Audio on VTRs
(first published in International Broadcast Engineer magazine)
There was a time when sound recordists carried Nagras with Neopilot tracks, transfer bays dubbed this audio to sprocketed tape and editors re-synchronised audio and video from the clapperboard! Of course this and some updated variants are still relevant to parts of our industry but much has and continues to change. This change process is almost invariably driven by changes in picture recording methods and sometimes the new technology may include real benefits to the audio quality or the ease of use of the sound system.
Sadly, there have been times when the audio issues seemed almost an after-thought. It is interesting to find that in some professional camcorder brochures there are no references to audio facilities at all. Another describes the audio input as balanced, on a 3 pin connector (type not specified) and either -60 dB or +4 dB. The output specification was described as +4 dB, 600 ohm and balanced. dB (without some reference) is not a signal level and it is ludicrous to be offering a 600 ohm output impedance when that was abandoned in favour of 50-100 ohm output impedances on most audio equipment 10-20 years ago! This is really the level of information we have come to expect on consumer products, and does not encourage a respectful reading from an audio engineer!
This is unfortunate, as lack of good flexible audio facilities on the video recorder mean that people are forced to adopt alternative sound strategies. For example if the VT sound tracks on an EFP/ENG camcorder cannot be controlled satisfactorily, people must resort to separate sound recorders such as DAT. Obviously the results can be excellent and there may be some benefits in recording flexibility but synchronisation once more becomes an issue to be resolved. "Issue to be resolved" can usually be translated into "cost" so that is never good news for the production budget!
This makes no claim to be all encompassing, but only to give an insight into the sound facilities available on some of the media now being offered to the sound engineer, with a brief backward glance to put this into an historical context. This can be useful when extracting material from the archives!
The 2" transverse scanned quadruplex system offered nothing of value to the wise sound engineer who would keep the audio somewhere better than the VT track for as long as possible! 1" C format provided the potential for much improved sound, especially when the Dolby noise reduction was used. Sadly, alignment of audio often seemed to come a poor second to video alignment and badly aligned Dolby encoders/decoders could wreak havoc on the subtleties of the audio.
We have now had Sony's Beta then Beta SP format with its four audio tracks for over ten years and this was a big step forward. The audio was no longer confined to a relatively low speed linear edge track and FM recording with a head on the drum provided the potential for high grade audio, good enough to sound "hi-fi" to all but the most demanding Nicam equipped viewer. Sadly, the track design of the Beta tape did not allow over-recording of the FM audio - the video had to be re-recorded at the same time. This obviously meant a generation loss for video as well as audio and this severely limited the usefulness of the FM tracks during the post production phase. High grade audio from the FM tracks often ended up "bounced" to the linear edge tracks, which now suffered an even lower tape speed than the old quad machines.
The audio track is usually important in determining the correct edit point. This might often be just before the start of a new word so as to leave as much ambience/reverberation around the previous word as possible; the start of the new word then hopefully masks any ambience change.
Recording audio in a digital form provides great sonic benefits and facilitates multi-generation copying. Of course quality is only part of the issue and facilities are important. For example, the ability to "scrub" audio backwards and forwards, as though "jogging" an analogue audio recorder is a common way to find an edit point. Doing this with a digital source was a serious challenge to early audio hard disc systems and are a sophisticated part of even today's systems, requiring complex software to produce forward and reverse audio at varying speeds! The first generation of digital formats helped sonically but fell silent when not in normal speed forward play mode.
Later systems such as DVCPRO provide an analogue cue track. Whilst this can be used independently, it usually contains a mono sum of the two digital tracks so provides a "scrub" audio feed to be used when locating edit points. Beta-SX adopts the alternative approach which is to decode the forward/backward variable speed digital audio into a real audio output.
The serious moves to digitise the entire broadcast chain in many organisations means that digital VTRs with digital audio, interfaced digitally are at last becoming more common. As with DAT and Mini Disc, at least a couple of the contenders began as a development from a consumer format. This has proved a stumbling block for the audio formats which did not make provision for professional needs such as time code within the initial specifications. More planning has gone into the new video formats and, unlike DAT, there is real provision for extended formats to facilitate professional working.
Format Tape Type Tracks Method Comments C - format 25.4 mm Oxide 4 Longitudinal, analogue Dolby alignment req. care Betacam SP 12.7 mm Metal particle 4 2 FM helical, 2 longitudinal 2 separately editable Beta-SX 12.7 mm Metal particle 4 16 bit/48K MPEG-2 Digital Betacam 12.7 mm Metal evaporated 5 4 @ 20 bit/48K linear, 1 analogue DV 6.35 mm Metal evaporated 2 16 bit/48 kHz linear DVCAM 6.35 mm Metal evaporated 2 16 bit/48 kHz linear DVCPRO 6.35 mm Metal particle 3 2 @ 16 bit/48 kHz linear, 1 analogue DVCPRO50 6.35 mm Metal particle 5 4 @ 16 bit/48 kHz linear, 1 analogue Digital S 12.7 mm Metal particle 4 2 @ 16 bit/48 kHz linear, 2 analogue
The DV format and all the extensions from it can provide four tracks, though usually with reduced sample rates or number of bits. Not all recorders implement the four channel mode.
The upwards compatibility of the consumer DV format to Sony's professional DVCAM or Panasonic's DVCPRO can provide both a strength and a weakness. Low cost and highly compact lightweight DV cameras can produce very acceptable pictures in situations where more bulky professional cameras would not be acceptable. If suitable camera mounts are used, the pictures can sometimes be so good that it becomes tempting for producers to exploit this cost saving even for situations where the compact size of the DV camera is not essential.
Whilst the pictures may be good, some serious sound problems arise. The main sound issues are audio level control, monitoring, sample locking and connector types.
A major limitation of some DV products is the inability to separate the automatic record level control for audio and video. Auto level has obvious advantages for the video in news and other unrehearsed situations and can obviously produce a controlled audio track. The human eye already has a fast acting "auto level" control in the iris and this gives us a natural acceptance of images that adjust in brightness to show the optimum amount of detail. The ear also has "auto levelling" functions but these operate much more slowly and we expect quiet sounds that follow loud ones to remain sounding quiet for a considerable time, if the effect is to be natural. Auto audio levels therefore generally do not sound good and audio quality can be irretrievably damaged by DV equipment. One major UK broadcast company has found that the record path of their DV equipment seems to behave more like an audio compressor than an automatic gain control system so when using it, they use low output microphones or attenuator pads to keep the audio compression to a minimum. Although this produces a low level on tape they don't worry as any increase in system noise is masked by the ambient noise likely to be present in the locations where this technique is adopted.
Traditionally, professional cameras have been used with a separate sound recordist as broadcast quality sound cannot be obtained with auto level setting devices. Even those companies now covering news with a single cameraman/recordist, expect the audio levels to be manually adjusted, though the ability to have only one audio track metered in the viewfinder is proving a problem for some Beta SP units now working this way.
It is common for the two audio tracks to be used quite independently, sometimes with only a non-critical camera mic guide track on one channel, but equally often with two equally important radio microphones. There are a growing number of cameramen's anecdotes of not getting the expected audio on the tape when two radio mics are used and only one is monitored! If single man crews are to be used, it is important that adequate monitoring and metering of all the audio tracks is provided - whether this leaves room to see the picture may then be in doubt!
The "playback" mode of monitoring on Beta SP equipment does not provide a good quality indication, but at least gives an assurance that some audio has got onto the tape! Most digital formats lack this and instead rely on digital error checking to ensure what goes into the recorder gets to tape. There are already tales of material shot overseas being found to be a "silent movie" on return to base! Hopefully such tales will become less common as people change shooting habits to improve verification of what has been shot. Perhaps the makers will find a way to squeeze an extra head on the drums of digital VTRs to provide an "offtape" playback monitor?
The four audio tracks of Digital Betacam are far less likely to suffer from problems of metering and monitoring as this format will normally only be used on productions that can justify a separate sound recordist, possibly recording one or more stereo tracks.
Serial digital interfacing has become common for post production and transmission systems and this includes up to four channels of audio embedded in the SDI stream. It is obviously essential that by the time audio samples are to be embedded, they should be locked to the video to avoid audio framing errors. Audio which is digitised by consumer DV format recorders is not locked to video and whilst some equipment may be able to handle such audio, unlocked audio samples is generally undesirable for post production as it can lead to problems ranging from no audio at all, to periodic clicks and other artifact becoming superimposed.
The smallness of the DV equipment which may be desirable also means there is much less space for connectors. Full size robust professional connectors are inevitably jettisoned in favour of the domestic style phono/cinch connectors for video and 3.5 mm unbalanced jack plugs.
Whilst these can pass signals of as high quality as anything else, they can also fall out then pass nothing! The potentially unreliability of connectors, coupled with the possibility of not monitoring all the audio can produce serious reductions in audio reliability. Having all audio circuits unbalanced may also be a drawback, though careful use can often avoid too many problems.
This review has focused on tape based systems only but hard disc server systems and hard disc recorders such as Ikegami's EditCam cannot pass without mention! Indeed, Ikegami's four discrete full specification channels recorded directly to hard disc has much to commend it, though issues of disc pack exchange/ownership pose some interesting dilemmas for those not part of a self sufficient corporation!
Digital VTRs offer many advantages for the audio recordist and can maintain audio quality much better than their analogue counterparts. The temptation to adopt low cost equipment needs to be considered fully in the context of the programme. Having to re-shoot because of sound problems never saves anyone any money! Low cost DV equipment is a useful tool but needs to be used selectively in those situations where the compact size are really necessary.
Hardly surprisingly the relative merits of the various digital tape formats systems will most often be assessed on the cost/benefits they offer on aspects of the video signal but audio should not be ignored. If the video format has inadequate audio facilities for the particular purpose, extra resources such as separate audio recorders such as timecode DAT may need to be introduced and the extra work of making, synchronising and handling a separate audio recording may offset all the savings made in choosing that "low cost" format!
Note: this article was written in 1998 so does not necessarily reflect the latest developments in video recording.
All material is copyright PHM © 2004.
P H M
(P H Music) : PO Box 383
Bury : BL8 4WX : GB
tel/fax: +44 (0)1204 887161