A II.1  Subheader Values


1.1  File Number


The file number is used to identify all  sectors that belong to one   file. 
A given file may be physically interleaved with  other files on the   disc. 
All sectors of a logical file have the same value in the file number  byte. 
The file number can  be used to select  sectors that belong  to the    same 
logical file and to reject any others.  The file number with a byte   value 
of 0  is  only used  for  a file  that  will  be read  consecutively.    No 
interleaving is possible with file 0.  Files 1 to 255 (see Figure   A.II.1) 
may be read consecutively or may be interleaved with other files.


Figure A.II.1

               ______________________________________
               _ Byte Value _     File Number       _
               ______________________________________
               _ 0000 0000  _ 0 (consecutive sector _
               _            _    reading only)      _
               _ 0000 0001  _ 1                     _
               _ 0000 0010  _ 2                     _
               _     :      _ :                     _
               _ 1111 1111  _ 255                   _
               ______________________________________





1.2  Channel Number

A real-time record may contain several information channels.  Two   notable 
reasons why this is allowed are:

-  to minimize  the amount  of disc space  wasted by  gaps between    audio 
blocks, and

-   to allow  one of  several mutually exclusive  sequences to   accompany 
  each other, e.g. one of a set of audio sequences with a video sequence.

A real-time  record is  created by  merging several  different  pieces   of 
information together.  When  this is  done, each logical  segment may    be 
given a unique channel number, which is recorded  in the second and   sixth 
bytes of  the  subheader.  Each  real-time  record has  a  maximum  of   32 
channels (channel  numbers 0-31).   Because the  maximum number  of   audio 
channels is 16, audio sectors can only use  channel numbers 0 to 15.     In 
order to conserve channel codes and simplify programming, a common  channel 
number can be used for sectors that are  always to be selected    together, 
regardless of  the  type  of  coding  information  (audio,  video,    etc.) 
contained in the sector.

Upon playback, any combination  of channels in the  record may be    played 
together if such a  combination is allowed in  the Base Case  System.   The 
selection is controlled by a 32-bit channel selection register.  Each   bit 
in the register corresponds to one subheader channel number.  If a bit   is 
set to  one  in the  channel  selection register,  the  sectors  for    the 
corresponding channel number are
selected.  If the bit is reset, i.e. to zero, sectors for the
corresponding channel are ignored.  The assignment of bits in the   channel 
selection register is illustrated in Figure A.II.2.


Figure A.II.2

           31        23        15                 0
           _________________________________________
           _1000 0000_0000 0001_0001 0000_0000 0001_
           _________________________________________

For the  channel selection  register depicted  in Figure  A.II.2,   sectors 
would be selected if their subheader contained channel  number 0, 12,   16, 
or 31.

Real-time records are always played back asynchronously.  This  means  that 
the  application  program   requesting  playback   continues  to    execute 
concurrently.  It is possible  for the application to  change the   channel 
selection  register   while  playback   is  in   progress,  allowing     an 
instantaneous switch between channels in real-time.

Non-real-time records can use any channel number (0 - 31).


Figure A.II.3

          _____________________________________________
          _ Channel Byte _ Channel Number _ Data      _
          _____________________________________________
          _  0000 0000   _      0         _ A _ V _ D _
          _  0000 0001   _      1         _ A _ V _ D _
          _      :       _      :         _ : _ : _ : _
          _  0000 1111   _      15        _ A _ V _ D _
          _  0001 0000   _      16        _ - _ V _ D _
          _      :       _      :         _ : _ : _ : _
          _  0001 1111   _      31        _ - _ V _ D _
          _  0010 0000   _   reserved     _ - _ - _ - _
          _      :       _      :         _ : _ : _ : _
          _  1111 1111   _   reserved     _ - _ - _ - _
          _____________________________________________

Figure A.II.3 gives which channel numbers may be used.

Channel numbers can be used for audio (A),  video (V) and program   related 
data (D) sectors, but audio channel numbers must be within the range of   0 
to 15.

The audio  channel selection  for the  ADPCM decoder  is  controlled by   a 
separate 16-bit  audio channel  selection register.   Any audio    channels 
selected by  the  32-bit  register  but  not  the  16-bit  register     are 
transferred to the host system along with the other (non-audio)   channels; 
the audio channel selected  by the 16-bit register  is transferred to   the 
audio processor.



1.3  Submode

The submode byte is bit-encoded as shown in Figure A.II.4.

Figure A.II.4

              _______________________________________
              _  bit number _ Meaning               _
              _______________________________________
              _      7      _ End-Of-File (EOF)     _
              _      6      _ Real-Time sector (RT) _
              _      5      _ Form (F)              _
              _      4      _ Trigger (T)           _
              _      3      _ Data (D)              _
              _      2      _ Audio (A)             _
              _      1      _ Video (V)             _
              _      0      _ End Of Record (EOR)   _
              _______________________________________

End Of File (EOF)   :This bit is set to one for the last sector of a  file. 
                          Only the last sector of a file can have this bit 
                          set to one.

Real-Time sector (RT):This bit must be set to one for all sectors that  are 
                            part of a  real-time file.   This means   that 
                            the  data   has  to   be  processed    without 
                            interrupting the real-time  behaviour of   the 
                            CD-I system.

Form (F)              :This bit is zero for all sectors recorded in   Form 
                 1 and set to one for all sectors recorded in Form 2.

Trigger (T)                :This  bit  is  used  for real-time   synchron- 
                      ization of video,  audio and  program related   data 
                      (e.g. text).  It generates an interrupt when set  to 
                      one (see VII.4).


Data (D)               :This bit  is set  to one  for  all sectors    that 
                 contain program related data.   When this bit is  set   to 
                 one, the form bit must be zero.

Audio (A)              :This bit  is set  to one  for  all sectors    that 
                 contain audio data.  When this bit is set to one the  form 
                 bit is also set to one.

Video (V)              :This bit  is set  to one  for  all sectors    that 
                 contain video data.

End Of Record (EOR):This bit must be set  to one for the last sector of   a 
                          logical record.   Only  the  last  sector  of  a 
                          record can have this bit set to one.  The use of 
                          the EOR  bit  is only  mandatory  for  real-time 
                          records.

The allowable combinations of  the form, audio, video  and data bits    are 
specified in Figure A.II.5.

Figure A.II.5

        __________________________________________________
        _    Coding    _     Form 1     _     Form 2     _
        _ (bits 3,2,1) _ (Form bit = 0) _ (Form bit = 1) _
        __________________________________________________
        _   Audio = 1  _       NA       _       OK       _
        _              _                _                _
        _   Video = 1  _     OK [NRT]   _       OK       _
        _              _                _                _
        _   Data  = 1  _       OK       _       NA       _
        __________________________________________________

where

OK = allowed combination
NA = not allowed combination
NRT = only to be used for non-real-time sectors

Only one of the data, audio and video  bits can be set to one at  the  same 
time in any sector.  If all bits of  the submode byte are zero except   the 
Form bit, then the sector does not contain any CD-I data and so is   either 
an empty sector or a message sector. One of bit 1, bit 2 and bit 3 must  be 
set to the value 1 for all sectors except empty or message sectors.

It should be noted that the combination of  the Video bit (bit 1) set    to 
one, the Form bit (bit 5) set to zero and the Real-Time sector bit (bit  6) 
set to one is reserved for full screen, full motion quality video.

An audio sector must contain 2304 bytes and  therefore cannot be used    in 
Form 1.  A data sector cannot be used in Form 2 because data errors  cannot 
be concealed and the integrity of the data cannot be guaranteed.

Since Form 1 real time sectors may not  be corrected in case of error,   if 
this would  endanger  the  real  time  behaviour of  the  system,  it    is 
recommended:

-   to group  all the  Form 1 sectors  that need  to be  corrected in    a 
  separate record placed just before the real time  record and to  declare 
  these Form 1 sectors as "non real time" (RT = 0)

-   not to mix, in the same record, real time and non real time sectors.
















1.4  Coding Information


1.4.1  Audio

The audio Coding Information byte is defined as follows:

   _____________________________________________________________
   _ 0 _Emphasis_Bits per Sample_Sampling Frequency_Mono/Stereo_
   _____________________________________________________________
bit  7  6        5             4 3                2 1         0

Reserved
Bit 7  = Zero

Emphasis (see IV.4.2)
Bit 6 
   0 = Emphasis off
   1= Emphasis on

Number of Bits per Sample
Bits 5-4
    0 0= 4 bits
  0 1 = 8 bits
  1 0 = Reserved
  1 1 = Reserved

Sampling Frequency
Bits 3-2
  0 0 = 37.8 kHz
  01 = 18.9 kHz
  1 0 = Reserved
  1 1 = Reserved

Mono/Stereo
Bits 1-0
  0 0 = Mono
  0 1 = Stereo
  1 0 = Reserved
  1 1 = Reserved








1.4.2Video

Details of the coding method and resolution of each video data sector   are 
contained in the Video  Coding Information Byte in  the subheader of   each 
video sector.  The format of this byte is:

              _____________________________________
              _ ASCF _ EOLF _ Resolution _ Coding _
         _____________________________________
           bit   7       6    5        4   3 2 1 0

where
  ASCF = Application Specific Coding Flag

  Bit 7
    0 = coding conforms to the specification of chapter V
     1 =   coding is application-specific, i.e. to be  inter- preted   by 
          application software.

  EOLF = Even/Odd Line

  Bit 6
    0 =Even lines
    1 =Odd lines

If error concealment is to be used, this flag indicates whether the  sector 
contains the even or the odd lines of the image.  If error concealment   is 
not to be used, this bit is set to zero.

  Resolution
  
  Bits 5-4
     0 0 = Normal
   0 1 = Double
   1 0 = Reserved
   1 1= High

  Coding

  Bits 3 2 1 0
       0 0 0 0 = CLUT4                1000 QHY
   0 0 0 1 =CLUT7                1001 to 1111 Reserved
   0 0 1 0 =CLUT8
   0 0 1 1 =RL3
   0 1 0 0 =RL7
   0 1 0 1 =DYUV
   0 1 1 0 =RGB555 (lower)
   0 1 1 1 =RGB555 (upper)
1.4.3Data


The data or program  related data Coding Information  byte is defined    as 
follows:

     _____________________________________________
     _                  _        (Reserved)      _
     _____________________________________________
 bit 7                                           0