Thank you for your very encouraging response.
>Could you clarify one thing: do you mean, actually use MARSEC and PROSICE,
>or do you mean "the type of material that is in them"? I thought we had
They indicate the type of material, though I feel PROSICE can be directly
borrowed because of its high level of recording specifications.
>agreed that we would collect some material of our own? I think our
>concerns are that MARSEC etc don't give us the type of info we need for
>the acoustic end of things.
MARSEC is not good enough in this respect but PROSICE is, as Mark can
confirm. In addition, we can collect some more material with a different
speaker. Because of the automated process (except transcribing), the
construction of the corpus won't take much longer than that of a list of
words and phrases.
Alex