Wow. Just wow.
I’ve never been involved in any formal sound training, and have never really been involved in sound production, so I’m not going to be condescending and say, ‘look what I’ve not been doing’ as I can honestly claim innocence of everything regarding proper sound. But at the same time, taking an abstract and global perspective, the evening turned out to be much as I expected – simply a stunning example of what is overlooked by pretty much everyone who is making a film from home movies, through amateur film clubs to film school students.
Unlike last week, we had quite a few film-club members, professionals and film students in the audience and all of them kept their mouths firmly shut in suggesting things that they believed were true.
I’m not even going to suggest that this article is a true representation of what I learned from this session – there was just so much to cram into a three hour session that we overran by more than half an hour even with Roland Heap, our tutor going at near breakneck speed.
I have a 25 page copy of Roland’s slides, and every one is covered in scribbled notes. I’m only going to heavily summarise the evening here, and in an order which makes sense to the article, not the original presentation, but I would suggest that anyone seriously looking at producing a short film from the club should seriously look at attending this session – it can be taken independently (as with all of the sessions) of the certificate for about £40+vat, and is certainly worth every penny.
Sound in theatres
While Roland was not totally sure of agreeing with the exact numerical value of this statement, someone once said that the film and video experience is actually 50% sound. Indeed, this session is billed with this header. What is certainly true however, is that with a few rare cases, a film is perfectly understandable if you remove the image, but not the other way around.
But first, lets break a myth.
In a theatre, whether 5.1 dolby or even today 7.1, a good 90% of the sound is actually mono, and generated from the single main speaker behind the screen. The reasons for this are simple – put the sound anywhere else and first of all it’s not coming from where you expect it to come from (i.e. the voice of an actor) and secondly depending upon where you are in the theatre a multiple-speaker setup will vary the experience and make it seem unreal. Everything that matters (dialogue, foley etc.) comes from this same single speaker and in mono.
The additional speakers (in 5.1, two either side of screen, and many on either side of the room which collectively represent a single sound source on each wall) are used for additional sound such as music and some sound effects which are used selectively. The final speaker (the .1 in 5.1) is the sub-woofer which generates the ‘vibration’ sound – the one that makes you feel the sound. This speaker is 1/10 of the range of other speakers (hence the term, .1) and is rarely rated above 100hz.
The exception to the rule (did we expect anything less?) are the French.
What was clear from Roland’s talk is that sound, far more than film, is regionally dependant. The French, more than all other film making nations, play with movement of sound between stereo speakers a lot.
Sound in TV etc.
The same is not totally true of TV and any shoots which are specifically intended for small-space presentation. This is purely because the distances involved mean that important sound (i.e. dialogue) coming from multiple speakers is not perceivable by the human ear and therefore can not cause confusion (this is of course, unless you watch your TV sideways with one ear pressed against one speaker on the side your set).
In these cases, sound may well be designed to move between speakers to take advantage of home stereo systems.
What is good sound?
One big question that was asked by several of the audience, even as we started, was ‘what is good sound’ – quoting people who have seen productions that they have been involved with as being good or bad sound. While bad sound is usually obvious, the question in some cases was, ‘why can’t I tell that something is good’. Roland’s view was that this is often bogus – there is good sound design (often relating to content), but no real range of ‘good’.
This one to me was perfectly understandable as the same rule is true of my own CGI world.
Good sound is invisible to the experience. Good sound is sound that furthers the story without the audience really understanding that it is there. The only way that you can really perceive good sound, is by not being able to listen to it and say, ‘wow, that’s bad sound’.
Roland made one really important point very quickly in the evening – getting good sound is not difficult to achieve (hire a good sound man, i.e. ME, at BIC2 rates, he then joked). Bad sound is usually the result of breaking certain rules – usually as a result of simply not caring and being content to live in just plain ignorance (my words, not his). Most amateur film makers get hot for a new camera, but few really think about getting good sound.
A good rule stated was that regardless of the budget for your film (excluding huge productions which may have specific budgets for particular things, like CGI, which skew the equation) you should always expect to pay 10% for the sound – this applies all the way to £4 mil films, where you would expect to dedicate £400,000 just for sound production.
A second golden rule is one that sounds obvious.
You’ll all read this and think, yeah, Good Lord, how can anyone be so stupid…but at the same time it is something which Roland admitted is the biggest single failure in sound on amateur and independent films.
What is it?
Simple: Make sure the microphone is actually pointing at what ever it is you want to record.
Some things to think about in regards to pre-production; things that are not done, but are as important for the sound as they are for the image:
- If you are hiring, a good sound man will go for £250 a day, but will work for significantly less is the film is pitched to them properly. This is sometimes very cost effective as sound is very different to other crew: expensive, specialised and often customised equipment is usually owned by the sound crew and will come as part of the deal (e.g. no hire costs). Hire a good sound man, and lots of important planning will be done for you. An important point made however, which is applicable to other crew as well: feed them.
- Sound is a very different skill to Music. Everything, from skills to equipment is totally different and there are very few people who can cross back and forth. In fact, the only similarities are the cables that link the microphones to the recorders (side note: Roland is ex-Abbey Road, so if anyone has the authority to make a statement like that, it’s probably him).
- Read the script intensely and look for everything connected with sound – either actual production of, or sound from external factors – and plan for it.
- Recce everywhere a scene will be recorded (interior and exterior), at the same time and day when filming will take place. Sometimes, problems can be temporally relevant, such as a school run at a certain time of the day, or children in a park during a key duration etc.
- For anywhere where there are hard surfaces, plan the availability of sound baffles (e.g. soft furnishings, mattresses, duvets etc.) to absorb sound. Worst places are rooms with hard parallel walls and waterproof surfaces – top of the list for example, bathroom scenes (line one wall, out of sight of the camera, with a big mattress).
- Don’t just look for sound issues, look for physical issues as well – low ceilings for example, means that the boom man is going to have issues.
- Plan what microphone is best for each shot (this varies hugely, see later). Roland gave one very golden rule which he could not stress enough – don’t buy your good equipment, rent it. An exceptional microphone, can be rented for as little as £5 a day including all of the carrying cases, fittings etc. Only ever buy if absolutely necessary, or it’s very long term use (even renting for a month can come out cheaper). This is particularly true if you need different microphones for different shots.
Production (i.e. Recording the sound)
There are some golden rules when it comes to actually recording the sound, and there are lots of other tips that go along with them. The important part is that if you break any of these rules, or are otherwise lax about them, your final sound will suffer.
- Make sure that the microphone is pointed at the subject making the sound at any moment. This is where an experienced boom operator comes in – not just in ensuring that the microphone (particular important if it’s a very tight directional shotgun mic in use) is pointed at the mouth of the actor (not the forehead, or chest…) but that he knows how the dialogue switches from person to person (i.e. he’s actually read the script).
- There is no such thing as too many microphones on the set – it just adds to the complexity of the final mixing but at the same time greatly enhancing the options available.
- Unless you are deliberately using VO (voice over), avoid ADR (additional dialogue recording) like the plague and record everything from the set, location etc., especially at amateur and independent levels. While you may not have experienced actors to repeat dialogue effectively off set etc., the biggest problem is cost. Additionally, ADR can actually ruin a previously good performance. Conversely, some US productions only use ADR, re-recording everything and using very little from the location recordings.
- If you must record ADR, especially if you are using it for VO, do so in a 100% sound-dead room. For amateur film-making for example, line the walls of the smallest room in your house with mattresses and duvets and record in there.
- When recording, try to record dialogue as best as is feasibly possible, and drop everything else out (a final recording like this, as used in demonstration during the session sounds very, very strange). There is more on why a little later, but one simple fact in play here is that you can add any sounds that you want to a recording, but it’s nigh on impossible to remove them – the killer being any form of reverb (i.e. reverberation, a more complicated version of echo). If you have any reverb on your recording, you may as well record it all again.
- Other extremely bad sounds that you certainly do not want to record (but add later in post) include music, waterfalls and traffic. In the case of music (and other similar noise, like crowd scenes and car-alarms) they represent continuity nightmares. In the case of things like waterfalls and traffic, these are very broad spectrum sounds that are impossible to identify using sound filters – it is therefore impossible to get rid of them. An additional problem with traffic, and other similar sounds, is that they are period dependant – 2012 traffic sounds totally different to 1940s traffic, which is totally different to 1900s traffic etc. In “The King’s Speech” they had to line the exterior walls of their buildings with sound-dampeners to prevent this type of modern noise reaching the location shooting.
- Any slight problem that you hear but decide is acceptable under the circumstances will sound 10 to 100 times worse when it’s recorded – kill as many extraneous sounds as possible, and especially electrical hums. The human brain makes lots of allowances, coupled with the fact the microphones record (at a technical level) in a very different manner to the human ear.
- Record wild sounds (a wild recording is background sound of a set or location) everywhere, and in different situations (e.g. record 60 seconds of empty room, another with the birds chirping outside, background crowd scenes if in a pub etc.). Not only do you use these to bridge sound cuts (there is always some sound going on, even if you are only aware of it subconsciously), but it also allows you to use foley (see later) recorded at the scene and is therefore more natural. Avoid sound libraries as much as possible.
- Record wild sounds of any continuity breaking audio, such as traffic / traffic events (no matter how minor). Once a scene is cut, a low volume copy of this specific wild track can be placed into the scene which hides the impact of the cuts (most important, if the continuity problem is volume or intensity related).
- Record mono for pretty much everything – there is next to no benefit in recording stereo for any sounds other than atmospheric wild tracks (see above), which are also sometimes done in surround sound if the final production / presentation method will take advantage of it.
There were lots of sound crew jobs discussed, most importantly stating what they actually do and how the job is important to the post-production assembly of the sound. I’m going to totally ignore this, as to be honest it is all redundant for our OVFM environment where a single individual will probably be doing everything. As a result, here’s another bullet point list:
- While the final edit might be a single image track and several sound tracks (e.g. dialogue will professionally be placed onto a separate track so that it can be replaced for international markets), this totally hides what really happens in production. The example that we had (a 3min long police siege / raid scene from a film called “Jack Falls’s”) started with 4 sound tracks in the first edit, rising to 192 tracks during sound design (of which only about five were dialogue tracks), falling to 4 tracks in the final mix.
- Each track is a very minor piece of the overall sound track of the film – while each track might be recorded at full volume, it is modified to fit the scene and required stresses. Crowds for example are recorded separately and added across the final edits of the dialog to flatten the sound-scape. In every episode of Eastenders for example, the background characters in the pub are actually near to miming so that the dialogue of the focus characters remain the only voices being recorded.
- Many of the sound tracks that will be added to a scene during sound design will be so quiet that the listener will not hear them in the final film and are present for sub-concious perception. They will however, hear them not being present.
- In feature films, almost everything that you hear on screen (other than dialogue) was created in the Foley room (Foley is the act of re-creating sounds using other devices to replace the real sound onscreen, including sound FX – named after Jack Foley, who oddly enough, also created the ‘CC in a TV’ close-captioning logo). In demonstration, we watched a sequence of the latest Robin Hood film in parallel to the sound engineers in the sound room – nothing in the three minute sequence was real sound, recorded on the set – everything from horse foot-steps (the obvious traditional one), to man walking on grass (how subtle is that?) to the quiet squeak of the leather jerkin as Russell Crowe turns around to face the camera.
- Most NLE (non-linear editors) are insufficient to do justice (editing wise) to a sound track (both Adobe and Final Cut fall into these categories). The simple fact is, they are designed for video editing (let’s face it, that’s what is written on the box), and a sound editor is always the best approach. In many ways, my opening lines of this article are born out by this – the software that we are using to put together our films are re-enforcing the lack of thought when it comes to sound.
- Dont make the final sound track too low in volume. Record initially at a good recording level (lowering intensity is not a problem, but boosting the intensity also boosts all of the imperfections, especially things like electronic hums), and don’t lower it too much in post. A big mistake of visually oriented film makers is subduing the sound generally – subdue only when it will enhance, especially prior to loud events like gun-shots and explosions, or to build tension.
The basic equipment list which might be utilised by a sound engineer (or you) breaks down to:
- Boom microphones
- Radio microphones
- Wind and shock isolation mounts
- Boom poles and stands
- Foldback (essentially, earphones but in specific use by the sound crew)
One of (if not the only in practice) concern in regards to selecting a microphone is the polar pattern – the shape of the area covered by any particular microphone. In other words, where will the microphone pick up sound best, moderately, and least. There are five basic patterns, but which can be summarised in three groups:
- Omni Directional
The omni-directional microphone picks up sound equally from all around it. From a film-making point of view, it is almost totally useless unless hidden in the middle of a group of actors. It’s main use is in the collection of wild-tracks and atmospherics.
- Figure of 8
The figure of 8 picks up sound best from directly in front and behind (some are oriented deliberately to get the sides and no front/back). This microphone is also of little use unless hidden between two actors.
The cardiod range of microphones is of most use – it is named because the shape of the pick-up area is a little like a ‘heart’. Most good sound is picked up from straight ahead, a little less to front-sides, sides and rear-sides, with little directly behind (note: because of a technical issue, there is a small ‘bubble’ of high sound pickup directly behind the microphone which is usually not shown on diagrams).This is the standard type usually referred to as ‘shotgun’ microphones. Hyper-cardiods (e.g. Schoeps CMC-641) have a broad forward bubble of reception, and smaller to the rear. Super-cardiods (e.g. Sennheiser 416 etc.) push the bubble further and longer to the front, with a gradual decrease of the area behind. These tend to be very directional, and are perfect as boom-microphones pointed directly at the actors mouth as they will pick up next to no ambient sounds from elsewhere.
Contrary to what might be considered common sense, very directional microphones such as the (some) hyper- and all super-cardiods (which prevent extraneous sound being picked up) should be avoided in interior locations. A technical effect of their pick-up mechanism means that they are worse at amplifying reverberation and extraneous noise in confined spaces (i.e. interior shoots) and should really be restricted to location. For interiors, use a more general cardiod of hyper-cardiod microphone, such as the Schoeps.
Quality of microphones really pans out in the resistance to RF interference and to a certain extent, build quality. If up close (e.g. a boom mic a few inches from the face) low priced microphones are more than comparable to higher priced manufacturers. However, as the mic pulls back such as when the boom needs to be further away for a wider camera angle, quality starts to suffer – quality microphones record quality sound for longer distances.
Radio Microphones are best used only where a boom microphone is impractical (e.g. long shots, very confined spaces). They are a nightmare to hide and are very susceptible to other sounds. As Roland put it – actors are not good microphone mounts – they rub, generate their own internal noises and don’t like their personal parts prodded (male and female) in the attempt to hide the microphones and transmitters. Additionally, radio microphones, because of where they are placed / hidden give flat or confusing sound – the ‘ear’ is in a different place to where it would naturally be (like the boom mic, out in front of where the actor is speaking). This can be corrected in post, but only by an expert and it is expensive.
A word of warning – right now, DO NOT BUY A SECOND HAND radio microphone. The upcoming 4G WiFi systems cover Channel 69, which previously was used by radio-microphones. Most older equipment can not be upgraded to use the new Channel 38, so the only safe way of get a working (and legal) radio microphone is to buy new. Roland himself estimates that he has about £15,000 worth of equipment which is about to become worthless.
Mixers are used to fine-tune recording levels, but not much else unless, as with Roland recently, he had a scene using 4 booms and 10 radio mic all working together – and even then, they are normally recorded to different tracks for use in the editing suite. At the level of amateur film making, they are not really much use, especially as today they are often built in to recording equipment.
Mixers come into their own in the sound-studio itself (e.g. see the Foley video above), but again are really beyond what we as amateurs and independents will probably ever need unless we are hiring sound engineer services – the facilities provided by our video or sound editing software is probably more than sufficient to make up the difference.
Recorders are used to record sound. Obvious, right? But there is a big difference in how they accomplish this, the quality of their connectivity (i.e the jacks etc.), and a lot of the ‘quality’ and ‘feel’ of their sound recording is subjective. It often comes down to tape (e.g. the Nagra) or digital (such as the Zoom H4N, quite a common digital recording device in our environment). Also in this category are the video camera themselves, capable of direct audio input to the recording medium – as sound tracks and image tracks can be split in the editing suite and re-written to disk, this is a very convenient way of ensuring that the sound is always in sync.
For separate recording, a simple hand-clap on camera is more than sufficient to trigger a spike that can be tracked for sound synchronisation. The use of a clapper board (which essentially does the same purpose) in modern times is often used for cinematography purposes to display a much wider range of information useful to the general editing process – many modern electronic clapperboards are even totally silent.
While most of us at out level of club film making don’t really need to hire, if we are making our masterpiece and really want good quality sound then hiring is the way to go, don’t buy (or even bringing in a sound-engineer who will have his/her own equipment in the first place). As a general guide, these are typical prices (all in, with all the mounts etc.) that you can expect to currently pay on a daily basis:
- £9 – Boom microphone (Sennheiser 416 etc.)
- £24 – Radio microphone (Audio ltd. 2020)
- £3 – Boom pole
- £20 – Mixer (Sound Devices 442)
- £45 – Recorder (Sound Devices 744T – quite a beast)
Conclusions from the evening
I think I said it all in my opening lines – this was an incredible evening simply in the gradual reveal of the range of things to think about in regards to recording sound. While a good many things were ‘obvious’ even if they still are not done in practice, it was the minor things that really made the information about sound recording all the more valuable.
Coming from a CG background however, where the laying of a vast array of different layers is a natural event however, I was most influenced by the sound design phases – looking at Roland’s laptop and seeing dozens (er, hundreds – he actually had the Jack Fall’s project running live on the MacBook) was something I could relate to and which confirmed my ideas of how it should work were I’ve always thought I must be over complicating it.
This is definitely a session for people to attend who want to produce great sound.