If it is a digital meter (made up of bars ‘lighting up’ in real-time) the highest bar stays lit up for a second so you can see the highest level easily.
A good rule of thumb is to set the meters so the loudest noise you are likely to encounter goes almost to the top of the meter. The amount of headroom you leave depends on the situation. If you are in a controlled environment (i.e. an interview) you don’t need much. If you are filming an exited crowd (football match/riot) leave more. You will learn from experience but as a beginner leaving extra headroom (having a slightly lower signal) is a good idea to be on the safe side. Ultimately you should be aiming to set the levels at a reasonable height without clipping. Having a slightly lower level is always preferable than clipping. For the details of the meters and how to set levels, refer to your camera manual (if you don’t have one, do a search on the Internet for ‘manual camera model pdf’).
If you are recording 2 channels (i.e. stereo) but only have a single mono mic there is a handy trick. Record one channel normally as above and set the second channel slightly lower. Any peaks which push one channel into the red should be recorded cleanly on the quieter channel.
So why not just set a very low signal? All audio recording equipment produces noise in its circuits. This is known as the noise floor, which is the level of background noise present on a quiet channel, introduced by the circuitry of the mixer and cables. As you increase the level during recording this unwanted noise is NOT increased, only the sound you are recording is increased. A recording with a low signal can be amplified during the edit to a certain extent, without hearing the noise floor. If the sound is too low boosting it in the edit will amplify the noise floor to the point that it becomes noticeable. Bear in mind you want the level of what you are recording to be well above the level of the noise floor, the higher the level you record the further above the noise floor your recording will be.
Get Close and Personal (M2 rule)
It seems logical that the closer you get to the source of the sound the louder the sound (i.e. there will be less interference from background noise). This is indeed true but is even more important than you may first think.
As with lighting, the M2 rule applies for sound. Halving the distance between mic and the sound source quadruples the usability of the sound, and the inverse is true. As you move away from the source, the sound becomes exponentially less usable. By usability I mean less unwanted background noise. The closer you get to the sound source the less background noise you get relative to the sound you are recording.
There are a number of things to consider when choosing a mic. The majority of this section talks about mono mics with a section on Stereo mics at the end.
Directional vs Omni-Directional
The first characteristic of a mic is how directional it is, and if it is directional what is the shape of the area it is designed to pick up the sound from (known as the mics ‘fields’). In video production the two most common are Super-Cardioid and Omni Directional. There is a discussion of other fields here.
Directional Mics (Cardioid)
- X-Y configuration is where there are two directional cardioid microphones close together, one pointing left and one right (at 90 degrees). This allows the two elements of the microphones to be close together (or in the same unit). This is convenient but is generally less “spacey” and has less depth compared to recordings employing using the below A-B setup.
- A-B Where two cardioid microphones are used spaced some distance apart (i.e. 50cm) pointing slightly outwards. The left/right separation is created through the time it takes sound waves to reach each microphone rather than the direction they are pointing. The ‘stereo width’ can be altered by moving the microphones further apart.
Separate Audio vs In Camera Audio
With ‘in camera’ sound the microphone is plugged straight into the camera (or into the camera via a mixer) and sound is recorded on the same media as the video (i.e. Memory Card). With ‘separate audio’ the microphone is plugged into a stand-alone audio recorder (again this can be done via a mixer).
For ‘separate audio’ analogue tape was originally used. In the digital world DAT (Digital Audio Tape) and hard disk systems were used. Recently there has been a move to ‘solid state’ recorders which record on memory cards of the type used in digital stills cameras (SD, SDHC, CompactFlash etc.). Prices vary, but Zoom do a range which include the budget Zoom H1 and the more professional Zoom H4n (pictured right), the latter has balanced mic level XLR inputs. Others to consider are the Edirol R-06 (pictured left) and the Sony M10. The Zoom H4n is generaly considers the best for XLR/Balanced line and the Sony M10 if you are using a stereo 3.5mm jack (non balanced input).
The advantages of recording audio separately are:
- Although the audio circuitry in consumer/prosumer grade camcorders can be good, specialist separate sound recorders tend to have better quality electronics optimised for recording audio. Therefore the quality of the audio tends to be higher. High end video equipment carry pro level inputs so this is not an issue.
- Ambient sound/wildtrack can be recorded separately by the sound recordist without the camera operator being involved. For example while shots are being setup.
- The sound recordist may also realise important dialogue is happening and it is possible to capture great dialogue when the contributors do not think they are being recorded as the camera is not running (obviously moral issues must be considered).
- There are no leads running between sound recordist and camera. This means that the sound recordist can more around more freely and there are no cables to get tangled up in/for people to trip over (less law suits).
- As well as the practicalities of having enough leads on set it also helps sound quality. The longer the leads the more interference you have (even with balanced line). Long leads often pick up some interference, this is normally not a problem as it is very low but it not ideal.
- Very long shots can be executed easily. For example you may wish to shoot from the roof of one high rise building onto the roof of another (i.e very long shot zooming in). This may even exceed the sensible range of radio microphones.
- Audio does not have to be synchronised afterwards. With separate sound the video and audio must be reintegrated so that they synch perfectly. This involves importing both audio and video into an editing program and (often manually) moving the audio track so it is exactly in synchronization with the video. A single frame out (a 25th of a second in UK video) causes problems.
- It makes recording run-and-gun without a sound recordist a lot easier. There is a lot to do when you are the only crew member and having also to start the separate recorder is often the straw that breaks the camels back. The sound recorder has to be mounted on the camera rig. Modern solid state recorders are small but still something else to hang off the camera that can get in the way.
A lot of the decent quality microphones require power. This will either be provided by batteries or Phantom Power. Phantom Power is a system where the camera/recorder delivers power to the microphone through the microphone lead (i.e. 3 pin XLR). This only works with balanced microphone lines. The power supply is applied to both signal lines identically, somewhat like interference. Therefore It is ignored by input circuitry which is looking for the difference in signal (see above). Despite being carried on the signal wires the power voltage “disappears” as far as recording equipment is concerned, hence “phantom” power. Phantom power draws current from the camera batteries, but not enough to significantly affect the battery life.
There is enough to go wrong in filmmaking without having to worry about microphone batteries dying at a crucial point, so phantom power is preferable if your camera can deliver it. Some of the cheaper ‘prosumer’ (consumer cameras with professional features) models don’t; generally if your camera has built in XLR sockets it will usually support Phantom Power but you should check with your camera manual. There are a few different Phantom Power voltages (12/48v) so you need to check that everything is compatible and set up correctly.
Leads and Adapters
Monitoring Sound, How and Why
Miking Up an Interview
The ideal setup for an interview is to place a lavelier on the contributor and use a directional mic on a boom pole/pistol grip. There are two reasons for using both. A lavelier will get the clearest sound but if anything brushes against it or there is clothing that rustles the sound will not be usable. If you are using a radio mic you may also get radio interference. If there is any problem with this you could use the boom mic sound instead, which will not be quite as crisp but if done properly will be sufficient. But booming when using the lavelier is not just as a backup. Mixing a little boom audio with the lavelier will give a richer more realistic sound as you will also have a little room tone. For an interview sound mix this is the icing on the cake.
Lavelier/Tie Clip Placement
A Lavelier should be clipped to the front of the contributor a little left or right of centre pointing downwards. It is pointing down as things are less likely to brush against it. Also if it’s pointing towards the mouth and the person breathes on it you can get an unwanted noise (popping). In drama it is necessary to hide the microphone either inside clothing, in a hat or behind a pen in a pocket etc. For documentary this is often not necessary but it does look better. The main issue is that if anything, even a hair, touches the capsule of the microphone you will get loud unwanted noise. When putting the microphone under a layer of clothing (e.g. a coat) you may pick up the clothes rustling or the clothes may touch the microphone. Try it out and get the contributor to move around. If you are not also booming its better to play safe and mic outside clothing. Miking up contributors with a lavelier can be a tricky business so a good sound recordist with knowledge of the dark arts is a real bonus.
Boom Mic Placement
Booming requires a boom pole and an appropriate microphone holder/clip. You may want to use a Softie or blimp. Often the boom will be held above the head with the microphone just above contributor off shot. You are trying to get the microphone as close as possible to the mouth and pointing at it. Coming in slightly to the front pointing the microphone down at 45 degrees works well (see picture to right). An alternative is to use a Pistol Grip to hold the microphone below frame and pointing up (See Holding Microphones). You will have to work with the camera operator to ensure the microphone is just off shot. Move it in and get them to tell you when it is in shot. You should also arrange a hand signal for them to use if you dip in shot. If doing run-and-gun this requires real coordination. The boom op must be able to judge when the microphone is in shot and also not get tangled up in the leads, even when the camera op re-frames. A real challenge! If you do not have a sound recordist it is still possible to use a lavalier and shotgun by putting the shotgun on a microphone stand but this is not ideal as moving the mic on the stand is often difficult during a shot.
Recording Gigs & Amplified Performances
The best way to record a gig is to either to use multi-track recording or have a second mixing desk (recording the mix to stereo) but this is beyond the scope of this guide and probably your budget. The more do-able solution is to record a 2 channel desk mix and ambient sound. If you talk nicely to the sound engineer (preferably beforehand) they should be able to arrange to record the mix from the desk onto your portable recorder (Solid State SD recorders are a good option). If you are on a tight budget you could use a late model mini-disc.
A decent line out signal from the desk will be necessary. Ask what connector the desk outputs and make sure you bring the correct leads/connectors (never assume the sound engineer doing you a favour will have the correct leads). You should also get another recording of the room/audience. The issue with desk mixes is that they are mixed for a room full of people so can come out a bit flat but they can sound good enough mixed with some ambient. The best way to do this is do a stereo recording at the back of the room. As the mixing desk is often at the back of the room this may be a good place to position the mic(s). This does mean you need 2 recorders and will get 2 recordings which you will need to synch up in edit. Mixing the desk mix and room recording will give you a richer more live sound.
If you can’t get a desk mix putting the microphone very close to the PA speakers can work but it does depend on the quality of the PA. This is very much an emergency option.
Wildtrack & Roomtone
Wildtrack is a general term for non-sync ambient sounds recorded without video. A recording of a person walking on a gravel path is a good example of sync (i.e. in sync) sound. You would normally record the audio at the same time as you video the person walking. The sound of the feet hitting the gravel would have to be in sync with the video of feet hitting the gravel. General environmental sounds like birds singing, motorway noise or roomtone are examples of (non-sync) wildtrack; they do not have to be recorded in synch with the video.
Roomtone is a special type of wildtrack. It is a great idea to record the ambient background sound of a location/room when no one is talking for a few minutes. Generally when an interview is edited bits are removed. Laying a track of roomtone under the whole interview (at a low level) will help to hide the cuts. It also allows you to add extra pauses which help you get the pacing correct. Without roomtone the gaps will be very noticeable by the total silence, It also adds texture to the sound design. If there was music edited in under the interviews, room tone may not be needed as music is another great way of hiding edits. However It is also still good to have the wildtrack recording so you still have the option of using roomtone when editing.
Recording location wildtrack is a good idea and it can be recorded at a different place or time. If you interview someone in a rather dull office environment you could go and get audio wildtrack of a busy office and add it to the mix later. This then gives the impression of a more dynamic environment. Obviously there are ethical issues here, as you don’t want to mislead the audience, but that’s a judgement call you have to make.
It is also common to add foley sound (e.g. footsteps recorded in a sound studio) in post production as well, but this may exceed the requirements of many productions and is beyond the scope of this location sound recording guide.
Controlling & Moving the Environment
ocation sound recording is not simply a case of making the most accurate recording of what exists. It is often a case of altering the environments or moving the contributors to get the best usable sound.
Controlling the environment includes reducing noise pollution from outside the immediate environment, such as closing doors and possibly windows. It may include asking people politely in other parts of the building if they can keep the noise down while you are filming. Let them know when you plan to finish and arranging to tell them when you have finished is vital.
Mobile phones are one of the biggest problems on set. Often if you ask people to turn them off they will simply set them to silent. This is no use as mobile phones transmit a burst of radio interference every 60 seconds. Some phones have airplane/flight mode. This stops them transmitting anything but does not totally power them down. You can tell people to turn their phones to offline but need to make sure they understand what you mean (it tends to work a little better than ‘phones off’).
Sometimes controlling the location is not enough, you may have to change it. Moving to an alternative (hopefully nearby) location may not degrade the picture and may massively improve the sound. For example you may be filming the outside of a building by a road; moving to the other side of the building away from the road will give you better sound and may not be to the detriment of the shot.
With the above in mind it is essential to consider sound when searching for and considering locations. Ideally, the sound recordist will be present during any reccie. If someone else is scouting for locations they should make comprehensive notes of sound pollution and make a quick recording. Recording next to a motorway on the flight path to Heathrow may seem obviously a bad idea from a sound perspective. I have however experienced a situation where the director had chosen such a location and did not realise this would cause major sound problems. Stopping the shoot for avoidable audio issues wastes time and money – it must be thought through by everyone before shooting.
As has been previously mentioned there are a lot of good reasons to record sound separately from video. If you are doing this you want to make it as easy as possible for the editor to sync up the audio and video.
Using a Clapperboard
Syncing Without a Clapperboard
Although clapperboards are still very useful for syncing (and can have scene/take written on them) they are not strictly necessary. Most camcorders now have built in mics or you can plug a cheap mic into them. This means that even when recording sound separately it is easy and cheap to record reference (‘scratch’) sound on the camcorder. It is still necessary to mark each shot (snapping your fingers in shot is a favourite) but you can just clap your hands in shot to mark each take. Saying something to indicate which shot is which will help everyone later. The editor will then look at the sound wave of the separate audio and the scratch audio, find the mark and sync just by looking at the shape of the two waveforms. Modern edit software tends to be able to automatically sync using scratch audio so it is becoming increasingly essential.
When recording for long durations the audio can move out of sync as the speed the video and audio runs may not be the same. For analog this is related to the exact speed of the motor moving the tape. It is not as much a problem for digital but does still happen so it is still good to always have camera scratch audio. Even if you are using a clapperboard having scratch audio is very useful. It will make syncing easier and help to sort out sync drift (some edit software may automatically sort this out for you).
Which Microphones to Get?
Most Rode, Sennheiser, Beyerdynamic and Audio Technica mics are generally very good. Out of these a good place to start is with Rode who offer great value for money. For shotgun mics the Sennheiser MKE 416 is the industry standard. It is exceptional but expensive (> £500). It is also very directional, maybe too directional for some situations. A good budget directional microphone is the Rode VideoMic. It is not balanced line and the build quality is very plasticy but the sound is good. Rode also do excellent traditional XLR shotgun microphones (NT1, NT2 and NT3; get the best you can afford).
Lavelier mics tend to be less expensive. There is an Audio Technica ART-3350 which is under £20, it produces OK sound but not broadcast. It is also not phantom power so can be used with cheaper camcorders. The Rode lavelier is also worth considering at around £170 and the Sanken COS-11 is one of the most respected mics but is more expensive (£250). The ones that come with Sennheiser radio mics are good and can be picked up for around £50 (a great option as they can be used for broadcast).
Bear in mind that although a lot of laveliers have a 3 pin stereo jack they cannot be connected to stereo mic inputs on camcorders or the cheaper solid state recorders. The 3 pin jack is actually a special type of phantom power connection specifically designed for lavaliers. It is physically the same as a stereo 3.5mm jack. Some Solid State recorders provide 5V phantom power and work with these but you should check in the manual. A cheapish configuration is a Edirol R-09 and a Sennheiser Lavalier (they both support 5v phantom power). All the lavaliers apart from the Audio Technica mentioned in the previous paragraph are 5V phantom powered. It is however possible to get leads and connectors to plug these into the balanced 3 pin XLR sockets that are used in good prosumer camcorders.
Choosing a microphone is a far from a simple task. Do your own research and attempt to get to try before you buy. Discussing this on Internet forums is useful but make sure you have done some research on the net first, reading a few reviews. Including links to some of these reviews in your question will show you have done some groundwork and will make it a lot more likely that people will spend the time to help you.
Please feel free to suggest additions to this section but if you do please include links to favourable reviews. We don’t have the budget to get all book on sound or time to read them. lazy I know but this is but it is a free guide.
Add Web Articles
- The Ten Commandments of Sound for Picture, part one and part two – Makes interesting and Entertaining reading.
- An Open letter from the Sound Department – This letter is being written by audio professionals to help directors and producers understand how good sound can be recorded on the set. We want to help you make the best film possible.