Your camera's on-board mic might be OK for home movies, but using a separate mic and audio recorder will yield much better results.
Synchronising sound with picture really couldn't be easier, could it? Just point the lens of your camcorder at whatever it is you are shooting, and its built‑in microphones will face in the same direction. Everything's recorded onto the same tape, disk, or removable media and the job is done. If you're shooting for YouTube that's probably good enough, as standards there are generally low. No, strike that, those standards are generally abysmal! I think we can aspire to better, and it isn't too difficult to achieve it.
It has always been the case both in cinema and television that sound has played second fiddle to picture, and while I'm a sound person at heart, I can see why this is the case. Since this is Sound On Sound magazine we ought to make a stand and say that sound is fully the equal of the picture, and perhaps on occasions more important. George Lucas famously said that "Sound is 50 percent of the movie‑going experience, and I've always believed audiences are moved and excited by what they hear in my movies at least as much as by what they see.” You'll be glad to know that it isn't all that hard to record great sound for your own pieces.
So what's wrong with recording sound through the camera's microphones? The camera operator (likely to be you) will position and set up the camera the lens for the best image. By 'best image', I mean either the most artistic composition, or the image that tells the story best. When you have selected your vantage point and composed the shot, your camcorder will be in the best position to capture a picture, but it probably won't be in the best position to capture sound.
Most of the time, to capture the best sound (the best subjective quality or the best for telling the story) you would need the microphones to be a lot closer, or be more directional than those on your camcorder. It is also worth saying that a camcorder can have moving parts, like tape or an internal hard disk, as well as plenty of complex electronics that can generate noise and interference that can be picked up by the internal microphones. A lot of people might tolerate this, but I'd say it's fine for nothing more than home movies.
The first step to take towards recording good sound is to use an external microphone and plug it into your camcorder's microphone input, assuming it has one of course! If it has a simple 3.5mm mini-jack input, you can't use a capacitor microphone unless it is battery‑operated or you have a separate phantom power supply. Oh, and you won't be able to use two microphones unless you use a splitter cable or a mixer; nor can you use three or more without the latter, as you'll need to mix the sound down to the two channels available on the jack plug.
Mini‑jack inputs are quite insecure, as they don't latch at all. For me, the worry that a connector might be pulled out or not make good contact is significant. I would want to monitor on headphones at all times to be sure — but that would disconnect me from the talent I'm directing. If you're lucky enough to own a camera with XLR input or two then connections will be more secure, plus you'll be able to use a mic that requires phantom power, opening up some more equipment options over the 3.5mm jack setup.
Either way, if you're plugging your mics straight into a camera, you'd best hope it has decent preamps. I don't mean that in the sense of a high‑end 'boutique' preamp versus the preamp in a standard audio interface, where the differences tend to be subtle. Some cameras really do have bad preamps, in that they are very noisy or have very little headroom, and you'll hear the difference between these and the preamps on an external recorder.
Often, the microphone inputs on cameras have to do double duty as line inputs, and there is no way they can be optimised for both functions. Sometimes they are the victim of poor compression known as AGC (automatic gain control), and on some models this can't be deactivated. When you think about the cost of the technology in a camera, it's easy to see why preamps are an afterthought.
Not so long ago I was making a video using a vintage Neumann U47 microphone for the audio, which I'll talk about later. To have plugged that directly, via its power supply, into the camcorder would have been a bad idea!
Audio metering in cameras is often poor. Since you're a Sound On Sound reader, you won't want to use the automatic gain control, meaning you'll have to keep an eye on the sound level using whatever display the camcorder gives you, probably on the LCD screen. That isn't going to end in love, peace and happiness. It's going to end in something nasty happening to either the picture or the sound because your attention is divided. And if you do have a separate sound recordist, do you really want them trailing after the camera operator all the time to keep the meters in view? That's not the way to put sound on an equal footing to the picture.
There's also an issue with mic positioning if you're recording straight to the camera, in that sometimes you'll want to get closer to a subject's voice. You can't position the mic in shot (unless you're shooting a documentary‑style production where it's appropriate for the talent to wear a miniature microphone), so you may want a separate crew member to hold a 'boom pole', which allows the mic to be held just out of view.
At this point I think you probably realise that I prefer to record sound separately to picture. I'm not saying that I will never record sound to the camera, just that recording sound separately has great advantages that I nearly always prefer. So, how is it done?
In order to record sound separately, you'll need an audio recorder of some kind. This could be a model with built‑in mics or XLR inputs; either will probably better the recording quality of your camera. Using a separate recorder will allow you to plug in a directional shotgun mic, which can be mounted on a boom pole in order for a sound person to get closer to the action. Even if you aren't using a sound person, having a separate recorder capturing the sound via a camera‑mounted shotgun mic is usually a step up from using on‑camera sound.
While it brings with it better preamps and greater control over metering, gain and recording format, the use of a separate sound recorder also requires that you find a way to synchronise the sound and picture in the edit. Modern digital technology is pretty good at keeping time, but just to make sure, a global timecode (see box on previous page) will be used on bigger‑budget shoots to synchronise all cameras and recorders on set to a single clock. You probably won't need timecode at a semi‑professional or enthusiast level unless you expect your camera and sound recorder to run at exactly the same speed for a long period of time. With less expensive gear there will usually be slight deviations after a long time, but they're easy to correct for after the fact in your non-linear editor.
It's easiest to explain how to get simple sound sync using external audio if I run through my thought processes for a typical shoot. When I make a video, I'm usually using it to explain a certain point about audio. Recently I made a film demonstrating the audio qualities of various microphones. I say "made a film” rather than "shot a video” because I used to live next door to Ridley Scott (really) and some of his genius rubbed off. Well, it didn't, but I like to pretend so.
For my film (of which I'll call myself the director) I had the services of a singer, an assistant camera operator, a recording engineer and a mighty sound studio with several of their most excellent microphones at the ready. I was operating two cameras myself, both on tripods.
When I originally conceived the project, it was going to be audio‑only, the reason being that the audio had to be absolutely first-class and I thought that video would be too much of a distraction for me.
The opportunity of having pictures to go with the audio, however, seemed too good to miss, especially since promotional visuals are becoming more and more important. I considered hiring a videographer, but I didn't at the time know anyone reliable and I am well aware from personal experience that there are many dodgy operators out there among the good ones! So I decided to handle the video production myself while the audio was recorded into Pro Tools by the engineer.
I decided to use two cameras on tripods that I could pretty much set and forget, and a third camera on a monopod, operated by an assistant. Call me old‑fashioned but I like my images as steady as possible, and the monopod is an excellent compromise between steadiness and mobility, as well as being much less expensive than a Steadicam.
The reason there were three cameras rather than two is that directing is a lot to think about, and I wanted to cover many angles without having to think too much until the edit. There's a lot that can go wrong, particularly when you're concentrating on getting the sound perfect. Having three cameras means you're covering yourself twice over, and it is nearly always my preferred way of working outside of the very most controlled conditions.
So there we are in the studio: me, the singer, the camera assistant, and the sound engineer in the control room. All we have to do is start the cameras, roll sound, and roll the backing track. The singer sings, it all gets recorded, job done. Well, there is a little problem: the three cameras are not synchronised in any way, and neither is the sound. Everything is 'wild', to use a technical term. So considering that over the course of a day there would be dozens of shots (including cutaway material that I was shooting too), putting it all together again could be been a nightmare.
Fortunately, the solution for synchronising all these devices is simple. I can't say I invented it, but someone must have done in the past, and hearty congratulations to whoever it was! To synchronise three cameras and sound, all you need is a clapperboard. If you don't have one, you don't really need the 'perboard' part: all you need is a simple handclap, captured on all three cameras and all of the mics. Suppose you have a one camera, or any number of cameras, and a separate sound recording system: all you have to do is put everything into record, confirm that all is rolling, and clap once in view of all the cameras and range of the recorder.
The important part here is "confirm that all is rolling and clap once”. Here are two things that, if not done properly, can lead to difficulty or disaster. If you are the person in charge of the production, you will be focused on working the camera. Either someone else will be working the sound, or you will have allocated a small part of your brain to that function. You'll know the camera is running because you pressed the button and waited till you saw the red light in the LCD screen. In my case, I would know that camera two was running because I was in charge of that too. I could see my assistant 'turning over' camera three. But what about sound? Sound tends to feel distant — in my case, it was being recorded in a different room. Confirmation that sound is rolling is essential.
If you look into the history of film production you will find that there are certain conventions. These make filming easier and less prone to difficulties. So just before the director or assistant director calls "action”, the assistant director (or the second assistant) will shout "turnover”, whereupon the sound recordist will start his machine and reply "speed” when it is running and recording. The camera operator will start the camera (film is expensive, so best not to waste it) and call "mark” for the clapperboard to be clapped in front of the lens. When you work with a sound recordist, you have to do exactly this. When you have started the cameras, call "turnover” and wait for the sound recordist's reply: "speed”.
I can tell you from quite a number of experiences that people find it rather uncomfortable pretending they are in Hollywood and using all this fancy language. So I instruct the sound recordist in what is required, then I call "roll sound” and expect the reply "sound rolling”. I would find it far too pretentious to call "action”, so I start things going for the talent with "cue”, like they do on TV. It's also worth bearing in mind what you need to do at the end of the take. Sometimes you will want to go again without stopping the cameras or sound; sometimes you will have come to a natural break in proceedings and you need the cameras and sound to stop. You're the director, so you're in charge: decide in advance what your language will be and make sure everyone else knows.
My other key point was that you should clap once and only once, each time you start recording. Suppose you clap, then wonder whether you were in view of camera two, so you clap again? Then you will have two claps on picture and sound. You'll spend a lot of time in editing figuring out which is the right one. You will also benefit from developing a style of clapping that reveals your hand movements clearly to the camera. As silly as it might sound, practice at home on camera before the shoot.
So you're back at base, ready to load your pictures and sound into your favourite editing system, and you have lots of written notes so you know exactly what's what. You might even have commented on camera and sound what each take was all about, a procedure known as 'slating'. Let's suppose that you have already loaded two or three camera 'reels' and your sound 'reel' into the system, and you are going to work on the first take.
My editing software, like most, allows multiple video tracks, so I put all of the picture and sound reels onto the timeline. The next thing I'll do is 'solo' a camera and find the handclap. I'll find the point where the hands meet and cut there. This point is the beginning of that clip. I'll do that to the other picture tracks, then to the sound, which obviously has to be done by listening or by looking at the waveform.
I'm listening for the initial transient of the clap, and I cut there. Once that is done, it is a simple matter to align all the tracks. When I've done that, I can hit the space bar and all will play in perfect sync. I'll scroll to the end of the take and make sure that sync is still OK. If it is, I can start editing. But this sounds so simple: can anything go wrong?
There are three problems that can easily crop up. The first is that a clap is missing. If your mind is on many things at the shoot, missing a couple of claps could easily happen. Another is that over a long take, sync slips because the various machines don't run at precisely the same rate. Thirdly, if you're using tape (though it's becoming rare quite quickly) there may be a 'dropout' and sync will slip. I've heard that you can get dropouts even on solid‑state media, but I have yet to enjoy that experience.
Losing sync isn't a disaster, it will just take a little longer to get the job done. To solve it you have to find something that you can identify easily right down to the individual frame, in both picture and sound. I've found that there's nothing more reliable than the phoneme 'b'. Find a 'b' (or 'p') in a headshot and align the sound to that. The reason this works is that the exact 'b' occupies no more than a single frame of video. If you can align the sound this accurately, most viewers will be satisfied.
It's worth saying that in professional film or television, sound tends to be sync'ed to an accuracy of a half or quarter frame. But computer video has taught the masses to be tolerant of relatively poor sync, so whole‑frame accuracy is pretty good going. If you can't find a 'b' or any other obvious feature, then it's going to be a matter of trial and error to find the best sync. It can be done, but you'll long for that missing clap, so the best solution is to get it right on set!
If you like your sound to be as good as your pictures, it deserves its own special treatment. Given that you can link the two with a simple handclap that costs nothing, what reason is there not to do so?
The method of simple sound sync covered here works well, but there might come a point where you aspire to something even more accurate, and that is timecode. The problem with a simple clap is that it identifies a point in time, but after that the camera and sound recording systems go their own sweet way, which can result in loss of sync. Timecode fulfills the function of the clap each and every frame, with an 'address' that uniquely identifies a frame of video and its corresponding 'frame' of sound. It also contains a 'clock' that keeps the rate of flow tightly aligned. When the two are played back in such a way that their addresses match and their clocks run at the same speed, perfect sync is achieved and maintained throughout the full duration of the recording. Timecode can be quite complex to deal with, especially when mixing formats, so we won't go into too much detail in this introductory article. Here timecode is shown on a high-end Denecke clapperboard.
When picking an external digital recorder for audio there are quite a range of devices available for all budgets. Due to the relatively low size of uncompressed audio files (certainly compared to uncompressed video) as well the high capacity and low cost of solid-state storage, most recorders either use removable solid-state media or internal flash chips to store data.
The sound of budget video cameras can be improved by something as simple as the Zoom H1, a tiny solid-state recorder with an X/Y stereo mic setup on the front. The H1 records to Micro SDHC cards, and using the simple sync methods described here, will capture much better audio than a typical DSLR or hybrid camera.
Recorders with XLR inputs range from two-channel affairs like the Tascam DR100 through the Edirol R44 and R4 (complete with timecode) right up to devices such as the Sound Devices 744T and the Nagra V and VI. A relatively inexpensive two-channel device will give you the option of two dialogue mics, while a four-channel device will allow for two dialogue mics and a stereo ambience track for capturing atmosphere and background sounds.
Sound Devices and Nagra gear competes with high-end studio preamps in terms of recording clarity and build quality, and such recorders and mixers are often found on the sets of high‑budget films. Such recorders will set you back thousands, but can give you some studio quality in the field! J G Harding