1) EP 0 660 540 B1 Philips "Leap Year" patent

Decoder for decoding an encoded digital signal and a receiver comprising the decoder

Claim 1:

A decoder that

• decodes an encoded digital signal that can be described as a sequence of 'frames':
  • 'frames' consist of multiple 'information packets' (IP)
  • Each 'information packet' contains N bits (N > 1)
  • the 'frames' all start with a piece of synchronisation information
  • The signal is obtained by encoding a digital signal S with a sampling frequency Fs
• has an input (10) for receiving the encoded digital signal
• is capable of reproducing a replica of the original encoded signal S on an output (8)

If we define:

• B as the number of 'information packets' in a 'frame'
• BR as the bitrate of the encoded digital signal
• Ns as the number of samples of the original signal S corresponding to the information contained in one 'frame' of the encoded signal
• N ans Fs as described above

The relation P = BR * Ns / ( N * Fs ) gives P, where:

• If P is an integer, it is equal to B
• If P is not an integer, B is the integer below P. In other frames B is the integer above P, whereas the average framerate of the encoded signal is substantially equal to Fs/Ns

Comments

The numbers (8) and (10) are the numbers as mentioned in the original claim, and refer to numbers in the figures contained in the patent. (Fig. 4, 12, 14)

'Synchronisation information' is information needed to maintain the correct speed while the original signal is played back.

The given relation simply comes down to:

• If the number of IP's per second, Q = BR / N
• If the number of frames per second, W = Fs / Ns
• Then P = Q / W, which means P is the number of IP's one would need per frame to maintain the given frame- and IP-rate continuously
• However, IP's are not to be divided in parts. To resolve the problem of non-integer values of P, B is varied between round_down(P) and round_up(P): the number of IP's per frame is varied slightly to make the effective value of the framerate (W) of the encoded signal equal to Fs/Ns

Schematic of the signal and it's components:

Encoded digital signal:
    +--------------+--------------+---
    |  frame 1     |  frame 2     | frame ..
    +--------------+--------------+---
   /                \
  +----+----+---+----+
  |Ip1 |Ip2 |...|IpB |
  +----+----+---+----+
 /      \
+--+--+--+
|b1|..|bN|
+--+--+--+

At least Ip1 would contain 'synchronisation information'.
b1 .. bN are single bit's (valued 0 or 1)

An example including a timeline:

Time:   0s                                      1s                                      2s
Bits:   01234567890123456789012345678901234567890123456789012345678901234567890123456789012345678
Frames: |    1    |    2    |    3    |    4    |    5    |    6    |    7    |    8    |    9   
(ideal) +---------------------------------------|---------------------------------------|--------
IP's:   | 1| 2| 3| 4| 5| 6| 7| 8| 9|10|11|12|13|14|15|16|17|18|19|20|21|22|23|24|25|26|27|28|29|
        +---------------------------------------|---------------------------------------|--------
Frames: |    1   |    2   |    3      |    4   |    5   |    6      |    7   |    8   |    9
(real)  +---------------------------------------|---------------------------------------|--------

Note: the integer number of frames per second is a coincidence of this example.