The data reference index to be used for a given media sample is stored within that sample’s sample description. Therefore, a track must contain multiple sample descriptions in order for that track to reference multiple data files. A different sample description must be used whenever the data file changes or whenever the format of the data changes. The sample-to-chunk atom determines which sample description to use for a sample.
The sample description atom would contain the following data values:
Atom size |
… |
|
Atom type |
|
|
Version/Flags |
0 |
|
Number of entries |
2 |
|
Sample description size[1] |
… |
|
Data format |
|
|
Reserved |
0 |
|
Data reference index |
1 |
|
(sample data) |
… |
|
Sample description size[1] |
… |
|
Data format |
|
|
Reserved |
0 |
|
Data reference index |
2 |
|
(sample data) |
… |
If there is only 1 sample per chunk and the first 10 samples are extracted from sample description 2 and the next 30 samples are extracted from sample description 1, the sample-to-chunk atom would contain the following data values:
Atom size |
40 |
Atom type |
|
Version/Flags |
0 |
Number of entries |
2 |
First chunk[1] |
1 |
Samples per chunk[1] |
1 |
Sample description ID[1] |
2 |
First chunk[2] |
11 |
Samples per chunk[2] |
1 |
Sample description ID[2] |
1 |
The data reference atom would contain the following data values:
Atom size |
… |
|
Atom type |
|
|
Atom size |
… |
|
Atom type |
|
|
Version/Flags |
0 |
|
Number of entries |
2 |
|
Size[1] |
… |
|
Type[1] |
|
|
Version[1] |
0 |
|
Flags[1] |
0 (not self referenced) |
|
Data reference[1] |
[alias pointing to file #1] |
|
Size[2] |
… |
|
Type[2] |
|
|
Version[2] |
0 |
|
Flags[2] |
0 (not self referenced) |
|
Data reference[2] |
[alias pointing to file #2] |
Last updated: 2007-09-04