By far the biggest technical shortcoming of many videos and independent films is the sound. The dialogue recorded on location, which in most cases is the project’s only source of dialogue in post-production, is not always recorded with correct technique, which results in poor audio quality. Location sound recording is a huge topic, but I will attempt to cover the salient points based on my past experience.
To record high-quality location sound the right type of microphone must be used: ultra-directional for external locations, directional (shorter) for interiors, and non-directional for cramped interiors.
The more directional the microphone, the greater the extent to which it selectively picks up sounds from its front end, and the higher the signal-to-noise ratio will be; but bear in mind that excessively directional microphones will pick up too much echo in tight interior locations,so the right compromise must be struck in those situations. The Audio-Technica BP4073 Lightweight Shotgun Microphoneis an outstanding dialogue microphone – the favorite of many professionals. I have personally used it and it produces truly awesome location sound, if used correctly.
Having picked the right microphone, the key to recording rich, clean, high-quality location dialogue is to place the microphone as close as possible to the subject. In this way the actor’s voice will be much louder than the background noise. The microphone should also ideally be overhead, pointing downwards at the actor’s mouth. The second best choice is to place the microphone below the bottom frame edge, with the microphone pointing upwards at the actor’s mouth.
The best way to do this is to make the actor(s) get into position, frame them, instruct the boom operator to dip the microphone into the frame, lift it out until it is just out of frame, set the level using a sample line from the actors, and roll!
If you use this technique consistently I guarantee that all you will have to do in post-production to obtain professional, clean-sounding dialogue is set a consistent sound level across the movie, patch up the occasional extraneous noise with clean sound from another take and place the audio cuts wisely (more about this later).
With this strict technique the microphone will occasionally dip into the frame, requiring a re-take of at least that part of the shot, but the amazing sound quality will more than compensate for that. I shot a 30-minute movie in this way and although my insistence on this technique made for a demanding shoot, the resulting sound was really worth it. I was very glad I insisted on it.
You should definitely do several takes of each scene, regardless of how well the actors are performing, because in this way the entire scene will be covered with good sound, even if no single take was flawless from beginning to end (they rarely are). If there are cars driving past outside, this will definitely be the case. With multiple takes, finding a clean version of any one section will be a breeze. You can (and should) listen to the sound during and after each take on location, but I guarantee there will be little annoying sounds that you missed in the hustle and bustle of shooting. That’s when you’ll be glad you did several takes.
I once shot in an interior location where the ancient heating system would occasionally emit a diminutive but very annoying “ding”. Not only was this not noticeable while shooting – it also occasionally overlapped with words, which was subtle but annoying. I was able to produce a truly flawless soundtrack in post-production by replacing the word or words that overlapped with the noise with clean versions from other takes – all done manually with Adobe Premiere. You’d be amazed at how seamless you can make your soundtrack by using such techniques. Time-consuming maybe, but immensely worthwhile.
If the extraneous noise, like a car passing by, happens in between lines of dialogue, it’s much easier: just splice out the noise and replace it with clean ambient sound. Which brings me on to the next sound recording tip:
For every location and every shot in that location, be sure to record at least 30 seconds of ambient sound. That means shutting everyone up and recording 30 seconds of silence. Of course it’s not real silence; it is ambient sound, and the “silence” will be different for every location and every setup in that location (because the loudness of ambient sound depends in part on how close the mic is the actors, which in turn depends on how the shot is framed).
Make sure the boom operator does not move his/her hands along the boom pole during takes, as the sound will be conducted by the pole to the mic and will produce unacceptable noise, notwithstanding the shock mount that you should be using. The shock mount is a contraption that holds the microphone in a web of elastic bands – this is to insulate it as much as possible from vibrations traveling along the boom pole as the boom operator moves. Professional microphones are incredibly sensitive! Taking rings off is also a good idea.
Whatever you do, remember that the surest way to make your production sound hopelessly amateurish is to record sound with an onboard mic (a microphone mounted on the camera). It will be an easier shoot but your audience will hate the poor sound.
Being attached to the camera, the mic will almost always be far away from the actors, resulting in noisy, echo-ridden dialogue, which will relegate your project to the amateur category. Get that microphone as close to the subjects’ mouths as possible!
You could of course use an onboard mic for the entire shoot and replace this poor location sound (the “scratch track”) with clean dialogue re-recorded in a studio, but this is difficult to pull off, time-consuming and expensive. Even the biggest movies use location sound as much as possible, only resorting to ADR (automatic dialogue replacement) when it’s absolutely necessary. There simply isn’t anything quite as good as well-recorded location sound.
If you are recording digital audio, regardless of whether it’s a DAT recorderor digital video, it is important to set the levels lower than you would with an analogue device (such as the Nagra). This is because with digital audio, over-modulating the sound recording level produces intolerably ugly distortion. In digital sound, the transition from “intense signal” to “distorted signal” is sudden and unacceptable – just as with the exposure levels of video itself (see digital cinematography tips).
To set digital audio recording levels correctly, before rolling ask the actor to give a sample of the loudest line in the shot, then give yourself a good 6dB of headroom above that. This is good practice with digital audio recording.
For a truly seamless soundtrack, you will also have to disguise the sound cuts by not aligning them with the video cuts — in other words, the video and audio cuts should be staggered.
The reason for which this is necessary is that the quality of the ambient sound may vary between setups and throughout the shooting day, and this change in background sounds will be audible when you cut from one shot to another. You can disguise the cuts by making them coincide with a louder sound (such as someone beginning to talk). I once shot in a location where it was quiet in the morning but full of loud birds in the afternoon. Making those shots match required a lot of creative editing, but it worked.
It is worth giving high priority to the quality of the sound you record for your projects. Remember that an audience can forgive imperfect camerawork if the subject is compelling; what they will never accept is poor sound.