-
Quote:
Originally Posted by maxg
FLZ,
Thanks for the reply - I am now formally giving up on ever understanding this stuff!! The 65,535 points represent volume? OMG - er....wait a moment...volume = amplitude of the waveform...nope.
Yep, you got it!
-
Quote:
Originally Posted by E-Stat
Moving from the theoretical to the real world, however, we find compromises. While the DACs themselves may be 16 bit, the effective number of bits (ENOB) is always fewer. As the level of a signal drops, it is represented by fewer and fewer bits. The result is called quantization distortion - the point at which the DAC goes deaf. The "solution" is a novel scheme called dither. Although it sounds counterproductive, dither introduces noise intentially to the signal to keep the bottom from falling out and is later removed. While there are many flavors and the process is largely successful, I think that some of the apparent quiet and lack of low level detail from the RB standard is simply uncorrected amounts of this distortion, IMHO. I have high hopes for the high rez standards.
rw
You couldn't be more wrong. First, the DAC doesn't go deaf. Second, the proper term is quantitization error. Third, the 16 bit word is a constant and every bit represents every sample. Quantization error is the result of non linearities in the conversion process plus the uncertainty of the least significant bit(LSB). Fourth, you have the dither process backward. Dither IS noise that is introduced to mix with the quantization error signal and spread it's energy out over a large number of frequencies. Who said jitter is a bad thing.
Since you like to speak of the "real world", the vast majority of recordings do not exceed 40dB dynamic range. Leaving 6dB headroom, you'll still be 50dB above the noise floor. So in the real world, it is very unlikely that quantization error is going to be audible since the noise floor of the medium will be below the noise floor of the listening environment anyway.
-Bruce
-
Quote:
Originally Posted by FLZapped
Second, the proper term is quantitization error.
I've seen it both ways. Digital errors are distortion.
Quote:
Originally Posted by FLZapped
Fourth, you have the dither process backward. Dither IS noise that is introduced to mix with the quantization error signal and spread it's energy out over a large number of frequencies.
I though that is what I said.
rw
-
Quote:
Originally Posted by mtrycraft
Oh, aren't they multi channel formats?
Logic would follow that if design engineers on two continents believed that the RB "perfect sound forever" standard was sufficient, then both SACD and DVD-A would simply be multi-channel variants using the same resolution with extended playing time. They aren't. With nearly seven times the storage capacity of the CD, they no longer face the same limitations.
rw
|