Is it true that the magnitude response remains same but only the phase response is different, between convolution and correlation? Bharat
correlation v/s convolution?
Started by ●October 3, 2008
Reply by ●October 3, 20082008-10-03
On 3 Okt, 15:10, "bharat pathak" <bha...@arithos.com> wrote:> Is it true that the magnitude response remains same but only > the phase response is different, between convolution and > correlation?Why not calculate the FTs of the two fomulas and compare? Rune
Reply by ●October 3, 20082008-10-03
On Oct 3, 9:10�am, "bharat pathak" <bha...@arithos.com> wrote:> Is it true that the magnitude response remains same but only > the phase response is different, between convolution and > correlation? > > BharatConsider a PN sequence. If you convolve it with itself you don't get a peak. If you correlate it with itself, you do get a peak. John
Reply by ●October 4, 20082008-10-04
bharat pathak wrote:> Is it true that the magnitude response remains same but only > the phase response is different, between convolution and > correlation? > > BharatMaybe you could define your terms a little better. Magnitude response refers to a system / filter / etc. Ditto phase response. Convolution is an operation. Correlation is an operation. Sure, the operation of a system on a signal is a convolution. But, the system isn't a convolution as such. I think you mean this: "Is it true that the magnitude response of a system remains the same and the phase response is different IF we reverse the system response in time?" Fred
Reply by ●October 4, 20082008-10-04
>Is it true that the magnitude response remains same but only >the phase response is different, between convolution and >correlation? > >Bharat >Hi Bharat, Yes, this is true. Also, the phases are simply opposite in sign. Here's a link that expands on this, describing how to use this property to create zero phase IIR filters. Regards, Steve http://www.dspguide.com/ch19/4.htm