I'm from Germany and this channel is truely worth gold. Thanks for all the effort you also very often, very well presented clearly. Do you also have videos on coding theory e.g. linear block codes (syndromes, generator matrices), Huffman, Fano, Reed Solomon or convolutional codes?
Thanks for the suggestions, I don't have much on coding at the moment. I'll put these topics on my "to do" list. You can check out all my videos in categorised listing at iaincollings.com I've got a basic video on convolutional coding, and some on decoding: "What is a Convolutional Code?" ua-cam.com/video/EgYKMDBj_zQ/v-deo.html and "Decoding Convolutional Codes: The Viterbi Algorithm Explained" ua-cam.com/video/IJE94FhyygM/v-deo.html
Basically FT is simply your "shopping list" of all cos functions (Power vs. freq) which you need for a given signal (Power vs. time). The more input signal looks like cos-form, the less "exotic" cos functions you have on that list. Extreme example here is cos(t) itself - just one freq. (plus its -f "mirror"). The more "wild" the input signal looks, and less like cos, the more variety of cos functions you require. Opposite example is some random non-periodic pulse - you need a lot of cos functions, and they form the "spectrum shape" (Power vs. frequency) which looks like a solid line, but this is just a sum of endless frequencies of those cos functions. This is how I see it. Thank you.
Good intuitive explanation. I'm practice when we deal with signals that are time limited, we should see in the spectrum a series of sincs, centered around multiples of 1/T
I think you're confusing things. You say that "when we deal with signals that are time limited, we should see in the spectrum a series of sincs, centered around multiples of 1/T" ... but that's not true. Signals that are time-limited have a spectrum that is frequency-unlimited. eg. the rect function has a Fourier transform that is a sinc, that goes from negative infinity to positive infinity ... but it doesn't repeat. You seem to be mixing this up with the situation when signals are time-sampled with a sampling period T - then their frequency response repeats at 1/T. You can find many videos on my channel related to sampling, and also to the Fourier transform of rect. A full categorised listing can be found here: www.iaincollings.com/signals-and-systems
@@iain_explains why doesn't the sinc repeat? If I apply a rectangular window, ie I multiply it in the time domain, that'd be convolving in the frequency domain with the transform of the periodic signal, that is discreet
This may be a strange request but could you made a video on "How to find an answer to a wireless concept ?". I have a feeling that sometimes a viewer might request something that you are not familiar with but you have been able to find the answer by yourself and explain it beautifully :)
Well, I'd say that the best way is to be curious, and always ask yourself "What", "How" and "Why", and if that doesn't get you the answer, then check out the videos on my channel 😁
So, the DC value can be calculated directly as an average in time-domain, without resorting to a DFT. (with proper scaling "2pi" based on the Fourier transform definition that has been used).
Great explanation, electronics and communication engineers are lucky to have people like Iain. Thank you!
I'm from Germany and this channel is truely worth gold. Thanks for all the effort you also very often, very well presented clearly.
Do you also have videos on coding theory e.g. linear block codes (syndromes, generator matrices), Huffman, Fano, Reed Solomon or convolutional codes?
Thanks for the suggestions, I don't have much on coding at the moment. I'll put these topics on my "to do" list. You can check out all my videos in categorised listing at iaincollings.com I've got a basic video on convolutional coding, and some on decoding: "What is a Convolutional Code?" ua-cam.com/video/EgYKMDBj_zQ/v-deo.html and "Decoding Convolutional Codes: The Viterbi Algorithm Explained" ua-cam.com/video/IJE94FhyygM/v-deo.html
Basically FT is simply your "shopping list" of all cos functions (Power vs. freq) which you need for a given signal (Power vs. time). The more input signal looks like cos-form, the less "exotic" cos functions you have on that list. Extreme example here is cos(t) itself - just one freq. (plus its -f "mirror"). The more "wild" the input signal looks, and less like cos, the more variety of cos functions you require. Opposite example is some random non-periodic pulse - you need a lot of cos functions, and they form the "spectrum shape" (Power vs. frequency) which looks like a solid line, but this is just a sum of endless frequencies of those cos functions. This is how I see it. Thank you.
Yes, that's one way to put it. Here's a video showing what you've said: "Visualising the Fourier Transform" ua-cam.com/video/U7ii8agAhIs/v-deo.html
you are best , there is always so much more to learn from your videos
Thanks for your nice comment. I'm glad you like the videos.
Hi Ian, thank you for your explanation! I struggled with frequency analysis in structural dynamics, thank god I got to find your UA-cam channel!
It's great to hear you've found the videos helpful!
Very intuitive explanation.. Excellent and 👍 for all your videos.. Appreciate the efforts in making these videos
Thanks. It's great to know that people like the videos.
So intuitive! I stumbled on this at work, and didnt quite know where to look for. Thanks 😊
Glad it was helpful!
Good intuitive explanation. I'm practice when we deal with signals that are time limited, we should see in the spectrum a series of sincs, centered around multiples of 1/T
And what does the phase look like of periodic signals by the way, that's not discrete is it?
I think you're confusing things. You say that "when we deal with signals that are time limited, we should see in the spectrum a series of sincs, centered around multiples of 1/T" ... but that's not true. Signals that are time-limited have a spectrum that is frequency-unlimited. eg. the rect function has a Fourier transform that is a sinc, that goes from negative infinity to positive infinity ... but it doesn't repeat. You seem to be mixing this up with the situation when signals are time-sampled with a sampling period T - then their frequency response repeats at 1/T. You can find many videos on my channel related to sampling, and also to the Fourier transform of rect. A full categorised listing can be found here: www.iaincollings.com/signals-and-systems
@@iain_explains why doesn't the sinc repeat? If I apply a rectangular window, ie I multiply it in the time domain, that'd be convolving in the frequency domain with the transform of the periodic signal, that is discreet
Which sinc are you talking about? Maybe this video will help: "Visualising the Fourier Transform" ua-cam.com/video/U7ii8agAhIs/v-deo.html
As usual, another great explanation!
Thanks. Glad you like the videos.
This may be a strange request but could you made a video on "How to find an answer to a wireless concept ?".
I have a feeling that sometimes a viewer might request something that you are not familiar with but you have been able to find the answer by yourself and explain it beautifully :)
Well, I'd say that the best way is to be curious, and always ask yourself "What", "How" and "Why", and if that doesn't get you the answer, then check out the videos on my channel 😁
thanks professor very informative lots to learn from this.
I'm glad you found it useful.
great explanation
thanks
Glad it was helpful!
thank you very much, this is really elucidating
Glad it was helpful!
Thanks a lot. To the upper right plot: Without that offset, the f=0 component would vanish, right? (That means a signal with a mean value of y(t)=0)
Yes, that's right.
So, the DC value can be calculated directly as an average in time-domain, without resorting to a DFT. (with proper scaling "2pi" based on the Fourier transform definition that has been used).
Tnx sir again
Thanks a lot, please can you make a video for the detailed steps for LTE transmitter (the form of signal at each stage) Thanks again.
Thanks for the suggestion. I've put it on my "to do" list.
Ist view