Recording Sound for Video – A Definitive Guide
We thought it was time to post a technical article for those that may be interested or would like to know a little more about how we go about recording sound with video.
This is often quite daunting for newcomers or those who are just starting with the technique of ‘double heading’ when shooting on DSLRs or similar camera technology. By the way ‘Double Heading’ is simply an industry term for simultaneously recording video and sound on separate systems i.e. a camera and a sound recorder.
So why is it necessary to do this? Well the technique dates back to the earliest days of sound with film. Obviously film cameras were (usually) incapable of recording sound, (until the invention of optical recording and magnetic edge ‘stripes’) so a separate audio recorder was required on location. The arrival of sound in film also required some form of ‘slating’ to effectively mark the starting point of a take in both vision and sound simultaneously and so, the clapperboard was born.
Today many film makers continue to use double heading techniques and it is still common practice on productions using state-of-the-art camera technology that is equipped with high quality integrated audio recording. In this case filmmakers will ‘go double headed’ for audio backup purposes.
In this article we are considering the importance of double heading with the perennially popular DSLRs. These cameras are very common on many, if not all productions intended primarily for web distribution and they can provide extremely high quality results.
However, many serious DSLR shooters will need to consider double heading as the audio recording facilities on these cameras are somewhat less than ideal. They offer minimal control, quite low quality input pre-amps and non-pro connectivity and there is no phantom power for capacitor mics.
Kit Items Required
The Zoom H4N Recorder
Today there are many portable, digital sound recorders available that have been specifically designed for this purpose. The Zoom H4N has been a very popular model for at least four years.
The recorder needs to offer ‘balanced’ mic inputs on three pin XLR connectors. It should be able to provide 48V phantom power. Adjustable bit depth and sample rate is important too and the latter is particular important here as it relates directly to sync locking. The industry standard sample rate for sound with video is 48KHz. Many sound recordists will want the option to record in uncompressed WAV format and even ‘broadcast WAV’ which embeds a lot more meta data such as SMPTE timecode.
The rode NTG1 is a very popular gun mic
At least two good radio mic sets will be a real boon to any filmmaker. They are often an essential item on our own shoots.
A good riffle mic for ambiance, wild track and fill recoding. There are many available ranging from a few hundred pounds to many thousand.
Sound Devices 302 mixer – a very high quality and popular machine.
This is probably one of the more contentious items and can often be one of the most expensive items in the audio kit. However, many pros will want to work with a field mixer to provide the best quality ‘front end’ for all of their input sources. It also increases the number of possible input sources (or mics) that can be connected to your two channel audio recorder.
This is usually where most inexperienced people will make mistakes. If you are running a complex system with mixer and recorder it is very important to know how to ‘line up’ all of the gear in your audio chain and manage its overall ‘gain profile’ correctly in order to avoid clipping and excessive levels.
The industry standard (BBC) for broadcast safe audio stipulates a maximum peak level of -10dB on the new dBFS scale, which is common to all digital equipment. Line up is carried out such that an analogue test tone at 0dB from the mixer is correlated with -18dBFS on your recorders’ meters. This ‘profile’ will give you loads of safe ‘head room’ thus reducing the risk of clipping on unexpected peaks and transients.
Don’t forget to make use of ‘limiters’ if you can – but try to avoid dynamic compression, as this cannot be undone in post.
It is important to remember to record wild track. This is basically a recording of the natural ambiance of a setting or environment, which is then laid up with your other sound track components such as dialogue, which may have been recorded at very close range. Wild track adds depth and naturalistic ambiance to your audio and it is an essential part of a drama or feature film soundtrack. Remember that even when recording in an environment which is ostensibly silent such as a still open landscape you will still need to capture ‘air noise’. The renowned Manchester Recording Engineer Martin Hannet referred to this as a ‘recording of silence’.
Record encoding: WAV 16bit at 48KHz
Peak Level at approx. -10dBFS
Line up 0dB to -18dBFS if using old analogue mixers.
0dB should correspond to number 4 on the old BBC style PPM meter (white numbers on a black background).
Always record wild track
Low ‘noise-floor’ margin: BBC practice stipulates that minimum audio levels should just be above the system noise floor. On a high sensitivity dBFS meter you should be able to calibrate for this as system noise will register typically in the -30dB range.