Glad I could help, thanks for watching. Make sure to check out my website adampanagos.org for additional content (600+ videos) you might find helpful. Thanks, Adam
your explanation is quite excellent..... really it helped me in my signals and systems lab....hoping for more videos like this on different topics in (s and s)......thank u very much adam................
+Junchao Lu Keep in mind that Matlab (and computers in general) can ONLY work with discrete signals. Storing a continuous-time signal (even for a short amount of time) would require an infinite amount of memory since a continuum has an uncountably infinite number of points. So, when calling the "conv" function in Matlab, the signals passed in (such as the variable x) are always vectors of a finite length and Matlab actually performs discrete-time convolution. This is a multiplication and summation of the vectors. The quantity delT = median(diff(t)) is just a scalar and is essentially equal to the "delta t" of the time vector used in the problem. Multiplying by delT normalizes the summation that Matlab did inside the "conv" function to account for the sampling interval used in the problem to yield a final value that is equal to the area one would have obtained doing a real continuous-time integration. You can think of the multiplying and summation performed inside the "conv" function as doing a rectangle method approximation of the real integral we'd like to work. The product of the signals is the function/rectangle height, delT is the width, and the product of these two provides the rectangle area which should closely approximate the continuous-time integral/convolution. See the equation near the top of the following link where a continuous-time integral is approximated by h * sum. The quantity h plays the same role as our delT here. en.wikipedia.org/wiki/Rectangle_method Hope that helps! Adam
I need learn about linear time invariant systems ( LTI ), finite impulse response ( FIR ) and IIR ( infinite impulse response ) can you help me? thank you very much
Check out and subscribe to my channel: ua-cam.com/users/agpanagos I have about 200 videos, many of which are related to continuous-time and discrete-time signals and systems. I also continue to add videos as I teach each semester. Hope you find something useful.
![Convolution in Matlab and Python | 3 Modes of Convolution [DSP #11]](http://i.ytimg.com/vi/9yVowuBuASQ/mqdefault.jpg)
You're an absolute master, my friend. Just saved me several hours of reading forums.
Glad I could help, thanks for watching. Make sure to check out my website adampanagos.org for additional content (600+ videos) you might find helpful. Thanks, Adam
Thank you...This is really good tutorial,have been struggling with this for long ,you can explain really well.
Thanks!
your explanation is quite excellent..... really it helped me in my signals and systems lab....hoping for more videos like this on different topics in (s and s)......thank u very much adam................
That was greaattttt....It was just what I was looking for!!! thanks for uploading!
Thank you very much, you really helped me
Glad I could help, thanks for watching. Make sure to check out my website adampanagos.org for lots of additional content that you might find helpful.
You are amazing . Thanks
+Wawawy1994 Glad to help, thanks for watching!
i need a code for "discrete" convolution by building the function by your self not using the "conv" built in function , can you help?
Thank you!
10x a lot, fellow
In the 20th line of code, what have you entered at the beginning of the bracket? is it 'goa'?
It's "gca", which stands for "get current axis". Basically, I'm wanting to make change to the axis to make it bold, larger font, etc.
How does the "median(diff(t))" in line 27 make the discrete convolution become the continuous time convolution?
+Junchao Lu Keep in mind that Matlab (and computers in general) can ONLY work with discrete signals. Storing a continuous-time signal (even for a short amount of time) would require an infinite amount of memory since a continuum has an uncountably infinite number of points.
So, when calling the "conv" function in Matlab, the signals passed in (such as the variable x) are always vectors of a finite length and Matlab actually performs discrete-time convolution. This is a multiplication and summation of the vectors.
The quantity delT = median(diff(t)) is just a scalar and is essentially equal to the "delta t" of the time vector used in the problem. Multiplying by delT normalizes the summation that Matlab did inside the "conv" function to account for the sampling interval used in the problem to yield a final value that is equal to the area one would have obtained doing a real continuous-time integration.
You can think of the multiplying and summation performed inside the "conv" function as doing a rectangle method approximation of the real integral we'd like to work. The product of the signals is the function/rectangle height, delT is the width, and the product of these two provides the rectangle area which should closely approximate the continuous-time integral/convolution.
See the equation near the top of the following link where a continuous-time integral is approximated by h * sum. The quantity h plays the same role as our delT here.
en.wikipedia.org/wiki/Rectangle_method
Hope that helps!
Adam
+Adam Panagos thank you! and happy Chinese New Year!:)
can we reduce this code adam?
I need learn about linear time invariant systems ( LTI ), finite impulse response ( FIR ) and IIR ( infinite impulse response ) can you help me? thank you very much
Check out and subscribe to my channel: ua-cam.com/users/agpanagos
I have about 200 videos, many of which are related to continuous-time and discrete-time signals and systems. I also continue to add videos as I teach each semester. Hope you find something useful.
can i have the code link
Sure, you can download the m-file I used in this example here:
drive.google.com/open?id=0B1ClB5V-5p_uREluSUdERkZpckk&authuser=0
thank you