I've been studying dsp for a couple of years and your explanations are some of the bests I've found on youtube. Simple, short and understandable, thank you for your contribution!
This is the first time I have read such a clear explanation of Fourier series mapping from real concepts to mathematical formulas. thank you very much.
I suggest you can put a warning to tell audience you will play a high frequency music(my ears hurt a lot while wearing headphone in a high volume),By the way,what a fantastic video.--comment from a chinese student
Great intro to signal understanding. I was lucky enough to study electroacoustic music composition in my youth and exploring the synth output on an oscilloscope was best physics course I ever took. :)
Great video. At 08:00, why did you use 2f0 and 4f0 rather than 3f0 and 5f0 to reconstruct/decompose the signal? I presume it could also have been expressed perfectly in terms of 3 and 5 rather than 2 and 4? Or not? Is it therefore also possible to perfectly reconstruct any signal using f0, 2f0, 4f0, 6f0, 8f0, etc.? I imagine this would often require an extra frequency since we have started with a small gap between 1f0 and 2f0 rather than the allowed gap of two fundamentals between 1f0 and 3f0? Or does this just not work? Thanks!
Actually, you need all of the harmonics (multiples of f0) to guarantee reconstruction. If you go with 2f0, 4f0, 6f0, you will only be able to reconstruct signals with a fundamental frequency of 2f0 (and fundamental period of 1/2f0). In the example at 08:00, it just happens that signal contains no contribution at 3f0 (I intentionally did that to show how the cos() and sin() contributions cancel each other exactly at that frequency). Hope this is helpful!
Great Tutorial! Just a comment: the sound volume of the samples you present is higher than that of your voice. In my opinion, it should be the opposite. We need to hear what you are saying, and the samples are just illustration they don't need to be that strong (so they don
@@bigmistqke Actually, I tried manim, but I couldn’t easily do the things I wanted to for signals (at least not in the way I imagined). I use matplotlib to generate individual plot frames, combine those into videos, and then put them all together in a macOS Keynote presentation. I appreciate the comparison to 3blue1brown… they’re the gold standard! 🤩
ive never see dsp like this before
this is a hidden treasure
I've been studying dsp for a couple of years and your explanations are some of the bests I've found on youtube. Simple, short and understandable, thank you for your contribution!
by far the best explanation of the fourier I have seen. Thanks for making this
So glad I discovered this playlist!
just. wow. pedagogically this is outstanding!
This man really made studying for dsp fun hahah thanks for these videos very helpful in understanding key concepts
This is the first time I have read such a clear explanation of Fourier series mapping from real concepts to mathematical formulas.
thank you very much.
I suggest you can put a warning to tell audience you will play a high frequency music(my ears hurt a lot while wearing headphone in a high volume),By the way,what a fantastic video.--comment from a chinese student
Sir you are amazing. Thank you very much.
Great intro to signal understanding. I was lucky enough to study electroacoustic music composition in my youth and exploring the synth output on an oscilloscope was best physics course I ever took. :)
amazing video !! like your explains very much :)
It’s called T because the other synonym of Period is Tour (in french)
3:00 why
great and simple explanation!
why does this have 6k views? deserves millions…
Great video. At 08:00, why did you use 2f0 and 4f0 rather than 3f0 and 5f0 to reconstruct/decompose the signal? I presume it could also have been expressed perfectly in terms of 3 and 5 rather than 2 and 4? Or not? Is it therefore also possible to perfectly reconstruct any signal using f0, 2f0, 4f0, 6f0, 8f0, etc.? I imagine this would often require an extra frequency since we have started with a small gap between 1f0 and 2f0 rather than the allowed gap of two fundamentals between 1f0 and 3f0? Or does this just not work? Thanks!
Actually, you need all of the harmonics (multiples of f0) to guarantee reconstruction. If you go with 2f0, 4f0, 6f0, you will only be able to reconstruct signals with a fundamental frequency of 2f0 (and fundamental period of 1/2f0).
In the example at 08:00, it just happens that signal contains no contribution at 3f0 (I intentionally did that to show how the cos() and sin() contributions cancel each other exactly at that frequency). Hope this is helpful!
👌
Great Tutorial! Just a comment: the sound volume of the samples you present is higher than that of your voice. In my opinion, it should be the opposite. We need to hear what you are saying, and the samples are just illustration they don't need to be that strong (so they don
I'll try to do a better job of balancing those in future videos 🙂
Oh my ear!!
What tools do you use to make these videos?
Most likely manim, that library made by 3blue1brown.
@@bigmistqke Actually, I tried manim, but I couldn’t easily do the things I wanted to for signals (at least not in the way I imagined). I use matplotlib to generate individual plot frames, combine those into videos, and then put them all together in a macOS Keynote presentation. I appreciate the comparison to 3blue1brown… they’re the gold standard! 🤩
Just a simple nitpick: 1 does not represent nothing. 1 is 1 unit out of T units. 0 would be nothing.
don't skip the intro