at 33.57 i think we get wh*L instead of Wh/L because it is scaling property if you decrease time domain then you will increase frequency domain can any one have a ans please tell me.
at 33.57 i think we get wh*L instead of Wh/L because it is scaling property if you decrease time domain then you will increase frequency domain can any one have a ans please tell me.
I stand corrected.....see the previous lecture about 24 minutes into it...the professor derives an FIR recursively....similarly it is clear that at IIR can be derived non recursively...
at 33.57 i think we get wh*L instead of Wh/L because it is scaling property if you decrease time domain then you will increase frequency domain can any one have a ans please tell me.
at 33.57 i think we get wh*L instead of Wh/L because it is scaling property if you decrease time domain then you will increase frequency domain can any one have a ans please tell me.
He went by the basic DTFT formula but he went term by term (time intants -2,-1) until he got to the part where he already had a tranform availble (0 onwards).
at 33.57 i think we get wh*L instead of Wh/L because it is scaling property if you decrease time domain then you will increase frequency domain can any one have a ans please tell me.
+Deepjyoti Das because it is 2pik and k varies from - infi to + infi so this will be -infi *2 pi,....,pi,0,pi,....infi *2 pi, but now we are looking at the baseband, our interest of vision is in between -pi to +pi (integration limit) and in this k has no value as it is in discrete form it can only attain integer so in between this every time k will be zero intuitively.
at 33.57 i think we get wh*L instead of Wh/L because it is scaling property if you decrease time domain then you will increase frequency domain can any one have a ans please tell me.
at 33.57 i think we get wh*L instead of Wh/L because it is scaling property if you decrease time domain then you will increase frequency domain can any one have a ans please tell me.
Your argument is rit but,, actually you are misinterpreting that we are compressing the signal in time domain in an upsampler because of n/L term in x(n/L).The value we are getting at say n=3 i.e. x(3) after upsampling x(n) we will get x(3) at n=3L as puting =3L in x(n/L) gives x(3)(L>1).So actually an upsampler streching out the signal not compressing it.
Yes try the equation of an accumulator --- y[n]= Sum of x[n-k] where k goes from 0 to infinity. This is an IIR filter implemented non-recursively. Of course it is causal as well.
![ปฏิเสธไม่ไหว (Crush on you) - LIPTA feat. No One Else [Official MV]](http://i.ytimg.com/vi/NBw1jF342vw/mqdefault.jpg)
We do look your lectures Like a jeweller looks an ornament. Because they are pure gold.
thank you for your lecture I got PhD in BITS Pilani because of listening to your lecture
really??
what question did they asked u ? I am also going to give interview for phd
at 33.57 i think we get wh*L instead of Wh/L because it is scaling property if you decrease time domain then you will increase frequency domain can any one have a ans please tell me.
@@dhruvandangar9972 you are taking about downsampler, in upsample we increase time domain and hence shrink the Freq domain
n-> n/L
w -> w•L
NICE WAY OF TEACHING AND MAKES ANYONE UNDERSTAND EVEN IF THE LISTENER IS A NEW TO THIS FIELD OF DSP
My Concept is clear.Thank you NPTELhrd ❤
thanks to all nptel team ...awesome class............
The starting music is like that of Tom and Jerry lol. Good lectures :)
In the example that was give, there is a mistake in the sequence. It should be symmetrical x[n]= [1/4,1,7/4,0,-7/4,-1,-1/4]
i am also getting the values you mentioned
Same here
Yes...we will get these values only
at 33.57 i think we get wh*L instead of Wh/L because it is scaling property if you decrease time domain then you will increase frequency domain can any one have a ans please tell me.
Getting 3 in place of 7
I stand corrected.....see the previous lecture about 24 minutes into it...the professor derives an FIR recursively....similarly it is clear that at IIR can be derived non recursively...
at 33.57 i think we get wh*L instead of Wh/L because it is scaling property if you decrease time domain then you will increase frequency domain can any one have a ans please tell me.
Superb.. Sir !really awesome
the indian universities always have such class animations
BEST INTRO EVER
Why is upsampler your fav ?
55:15 X*(n)----X*(e^(-jw)) , missed a minus
no its correct
Best experience when u watch it at 1.25x speed
at 33.57 i think we get wh*L instead of Wh/L because it is scaling property if you decrease time domain then you will increase frequency domain can any one have a ans please tell me.
Watch at 1.5X to save time
Thanks man for sharing this! It is helpful.
@ 46:19, How did Professor Roy take the DTFT of the -2(alpha)^-2 and -1(alpha)^-1 term?
He went by the basic DTFT formula but he went term by term (time intants -2,-1) until he got to the part where he already had a tranform availble (0 onwards).
Look at the basic formula of DFT , for n=-2, put x[-2]= -2alpha^-2
at 33.57 i think we get wh*L instead of Wh/L because it is scaling property if you decrease time domain then you will increase frequency domain can any one have a ans please tell me.
what are the number o Complex multiplication and complex additions for 2D-FFT algorithm?
how we get sum of 1/n^2=pie/6
at 24.34 why k=0 for the range -pi to +pi
and where did the summation go
+Deepjyoti Das because it is 2pik and k varies from - infi to + infi so this will be -infi *2 pi,....,pi,0,pi,....infi *2 pi, but now we are looking at the baseband, our interest of vision is in between -pi to +pi (integration limit) and in this k has no value as it is in discrete form it can only attain integer so in between this every time k will be zero intuitively.
at 33.57 i think we get wh*L instead of Wh/L because it is scaling property if you decrease time domain then you will increase frequency domain can any one have a ans please tell me.
01:00
Respect ....
at 33.57 i think we get wh*L instead of Wh/L because it is scaling property if you decrease time domain then you will increase frequency domain can any one have a ans please tell me.
Your argument is rit but,, actually you are misinterpreting that we are compressing the signal in time domain in an upsampler because of
n/L term in x(n/L).The value we are getting at say n=3 i.e. x(3) after upsampling x(n) we will get x(3) at n=3L as puting =3L in x(n/L) gives x(3)(L>1).So actually an upsampler streching out the signal not compressing it.
Good evening sir.. Can i know what is the application which is using DTFT?
Speech processing, Image processing, and Digital signal processing
@@aniket4952 Thanks for the answer
@@rooyresh Wow. After 9 years it is still helpful? 😆
@@aniket4952 At least someone answered and it would help others too... haha
thank you
please update
radix-2 fft algorithm
is it just me or does this guy look like an indian Bill Murray?
I don't think an IIR filter can be implemented non-recursively, can anyone give me a counter-example?
Yes try the equation of an accumulator --- y[n]= Sum of x[n-k] where k goes from 0 to infinity. This is an IIR filter implemented non-recursively. Of course it is causal as well.
Did you know that if you take the Fourier Transform of milk you will get paneer?
haha clown you are
@alex3ff4rly I would have said Stevan Segal
Stephan Segal
1:00
no you will get molecules of milk :)
oru undayum manasilavunnilla
Superb.. Sir !really awesome