.
All about dat Hertz.
https://www.engadget.com/2019/05/04/...ampling-works/
Printable View
.
All about dat Hertz.
https://www.engadget.com/2019/05/04/...ampling-works/
Doesn't even begin to explain what's going on, and neither do other articles.Quote:
Originally Posted by SEAWOLF97
Fine on the sample rate..., but what happens next?
How is that sample quantified into a series of 1's and 0's and how do the 1's and 0s' relate? Does a square wave have the same 1's and 0's as a sine wave of the same freq (obviously not), and what of complex waves, and how were these values assigned to the 1's and 0's?
How do the 1's and 0's relate mathmaticaly such that they can be manipulated by sofware through various equations?
Are the 1's and 0's used for video or other formats similar/duplicate, with other encoding to describe the format, to those of audio quantification, knowing that there are only so many 1's and 0's combinations available at so many bits?
Inquiring minds want to know??? :banghead:
When they wonder, they'll do a web index search which will reveal all.
It all does come back to Harry Nyquist. As no one has so far ever proved him wrong about anything, though long deceased he is the number one dispeller of audiophile nonsense/myths/horsecrap. Neither audio nor electrical engineers have any trouble with this, but that leaves nearly everyone else to spend what they are spending on gear that delivers profits instead of superior sound. If searching the web is not your thing, pony up for The Audio Expert by Ethan Winer. You'll never be taken in again, and you will be able to ANSWER all these questions.
As an aside about the ones and zeros questions, a real World example. Sony developed and authored the digital methods of reproducing music (the CD, SACD, DSD and Blu-ray programming) and two-thirds of the data on an audio CD is error correction software. How good were they? No functioning player has ever output an incorrect bit. That's the power of audio software. The hardware (disc or player) can be damaged but the software never introduces errors.
https://www.youtube.com/watch?v=cIQ9IXSUzuM
One of my favourite (and technically correct) explanations, with real world examples, in under 25 minutes.
I remember stumbling across this man's work years ago. I couldn't agree more, a great treatment. I took away a great deal from this offering from (I think) the same team:Quote:
Originally Posted by Mitchco
https://people.xiph.org/~xiphmont/demo/neil-young.html
A good video, but does not address what I asked about in the least (e.g., once the signal is sampled, how is that sample assigned a "1" "0" pattern and how can that pattern be mathmatically manipulated to produce basically any other pattern (e.g., change pitch).Quote:
Originally Posted by Mitchco
Also, I have a problem with the Nyquist Principal and he demonstrates why at 6:47 but glosses over it. He shows a 20 kHz sine wave and samples HOW MANY TIMES??? to make it into a sine wave pattern. But what if the 20kHz sine wave only lasted for 1/10,000th of a second (and who's to say that it can't)??? He would only have the first four or five samples and in no way does this represent a complete sine wave.
Another problem with the Nyquist method that I've never seen addressed, is sampling white noise.
Assuming the white noise source consists of all frequencies from 20 Hz-20kHz there is more than a 50% chance that the frequency being sampled is between 10kHz (=4.4 samples per complete cycles per second) and 20kHz (=2.2 samples per cycle).
All frequencys in white noise are not produced simultaneously, and the noise is really "microbursts" of a frequency. If you turn on the "sample and hold" on a synthesizer and sample white noise you get a random series of individual tones as it samples each voltage of the waveform.
How long are these microbursts before the next microburst at a different frequency, knowing that the odds are in favor of fewer than 4.4 samples per cycle??? (Never seen it printed anywhere.) Obviously, we would get "noise" but, how do we know that the digitized white noise is an exact reproduction of the original? Obviously someone figured that at that point noise is noise, few actually listen to white noise, and no one would know the difference.
https://youtu.be/m9auf7dpMrU