Q about Signal Processing

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
imagine a FIR processor where the impulse response consists of just two impulses with the second one time delayed and inverted. it should produce a comb filter of sorts with the first null at DC.

now lets say i want to claim intellectual property rights for this filter for a certain application. but other guys are already using equalizers to deal with the same problem. if they wanted to steal my idea all they would have to do is to calculate the required frequency response and then design an EQ for it which would end up having the same impulse response that i specified. then they'd be able to claim they didn't use my idea but merely made an EQ which they had been using before.

question is. starting from that comb filter frequency response would they actually be able to arrive at the two impulses by any standard algorithm ? or are there many possible impulse responses which would produce the same frequency response ? i can only think of one other.

in other words i can't claim any rights to the frequency response as this is something they're already working on. but a more direct way of dealing with the problem is to produce a delayed and inverted signal. and i am asking whether a delayed and inverted signal is in fact exactly the same as an equalizer realizing the same response or is it somehow different ?

makes sense ?

HELP ! ! !
 
Last edited:
Considering that a fairly standard way to produce a comb filter is exactly that, delay and sum....

You can trivially go from an impulse response (which is essentially time domain) to a frequency response or vice versa by simply taking a Fourier transform of the impulse response to get the frequency response or vice versa to get from a frequency response to the corresponding impulse.

Sorry, I don't see any reasonably protect-able IP in that filter, possibly in your bigger application, but that approach to a comb filter is well known and well documented.

Incidentally, if you are trying to correct speaker time domain behaviour, Larvy has some interesting notes as to why it does not in general work in one of his presentations.

Regards, Dan.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.