Home Propeller Head Plaza

Technical and scientific discussion of amps, cables and other topics.

[ Follow Ups ] Thread: [ Display  All  Email ] [ Propeller Head Plaza ]
[ Alert Moderator ]

Re: How right? Don't challenge my aural intelligence, pls.

I've done a little test once:
Somebody on another forum recorded a hihat and there were 5 files all in 16/44.1, all recorded using the same equipment (ie same mic, same mic pre, same convertor).
The difference was that one or more were recorded straight at 16/44.1 whilst one or more were recorded 24/96 and then dithered down to 16bit and resampled to 44.1.
The files were not marked in any way, the only way to tell any difference was by listening.
And despite the resulting files all being 16/44.1 the difference was subtle yet clearly audible.
If its recorded in 24/96 it does sound more realistic even if it ends up on a redbook cd.

I guess there are some misunderstandings about Nyquists theorem.
It states that it is POSSIBLE to recreate a sampled waveform accurately if the sampling frequency is at least twice the highest signal frequency.
This does NOT mean that accuracy is guaranteed!
Accuracy is only guaranteed if the peaks of the original waveform coincide precisely with the instant the sample is taken. This instant occurs roughly halfway thru the sampling process and is very much shorter then a 1/44100 of a second.
Basically the sampling circuit consists of a switch and a capacitor and during the process it goes through three discrete steps: Firstly the switch opens to drain the previous charge, secondly the actual sample is taken and thirdly the switch closes and the voltage value is read out. All these steps each take a finite amount of time which means that the sample is not averaged over 1/44100 sec as this is technically impossible.
Four samples per waveform period are the minimum to reproduce the waveform with any degree of accuracy, eight would be better otherwise the sum truncation error can become very large indeed! Unfortunately this problem does not show up on standard distortion measurements as this only shows up anything that has been added (ie output - input = distortion). Since sampling will always leave things out rather then adding the resulting distortion would be negative (provided everything works perfectly which it never does).
That said Mister Lavry produces some of the best medium sampling rate convertors around, still I side with George Massenburg on this one.

This post is made possible by the generous support of people like you and our sponsors:
  Parts Connexion  

Follow Ups Full Thread
Follow Ups
You can not post to an archived thread.

[ Top Top of Page ] [ Contact Us ] [ Support/Wish List ] [ Copyright © 2024, Audio Asylum, all rights reserved ]
[ General ] [ Speakers ] [ Tubes ] [ Vinyl ] [ Digital ] [ Hi-Rez ] [ Video Asylum ] [ Cables ] [ Tweaks/DIY ] [ Music ] [ Films ]