Format cannot be fully quality assured in its generic state

[geoffmatt]

Looking at implementing this format I believe it is impossible to test in the general case.

We expect to see specific implementations of this format, where only the values required by the system are populated, and all others either ignored or, if mandatory, set to an arbitrary value such as 0.0.
Such implementations can be tested because it is possible to know what values are expected and how they should be used.

However the concept here is to provide a generic file format for any acoustic data, and this won't be practically testable. The reason is that the number of parameters and variables involved is extremely high, and therefore the possible number of combinations of these inputs is accordingly very high. It simply won't be practical to cover all these test cases, and therefore won't be possible to sign-off that the generic format can be read in all cases.