Re: findDelay.m

[Juan Pablo Carbajal]

On Tue, May 14, 2013 at 4:12 PM, MV <address@hidden> wrote:
> Ok, thanks for the tips
>
> Marcos Varella
>
> Em 14/05/2013 10:09, "Ben Abbott" <address@hidden> escreveu:
>
>>
>> On May 14, 2013, at 9:53 PM, MV wrote:
>>
>> > Em 14/05/2013 09:25, "Juan Pablo Carbajal" <address@hidden>
>> > escreveu:
>> >> On Tue, May 14, 2013 at 2:24 PM, MV <address@hidden> wrote:
>> >> > Hi there
>> >> >
>> >> > Thanks for the answer.
>> >> >
>> >> > http://www.mathworks.com/help/comm/ref/finddelay.html
>> >> >
>> >> > Marcos Varella de Almeida
>> >> > address@hidden
>> >> > (92) 8821-6511
>> >> >
>> >> > Em 14/05/2013 08:20, "Juan Pablo Carbajal" <address@hidden>
>> >> > escreveu:
>> >> >
>> >> >> On Tue, May 14, 2013 at 1:56 PM, MV <address@hidden>
>> >> >> wrote:
>> >> >> > Hi there,
>> >> >> >
>> >> >> > Does Octave have an similar function as matlab "findDelay"?
>> >> >> >
>> >> >> > How can I use it on Octave?
>> >> >> >
>> >> >> > Thanks!
>> >> >> >
>> >> >> > Marcos Varella de Almeida
>> >> >> > address@hidden
>> >> >> > (92) 8821-6511
>> >> >> >
>> >> >> >
>> >> >> > _______________________________________________
>> >> >> > Help-octave mailing list
>> >> >> > address@hidden
>> >> >> > https://mailman.cae.wisc.edu/listinfo/help-octave
>> >> >> >
>> >> >>
>> >> >> That is from hte communication system toolbox. Please check in
>> >> >> Octave-Forge
>> >> >> http://octave.sourceforge.net/packages.php
>> >> >>
>> >> >> If you do not find it, can you tell what is the algorithm their are
>> >> >> using? If you can then (maybe a publication?) then it can be
>> >> >> implemented.
>> >> >>
>> >> >> Cheers
>> >>
>> >> So do you understand how they calculate the delay?
>> >> Check the first row of the table of methods.
>> >> "Cross-correlating the columns of X and Y over a range of lags
>> >> -MAXLAG:MAXLAG."
>> >>
>> >> How would you get the delays from that?
>> >
>> > I have two periodical signals, not sinoildal. I don't know how the
>> > function works. I'll try to find a matlab m file...
>>
>> Please don't look at Mathworks copyrighted code and then describe it here.
>> We don't copy closed source code and avoid looking at it.
>>
>> You may want to avoid placing your email and phone number in these emails.
>> The result is replicated in across the internet.  No doubt you are exposing
>> yourself to spam and uninvited phone solicitations.
>>
>> One more point, please avoid top posting.
>>
>>         http://www.jefftk.com/news/2013-02-16
>>
>> Ben
>>
>>
>

Also, check the function hilbert in the signal package.
http://octave.sourceforge.net/signal/function/hilbert.html

The abs of the Hilbert transform estimates the envelope and its angle
(i.e. angle(hilbert(y))) gives you the the phase of the signal. The
phase of different signal (if you know their periods) can be used to
get a notion of delays as well.

Just to mention that delays can be estimated in many ways.