Re: Purpose of using a random seed for Maximum Length Sequence

On 2009-08-25, Nicholas Kinar <n.kinar@xxxxxxxx> wrote:

This is not cryptographically sound, but it probably doesn't matter
since all that we are doing is measuring the frequency response of a
concert hall or auditorium.

However, would I gain anything by selecting a random seed for the linear
feedback shift register?

Probably not. Note that, since you've said that the LFSR you're using
has maximal period, all its states (expect the zero state) belong to a
single cycle. Changing the seed merely changes the starting position
along that cycle.

That said, I know very little about the use of pseudorandom sequences
in signal processing, so I can't really say anything definite.

Why do you ask, anyway?

Ilmari Karonen
To reply by e-mail, please replace ".invalid" with ".net" in address.