3 ms·
I would like to add the maybe counterintuitive fact (for people that never looked into that theory) that for "bandlimited signals", i.e. in your CD example a pi
by anyfoo 5y ago
I would like to add the maybe counterintuitive fact (for people that never looked into that theory) that for "bandlimited signals", i.e. in your CD example a piece of music that does not have any content above 22.05kHz to begin with, that reproduction of audio will be perfect, i.e. look exactly the same as the input signal. (Well, assuming ideal filters and analog components, which isn't really possible, but you can basically get "arbitrarily close" to the representation of the original signal with your digital 44.1kHz signal assuming a bandlimited signal by choosing better and better analog components.)