My knowledge of these two is somewhat sketchy, but DSD is a one-bit approach, employing pulse-density modulation at a (1 bit) symbol rate of 2.8MHz (64X 44k1). NAD use a 7bit symbol at a rate of 16X 44k1.
NAD's scheme is based on hardware from Diodes Inc (formerly Zetex) and is a signal chain designed for a power output stage. Thus it uses PWM rather than PDM to minimize the number of transitions. Using DSD to drive a power output stage would incur much higher losses due to the high rate of transitions. In contrast NAD's approach has transitions at a 16X rate, but each symbol has a 'quantum of width' around 9nS from a 108MHz clock.
Given that Stanley Lipshitz has shown that DSD is a flawed format because its impossible to apply enough dither to linearize the single bit quantizer, NAD's approach appears to be superior because its by nature multibit, rather than single bit. However they're designed for different applications - there would be no point in storing PWM signals on an optical disk for example.
NAD's scheme is based on hardware from Diodes Inc (formerly Zetex) and is a signal chain designed for a power output stage. Thus it uses PWM rather than PDM to minimize the number of transitions. Using DSD to drive a power output stage would incur much higher losses due to the high rate of transitions. In contrast NAD's approach has transitions at a 16X rate, but each symbol has a 'quantum of width' around 9nS from a 108MHz clock.
Given that Stanley Lipshitz has shown that DSD is a flawed format because its impossible to apply enough dither to linearize the single bit quantizer, NAD's approach appears to be superior because its by nature multibit, rather than single bit. However they're designed for different applications - there would be no point in storing PWM signals on an optical disk for example.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.