In our examination of the AES3 digital audio signal, the only bit left to discuss is channel status.
A channel status bit is carried in timeslot 30 for each of the two audio channels. The channel status bits are arranged in 192-bit blocks and are subdivided into 24 bytes. As with the user bits described previously, the start of each block is indicated by a Z preamble in subframe 1.
Each channel (subframe) can contain a unique block of channel status bits. While often the same information is present in both of the channels, they could be different.
The channel status bits provide a wealth of information about its associated audio channel, which can be used by an AES receiver/decoder for subsequent processing.
Channel status byte zero contains bits indicating either professional or consumer use (bit zero) and whether the audio sample contains linear PCM (pulse-code modulation) audio information or something else, such as a compressed format like Dolby E, AC-3 or DTS (bit 1).
As discussed previously (TV Technology, Sept. 19, 2007), the function of subsequent channel status bits differ depending on whether bit zero is set for consumer or professional. The following describes the function of channel status for professional mode.
IN A PROFESSIONAL SETTING
Byte zero also contains bits indicating whether emphasis (a specific type of equalization) was applied to the signal or not. More bits indicate the source sampling frequency lock condition (lock not indicated or unlocked) and specific sampling rates.
The choices for sampling rates are: no rate specifically indicated by the channel status bits, 48 kHz, 44.1 kHz or 32 kHz.
The sampling rate does not need to be specifically indicated in channel status for an AES3 signal to be decoded. As discussed previously, the design of the AES3 signal is such that the sampling rate information is derived from the signal itself, and can be at rates other than those that have specific channel status codes in bytes zero and four.
Byte 1 contains bits for channel mode and user bits management. Channel mode bits indicate how the two subframes are to be used. In two-channel mode, the subframes represent two distinct audio channels. Single channel mode indicates a monophonic channel. Stereo mode, as the name implies, indicates a stereo signal, with Channel 1 being the left channel. There’s also a primary-secondary mode with subframe 1 as the primary.
There are three single channel modes for cases where the sampling frequency is twice the frame rate. In these cases subframe 1 and subframe 2 carry successive samples of the same signal, according to the AES3 standard, and the channel status bits indicate whether a single channel, stereo-left or stereo-right channel is contained in the audio sample.
If a channel mode is not indicated, the receiver should default to two-channel mode.
The most common use of the user management bits is to indicate if the 192-bit block structure is used for the user bits with preamble Z (indicating the start of the block) if there is no user information indicated, or if user data conforms to IEC 60958-3 (the consumer digital audio standard).
Byte 2 indicates maximum audio word length and whether auxiliary sample bits are to be used. The default is 20 bits for maximum word length with the use of the auxiliary sample bits undefined. Another option is a maximum audio word length of 24 bits to be used for the main audio channel only, with none available for any auxiliary sample bits. A third option indicates a maximum word length of 20 bits with the aux sample bits used for a low bit-rate audio “coordination” signal.
The next set of bits in channel status Byte 2 can be set to explicitly indicate how many bits are actually used to encode the audio signal, from 16 to 24 bits. The last two bits in this byte indicate if an alignment level is specified, and if so, what type (20 dB below full scale or 18 dBFS).
Byte 3 indicates details when the AES3 signal is as part of a group of channels (multichannel mode).
Byte 4 indicates whether the AES signal is a digital audio reference, and if so, what kind. Byte 4 also contains bits that indicate sampling frequencies different from those indicated in byte zero. Here sampling frequencies of 24 kHz, 96 kHz, 192 kHz, 22.05 kHz, 88.2 kHz and 176.4 kHz can be indicated. The last bit is a sampling frequency scaling flag, which if set indicates that the sampling rate is 1/1001 times that indicated in the previous bits of byte 4.
Zipping through the rest of the bytes of the channel status code:
- Byte 5 is reserved;
- Bytes 6–9 are used for alphanumeric channel original data;
- Bytes 10–13 are for alphanumeric channel destination data;
- Bytes 14–17 are for a local sample address code;
- Bytes 18–21 are for time-of-day sample address code;
- Byte 22 is for a flag that indicates if the channel status information is reliable; and
- Byte 23 is used for a channel status cyclic redundancy check character.
This last byte checks that the channel status block from byte zero through byte 22 was validly received.
How an AES receiver interprets channel status bits (if it does at all) isn’t set by the standard and depends on the individual piece of gear. Hopefully this information would be described in a user’s manual.
If an emphasis curve was indicated in channel status, for example, the receiver should be able to decode that information to allow proper application of a corresponding de-emphasis curve to the decoded analog audio signal.
If there’s conflicting information between the actual audio signal and the channel status, it’s not clear what a receiver would do. If, say, the audio sample was done at 48 kHz, but the channel status bit was erroneously set for 44.1 kHz, a receiver could possibly ignore the channel status and lock to and decode the source signal properly; or, as another possibility, it could try to unsuccessfully decode the source signal at 44.1 kHz.
Having test equipment that can provide a readout of not only channel status, but also the rest of the bits comprising the AES3 bitstream, is invaluable for troubleshooting these sorts of issues.