View Single Post
  #46  
Old 2008-02-19, 04:51 PM
FalloutBoy FalloutBoy is offline
 
Join Date: Feb 2008
Re: Vinyl records vs. Cds.

Quote:
Originally Posted by Tubular View Post
Quote:
We can deduce from this that the more discrete values available to us, the better it will sound.
Do you disagree with this?
It is true in some cases, depending on the signal being represented. In other cases it will sound the same.

It also depends on other factors, such as sample rate, dither and noise shaping.

Quote:
When I say more accurate I mean a more faithful representation of the live waveform
That accuracy is determined by the sample rate.

Nyquist Sampling Theory: A sampled waveforms contains ALL the information without any
distortions, when the sampling rate exceeds twice the highest frequency contained by the
sampled waveform.

Here is a good a paper on the sampling theory by Dan Lavry (developer of professional A/D and D/A equipment):
http://www.lavryengineering.com/docu...ing_Theory.pdf

Quote:
Originally Posted by Tubular View Post
Quote:
In order to make this sound better, we need to be able to have discrete values in between these values. A fading piece of music canít just go from very audible to silent, or it wouldnít be a smooth fade. A 4-bit recording would have 16 discrete possible amplitude levels. Can you again imagine what this would sound like? Definitely better, but its still a totally horrible representation of the sound. We can deduce from this that the more discrete values available to us, the better it will sound. Is there a limit to the human earís ability to perceive these inaccuracies? Definitely, but it unfortunately does not stop at the 65536 discrete values afforded to us by 16-bit technology.
What about this do you disagree with, FalloutBoy?
I think it is highly misleading, as you can't possible deduce anything about how a recording sounds just from the bit depth.

It also doesn't mention the nature of the "inaccuracies" (quantization errors) it mentions.

And it doesn't even mention dither and noise shaping which makes the conclusions drawn moot.

Quote:
With more bits per sample, the digital representation of the live waveform's amplitude can be smoother and less square and stairstepped.
True. More bits per sample -> less quantization error -> lower noise floor.
Reply With Quote Reply with Nested Quotes