I've been reading this post for a long while now and I have to say there are some things that everyone has missed. First, there are distortion mechanisms in digital that do not exist in analog LP, likewise there are distortion mechanisms in analog LP. Consequently, one cannot compare digital to analog without stating what they are comparing. As casually mentioned above in a few places, a very good turntable with good source will trash a cheap digital player with good source. Likewise, the reverse is also true. It isn't possible to compare pure analog to pure digital and make a consensus about which is better. What is worse, is I don't know of a single high end source material that was recorded in analog and digital simultaneously. Sheffield Labs did a few but I don't know if they were ever released since they were recorded in digital format that is not what we use today. If someone knows of one, please enlighten me.
Second, back in the day when digital was first designed, Sony had to go to Burr Brown to produce suitable DAC's since Japan didn't have the analog semiconductor processes to develop a suitable chip. The first thing BB told them was the sample rate was too low for a 20 KHz data rate. (Yes, I know about the 2x Nyquist limit, having published several technical papers on the subject.) What most folks don't know is the stipulation in the Nyquist limit - in order to sample only twice per bipolar waveform then the two samples must occur at the precise peak levels of the analog data stream. That is at the maximum positive peak and maximum negative peak. So, that means that the ADC will know when the peaks will occur - yet it isn't possible for the ADC to know that since 20 KHz may or may not occur at any given sample period. For a random occurrence near the Nyquist limit, the sample rate must be at least 5 to 7 times the maximum frequency (there are a number of papers published on this fact by Analog Devices, National Semiconductor, Burr Brown, and Linear Technology as well as others). That puts the ideal sample rate at about 3 to 4 KHz. But, in order to fit the amount of music onto a the predefined disk size, significant compromises were made.
Lastly, as we all know, there are a number of distortion mechanisms in analog LP. These are just as destructive as the ones in digital and we perceive them differently. There is no point in me repeating them all here.
If someone on this board is going to make a blanket proclamation that digital is better or analog is better regardless of what they play it on, then I argue that they prefer certain types of distortion mechanisms over the others.
There is no question that digital is getting better year by year as it has much growth potential in terms of sound quality. Analog LP technology is mature, so the growth potential for sound quality is slower and harder to come by. With digital downloads, there is great potential to eliminate the limitations of the standard Compact Disk. 32 bit data and 384 KHz is already in experimentation stages and I personally anticipate 32 bits at twice that rate in the future. With advanced digital signal processing, some of the early destructive distortions in digital can be reversed, although the resulting data file will no longer fit on the original CD, it will be a download only format.
Second, back in the day when digital was first designed, Sony had to go to Burr Brown to produce suitable DAC's since Japan didn't have the analog semiconductor processes to develop a suitable chip. The first thing BB told them was the sample rate was too low for a 20 KHz data rate. (Yes, I know about the 2x Nyquist limit, having published several technical papers on the subject.) What most folks don't know is the stipulation in the Nyquist limit - in order to sample only twice per bipolar waveform then the two samples must occur at the precise peak levels of the analog data stream. That is at the maximum positive peak and maximum negative peak. So, that means that the ADC will know when the peaks will occur - yet it isn't possible for the ADC to know that since 20 KHz may or may not occur at any given sample period. For a random occurrence near the Nyquist limit, the sample rate must be at least 5 to 7 times the maximum frequency (there are a number of papers published on this fact by Analog Devices, National Semiconductor, Burr Brown, and Linear Technology as well as others). That puts the ideal sample rate at about 3 to 4 KHz. But, in order to fit the amount of music onto a the predefined disk size, significant compromises were made.
Lastly, as we all know, there are a number of distortion mechanisms in analog LP. These are just as destructive as the ones in digital and we perceive them differently. There is no point in me repeating them all here.
If someone on this board is going to make a blanket proclamation that digital is better or analog is better regardless of what they play it on, then I argue that they prefer certain types of distortion mechanisms over the others.
There is no question that digital is getting better year by year as it has much growth potential in terms of sound quality. Analog LP technology is mature, so the growth potential for sound quality is slower and harder to come by. With digital downloads, there is great potential to eliminate the limitations of the standard Compact Disk. 32 bit data and 384 KHz is already in experimentation stages and I personally anticipate 32 bits at twice that rate in the future. With advanced digital signal processing, some of the early destructive distortions in digital can be reversed, although the resulting data file will no longer fit on the original CD, it will be a download only format.