Has anyone listened to the NAD C 390DD?

[darrenyeats]

If you feed a 16-bit computer generated -90dB undithered 'sine' wave, you're essentially feeding into the DAC a 1kHz (in this case) 1-bit *square* wave. This is exactly what the measurements show.

Despite my doubts and musings yesterday, I don't think so. What is the difference between a sine wave and a square wave at -90db in 16 bit? None. A 16/44 DAC should smooth this analoguely (I think this sinc function thingy is about this). A 24/96 DAC would upsample and interpolate the data points of this tone into a smoother shape (because the DAC assumes it's NOT a square wave - a square wave would require an infinite series of harmonics yet we know the data is band limited to satisfy Nyquist).

Please see here:

Phil,
See these posts:
Submitted by Miska on Sun, 02/12/2012 - 17:43.
Submitted by Miska on Mon, 02/13/2012 - 02:48.
Here: http://www.computeraudiophile.com/co...2496-not-24192 (search for 'miska')

Darren

[Phil Leigh]

Despite my doubts and musings yesterday, I don't think so. What is the difference between a sine wave and a square wave at -90db in 16 bit? None. A 16/44 DAC should smooth this analoguely (I think this sinc function thingy is about this). A 24/96 DAC would upsample and interpolate the data points of this tone into a smoother shape (because the DAC assumes it's NOT a square wave - a square wave would require an infinite series of harmonics yet we know the data is band limited to satisfy Nyquist).

Please see here:

Darren

That miska stuff is bollocks. he's using audacity plots to try to prove his incorrect point. Audacity has major bugs int the way it renders waveforms on screen.
If you put a real oscilloscope across a DAC you get to see the truth and it doesn't look like ANY of his plots!!!

A 10khz sine wave looks very fine to me coming out of my DAC, regardless of sampling rate or bit depth (assuming that it's level is NOT -90dB!)

DACs don't have intelligence and don't make contextual judgements. They just do what they do.


With a one bit symmetric waveform you can keep the bit on or turn it off. If you keep it on for a few samples then off for the same number of samples you have an approximation of a square wave ( which of course the analogue parts of the DAC cannot reproduce properly because of filter ringing). If you turn it alternately on or off for a single sample you get the approximation of the sine wave...which the filter will smooth...

If you only have one bit there isn't really much else you can do with it! - and the DAC cannot have any understanding of what you were trying to do with it :-)

[darrenyeats]

That miska stuff is bollocks. he's using audacity plots to try to prove his incorrect point. Audacity has major bugs int the way it renders waveforms on screen.
If you put a real oscilloscope across a DAC you get to see the truth and it doesn't look like ANY of his plots!!!

A 10khz sine wave looks very fine to me coming out of my DAC, regardless of sampling rate or bit depth (assuming that it's level is NOT -90dB!)

DACs don't have intelligence and don't make contextual judgements. They just do what they do.


With a one bit symmetric waveform you can keep the bit on or turn it off. If you keep it on for a few samples then off for the same number of samples you have an approximation of a square wave ( which of course the analogue parts of the DAC cannot reproduce properly because of filter ringing). If you turn it alternately on or off for a single sample you get the approximation of the sine wave...which the filter will smooth...

If you only have one bit there isn't really much else you can do with it! - and the DAC cannot have any understanding of what you were trying to do with it :-)

I think we're agreeing mostly again? LOL.

Except on for x samples and off for x samples is surely just a different frequency sine... (On for one sample and off for one sample being a sine at fs/2.)

Audacity has major bugs int the way it renders waveforms on screen.

Read all his posts all the way through - he explains that very thing.
Darren


Edit: in case anyone's reading this thread a long time later - I've corrected my own misunderstanding here: http://forums.slimdevices.com/showth...l=1#post856180

[Stereoeditor]

I was thinking of these (figure 5 in each case):

http://www.stereophile.com/content/b...r-measurements

http://www.stereophile.com/content/s...r-measurements

... It's not clear to what Stereophile think these measurements prove in terms of sound quality? - my guess is nothing, so they are only meaningful as an engineering benchmark in comparison with other DACs.

Apologies for the belated response. I offer this measurement, of an undithered 16-bit tone at exactly -90.31dBFS, because it easily reveals DAC linearity problems.

Perhaps to simplify overmuch, in the 2's-complement encoding used by16-bit LPCM, the two transitions shown, from the zero line to +1LSB's voltage and -1LSBs voltage, actually involve maximally asymmetric bit changes in the digital data presented to the DAC: one from all bits = 1 to all bits =0; the other from all bits off to just the LSB = 1. The fact that with a good DAC, the voltages are equal and the waveform is symmetrical tells you that the summed errors in the 15 MSB transitions are very much lower than the LSB transition, which is what you want.

With the vast number of DACs now sigma-delta types, this test of obsolete, But I still include it because it also shows you at a glance how high above the DAC's noisefloor the LSB transition lies.

John Atkinson
Editor, Stereophile

[prutten]

Getting back to the OP's question

I'm curious as this is a "digital" amp.

According to the NAD website "C 390DD has no analogue stages in the signal path, keeping music in the digital domain right up to the speaker outputs. All preamp functions are executed in the digital domain without the phase shift, noise and distortion that plagues all analogue designs regardless of price or pedigree."

It sounds like it would work well with a Squeezebox, as it is a DAC, pre-amp and power amp all in one.

Hi guys,

I just stumbled across this thread today and noted it had gotten quite detailed on the technical issues of digitization (which is rather applies to all digital amplification), and wanted to rather respond to the OP's question on the NAD c390dd. I have listened to the c390dd now a lot - firstly next to other amplifiers and also now at home ;-)

I can truly recommend it strongly from a listening perspective. I felt the amplifier doesn't have any fuzziness, no problems with project a wide soundstage even in complex pieces of music at strong volumes. I really like its transparency - you can hear everything. Now it might not have the charm and warmth of some tube amplifiers, but that warmth to me is often rather taking away from the original sound as it was meant to be, but hey, that's just my opinion. I found that the description of the M2 on Stereophile pretty much matches my experience listening to the NAD - so at least perception-wise I feel I am getting great value for money.

Also, what I liked is that this amplifier just invites me to listen a lot more. I don't get tired by extraneous sounds, additions.

As an engineer myself, I appreciate the beauty of just have a streamer and a straight digital path through the system - this is perhaps not a listener's concern, but it certainly is a cool feature

What I also liked about the product that - in good NAD style - it is just very pragmatic, offering connections for both my streamer, 2 speaker home cinema, my PC, a second DVD player, etc.

All in all - I would definitely recommend you to go out and listen to this amplifier!

Kind regards,

Paul
PS I listened to the NAD c390dd in conjunction with both KEF, Dynaudio, Dali and Sonus Faber speakers. I eventually settled on the Sonus Faber Auditor Ms - that's another story of course.

[SBGK]

Hi guys,

I just stumbled across this thread today and noted it had gotten quite detailed on the technical issues of digitization (which is rather applies to all digital amplification), and wanted to rather respond to the OP's question on the NAD c390dd. I have listened to the c390dd now a lot - firstly next to other amplifiers and also now at home ;-)

I can truly recommend it strongly from a listening perspective. I felt the amplifier doesn't have any fuzziness, no problems with project a wide soundstage even in complex pieces of music at strong volumes. I really like its transparency - you can hear everything. Now it might not have the charm and warmth of some tube amplifiers, but that warmth to me is often rather taking away from the original sound as it was meant to be, but hey, that's just my opinion. I found that the description of the M2 on Stereophile pretty much matches my experience listening to the NAD - so at least perception-wise I feel I am getting great value for money.

Also, what I liked is that this amplifier just invites me to listen a lot more. I don't get tired by extraneous sounds, additions.

As an engineer myself, I appreciate the beauty of just have a streamer and a straight digital path through the system - this is perhaps not a listener's concern, but it certainly is a cool feature

What I also liked about the product that - in good NAD style - it is just very pragmatic, offering connections for both my streamer, 2 speaker home cinema, my PC, a second DVD player, etc.

All in all - I would definitely recommend you to go out and listen to this amplifier!

Kind regards,

Paul
PS I listened to the NAD c390dd in conjunction with both KEF, Dynaudio, Dali and Sonus Faber speakers. I eventually settled on the Sonus Faber Auditor Ms - that's another story of course.

there is an interesting thread in an australian forum which is comparing the M51 and c390dd, some think the sound gets warmer after burn in.

[Daverz]

I'll definitely put this on my wish list if it

* Runs cool (or at least cooler than my Bryston).
* Can do the first-order (6 dB/octave) high-pass filter that my Vandersteen speakers need. (Yeah, I've done this with sox on the server, but I'd rather not have the filter at the source.)

[prutten]

Running cool - yes

I'll definitely put this on my wish list if it

* Runs cool (or at least cooler than my Bryston).
* Can do the first-order (6 dB/octave) high-pass filter that my Vandersteen speakers need. (Yeah, I've done this with sox on the server, but I'd rather not have the filter at the source.)

hi Dave,

This amplifier definitely runs cool - it is a class D design after all with very high efficiency. I don't know precisely how high, but simply put, the amplifier does not become more than hand-warm even after running for extended periods.

Unfortunately, I don't have enough technical expertise to judge the high-pass filter - someone else will have to help you on that one.

Kind regards,

Paul