This is the glossary for the WAV & BWAV Technical Specification.
- Channel or Leg:
A single audio stream. Stereo comprises two channels.
- Linear Audio:
Digital Audio that has not been subjected to 'lossy' data compression (bit rate reduction) (see below).
- Data Compression (a common term for bit rate reduction):
A process of compressing data in order to reduce file size or stream bandwidth. It is usually destructive (“lossy”) and is best avoided.
(Not to be confused with Dynamic Range Compression, where the difference in levels between the quietest and loudest sections of audio is reduced, so that the quieter passages sound louder).
There are many recording formats for digital audio, the majority of which use lossy data compression or bit rate reduction. These include mp2, mp3, AAC, mp4 (common on ipods) and ATRAC (Minidiscs).
The use of such lossy data compression should be avoided wherever possible. If it has to be used, e.g. for covert recording or when using an ISDN line, details of the type of compression and the bit-rate should be noted on the recording report. Most solid state audio recorders (e.g. Compact Flash) offer a choice of recording formats, and linear should always be used in preference to any of the data compressed formats.
It is possible to apply lossless data compression without compromising sound quality, using FLAC (Free Lossless Audio Codec). This has been adopted as a suitable means of file transfer. For more details of how you could use this to deliver your programme, contact the commissioning network.
Take care when moving files between devices. Certain proprietary software may irrevocably damage your audio. Depending on the settings, the software may add data reduction techniques or Digital Rights Management (DRM), either of which could prevent your programme from being broadcast.
Make sure you send us the original file generated by your digital editor, not a version that you've listened to on an mp3 player.
- Cascade (encoding):
Where audio is subjected to more than one pass through a lossy data compression encode/decode cycle. With each additional pass the audio quality is degraded, often to an unpredictable degree. Cascading is a serious issue for BBC Radio. The BBC uses lossy data compression for its digital transmission systems (e.g. DAB, DSAT, DTT and Internet streaming). If lossy data compression is also used during the production process it may interact, resulting in unsatisfactory audio quality as received by the listener.
Even if the quality of compressed audio sounds reasonable in the studio, it probably won't when the listeners hear it.
Peak Programme Level or Peak Level:
The highest audio level to which the programme consistently peaks.
Levels are often quoted as read from a BBC Peak Programme Meter (PPM), marked “1” to “7”, each division corresponding to 4dB. See: Meters.
The following reference levels are commonly used within the industry:
“Zero level” - which corresponds to PPM 4
“Peak level” - which corresponds to PPM 6
“Maximum Digital Coding” – which is 0dBFS (0dB with respect to Full Scale)
FOR BBC PROGRAMMES, THE PEAK PROGRAMME LEVEL SHOULD BE PPM 6.
If you are using a digital bar-graph meter, regular readings of -10dBFS (10dB below absolute Full Scale or Maximum Coding) are desirable.
Digital Peak Level:
The maximum level of audio that can be represented in a digital system.
Also called 0dBFS (dB Full Scale) or Maximum Coding.
The Digital Peak Level must be below 0dBFS, to avoid distortion. See LEVELS .
Any meter used in the production of programme material should be correctly calibrated. See: Reference Tone .
These are usually bar - graphs and their main characteristic is that they respond to programme transients more accurately than analogue meters (eg PPMs).
PPMs tend to under-read the true level of transient peaks.
Peak Programme Meters (PPMs):
The BBC PPM is defined by an international standard: IEC IIA.
It is designed not to show every peak of short duration. This is so that it reflects more accurately the loudness of the broadcast as perceived by listeners.
Single-needle PPMs conventionally have a white needle, and each meter will be labelled according to whether it is reading A, B, M or S.
Twin-needle PPM movements have their needles colour-coded as follows:
Red = A (Left) channel
Green = B (Right) channel
White = M (Sum)
Yellow = S (Difference)
Note: In some installations, twin-needle meters can be switched to show either A&B or M&S. In this case, the Red & Green needles may be used to indicate M&S respectively, or White & Yellow to indicate A&B respectively.
There is a recent variant of the M & S meter that has the yellow needle replaced with a black needle with a dayglo orange tip.
Single frequency audio (usually 900Hz or 1kHz sine wave ) whose level is relative to that of programme audio (as opposed to an absolute level). The level of the tone and its relationship to the programme audio should be stated in the metadata, for example: “reference tone recorded at 8dB below the peak level of this programme." See: Peak Programme Level or “tone at PPM 4 per leg”. See: Meters.
Not all networks require tone. Check specific delivery requirements with the commissioning network.
If submitting material with reference tone, tone and programme material must be have been monitored and measured through the same technical chain. The level of the tone must be equivalent to 0dBu and no less than 8dB below peak audio level for the programme material.
Listen to a sample of standard level tone at 900 Hz.
If you import this file into your programme equipment you should be able to reference your programme level to the level of tone:
On a BBC PPM, this tone should read 4:
On a digital bar graph meter, this tone should read -18:
On a VU meter, this tone should read -4 per leg:
On a BBC PPM, Peak Programme Level should be up to 8dB higher than this tone. See: Peak Programme Level.