[shortcuttomonct]

 
Mr. shortcut,
 
Where on earth did you find an Audio Note DAC and which one?
 
Are you a Brit?
 
Tony in Michigan

Eastern Canada!  It is called a DAC 2.1 and was built from the online kit by a local guy - electrical engineer.  He did some upgrades to the standard AN kit but honestly, I am not sure what those were other than he spent about 500 bucks extra on the upgrades.
 
I`m pretty excited to give it a shot in my system when he goes away for the holidays.  

 

[judmarc]

  Eastern Canada!  It is called a DAC 2.1 and was built from the online kit by a local guy - electrical engineer.  He did some upgrades to the standard AN kit but honestly, I am not sure what those were other than he spent about 500 bucks extra on the upgrades.
 
I`m pretty excited to give it a shot in my system when he goes away for the holidays.  

 
The Audio Note kits don't use oversampling as far as I know.  Perhaps your EE friend changed that (unlikely), or he or you use computer software to oversample prior to the DAC.  If not, with Redbook (16/44.1) material, you'll get harmonic distortion, similar to the original CD players.  Many people like this, as subjectively it's a "hot" or "live" sort of sound.
 
Not trying to rain on anyone's parade, just so you know.
 
You might try listening to some higher resolution music, or using computer software to upsample Redbook prior to feeding it to the Audio Note, and seeing what you think of that.

 

[HemiSam]

I thought that defeated the primary purpose of using a NOS DAC?
 
HS

 

[judmarc]

  I thought that defeated the primary purpose of using a NOS DAC?
 
HS

 
Depends whether you want accuracy or a particular "sound" from your DAC. 

 

[Articnoise]

One big (if not the biggest) reason why AN changed from oversampling to no oversampling was to get rid of the digital filters. The Phasure nos1a does the oversampling/up sampling in the computer before deliver it to the dac and don’t use digital filters, if I have understand it right.

 

[judmarc]

One big (if not the biggest) reason why AN changed from oversampling to no oversampling was to get rid of the digital filters. The Phasure nos1a does the oversampling/up sampling in the computer before deliver it to the dac and don’t use digital filters, if I have understand it right.

Almost. It (the Phasure) substitutes digital filtering in the PC for digital filtering in the DAC. That, or using higher resolution files, are the options I suggested in order to use the Audio Note but avoid harmonic distortion.

Edit: Filtering and upsampling/oversampling are nearly synonymous, because all upsampling/oversampling must be accompanied by filtering.

 

[Articnoise]

Almost. It (the Phasure) substitutes digital filtering in the PC for digital filtering in the DAC. That, or using higher resolution files, are the options I suggested in order to use the Audio Note but avoid harmonic distortion.

Edit: Filtering and upsampling/oversampling are nearly synonymous, because all upsampling/oversampling must be accompanied by filtering.

 
The Light harmonic davinci dac doesn’t oversample or use digital filter and I haven’t heard people saying that it has harmonic distortion, so maybe it’s more to it

 

[judmarc]

 
 
The Light harmonic davinci dac doesn’t oversample or use digital filter and I haven’t heard people saying that it has harmonic distortion, so maybe it’s more to it

 
This is what the folks at Schiit Audio call "buzzword compliant."
 
From the Audiostream online review of the DaVinci (which I've heard at an audio show - not the best venue, granted - sounding great):
 

The Duet Engine utilizes analog interpolation with parallel output modules utilizing highly precise clocks to double the sample rate for CD 44.1/16 files. The Duet Mode does not utilize conventional upsampling or digital filters. Light Harmonic claims that a sophisticated aligned timing technology is used to double the effective playback sampling rate to 88.2kHz. Light Harmonic feels that the Duet Mode will provide smoother high frequencies without the negative effects of a “brick-wall filter”, and at the same time, provide bit perfect playback.

[Emphasis added.]
 
In other words, they use an analog rather than a digital oversampling filter.  Still an upsampling/oversampling/interpolation filter.  (Interpolation is the more technically proper term.)  This is, as they mention, to avoid the negative effects of "brick-wall" filtering, i.e., harmonic distortion, that would be required if the DAC truly did no interpolation of 44.1 rates.
 
The Light Harmonic marketing folks couldn't resist one more buzzword phrase, "bit perfect playback," at the end.  This is nonsensical in two ways: (1) As we've seen, there's interpolation via an analog filter.  The moment you filter, by definition you are no longer bit perfect.  (2) You can't be bit perfect at the *playback* end, because you're outputting analog, not bits.  By definition you've employed a reconstruction filter to convert the bitstream to analog.  (If you've ever listened to a DAC that loses lock, you know how nasty the unconverted bitstream sounds.)
 
What goes on inside DACs is the subject of a tremendous amount of marketing.  There are buzzwords to be avoided (digital, upsampling) and buzzwords to be included (analog, bit perfect), and technical accuracy takes a back seat.  There are only a tiny handful of true NOS DACs, among them the Audio Note and Phasure.  And as noted above, the Phasure is explicitly meant to be used with software that does the digital interpolation filtering in the computer.  All in all, you should remain skeptical that a DAC is truly NOS unless you know what the DAC chip(s) inside is/are and have looked at a spec sheet for it/them.
 
With regard to there being "more to it" - digital audio at Redbook resolution needs a "brick-wall" reconstruction filter that causes harmonic distortion.  This isn't a matter of better or worse filter design, or digital versus analog filters, it's sheer mathematics.  The only alternative (for 44.1 resolution) is the one that engineers turned to decades ago - interpolation/filtering to a higher rate, and only then using a reconstruction filter that can be more gently sloped due to the higher rate bitstream.

 

[jacal01]

If for a continous analog output signal, what is the filter interpolating between?  And I thought clocking was in the digital realm to minimize signal bit timing synchronizing jitter.

 

[judmarc]

  If for a continous analog output signal, what is the filter interpolating between?  And I thought clocking was in the digital realm to minimize signal bit timing synchronizing jitter.

 
Filtering takes place at least twice in all but true NOS DACs:
 
- The *interpolation* filter(s) to raise the sample rate of the bitstream.  By far the most common is 8x oversampling, which raises the bit rate to 352.8/384kHz in 3 or fewer rounds of doubling.  So, for instance: 44.1 -> 88.2 -> 176.4 -> 352.8; or 192 -> 384.
 
- The *reconstruction* filter, nearly always a relatively simple analog low pass filter to take the bitstream and convert it to analog.
 
In sigma delta DACs (i.e., all except the very few R2R DACs), there is a step between these, sigma delta modulation.  There, the sigma delta modulator (which is another filter) modulates the bitstream (usually to DSD-type rates, though in ESS chips the rate is much higher).
 
Edit: Oh, and clocking - that takes place before the interpolation filtering stage (or before each round of it, in case of multiple rounds of interpolation).  The bitstream runs from the DAC input to a buffer; the clocks then clock the bits *out of* the buffer and into the interpolation filtering stage.
 
Yet another edit: See the block diagram for PCM data on page 2 of this datasheet for the chip used in the original Bifrost (don't know if it still is): http://www.akm.com/akm/en/file/datasheet/AK4399EQ.pdf

 

[arnaud]

  If for a continous analog output signal, what is the filter interpolating between?  And I thought clocking was in the digital realm to minimize signal bit timing synchronizing jitter.

I am confused as well, interpolation means between discrete points, can only be done in the digital domain?!
 
Otherwise, @judmarc: I have had good experience with NOS / analog reconstruction "brickwall" filter DAC and I typically dislike oversampling it seems (although my current DAC may actually be oversampling 8x for all I know yet it does not suck to these ears). So maybe there's more to it than the R2R / NOS buzzwords you mention indeed.
 
My thoughts on the topic, posted earlier in the stax thread ( post #3960 )but more relevant here actually :
 

 Why do we limit the frequencies?  Well, human ears can't hear sound over a certain frequency, so we just cut it off.

 
To answer simply: it's not quite that simple I am afraid.
 
I stand to be corrected with what's below as I am writing from memory and memory can be treacherous sometimes but...
- Indeed with a continuous fourier transform (and its inverse), we can go to and from time or frequency domain representations of a signal. This assumes you do not truncate any piece of information in either domains
- Issue is when we go from continuous to discrete (or analog to digital domains if you will),  as you run the risk of truncating some of the frequency content when doing the finite fourier transform
- Without filtering of this high frequency content (basically anything above 1/2 sample rate), it will "fold back" onto the visible portion of the fourier spectrum.
- Without even considering fourier transforms and sticking with the conversion of time domains signals from contiguous to discrete domains, the above explains why "anti-aliasing" filters are necessary prior to an A/D stage.
- Similarly, I believe "reconstruction filters" are necessary after D/A stage to smooth out (filter out) high frequency noise that results from the quantization process. e.g. a stair case has effective very broad frequency content well above the sinusoidal it is trying to replicate.
 
Now where that relates to practice:
- Astrostar is trying to share his bad experiences with oversampling and other reconstruction filters used in D/A stages.
- For instance, no reconstruction filter is gentle if effective, hence the nasty name "brickwall".
- The steeper the slope of the filter (required if you want to maintain bandwidth right up to half the sample rate), the nastier it's transients (e.g. you get lots of pre-ringing or lots of post-ringing or some mix along with phase distortion)
- The solution has long been to either a) get rid of the reconstruction filter (that assumes you trust the high frequency hash you may get as a result isn't going to bite you back), b) oversample the data so that the filter can have gentler slope, c) forget about trying to reproduce 20kHz tone without attenuation and leave with a slight HF roll off from the filter at the benefit of reduced ringing.
- You'd think oversample is the most elegant way to deal with the issue but it turns out an oversampling filter is just another filter with its own issues in terms of transient response and such.
 
 
My personal experience:
- I've heard nasty artifacts with a NOS/filterless DAC I tried a few years back so typically running away from filterless designs.
- I've heard no better than NOS / R2R DAC (with a brickwall filter).
- I've never ever preferred a resampled / up sampled version of a 44.1k, be it software or hardware based upsampling
- My take is that oversampling is thus more detrimental than a well designed NOS reconstruction filter and that is the direction I am headed to.
 
cheers,
arnaud

 

[jacal01]

@ judmarc:
Yes, but what is analog interpolation filtering?  That part confuses me.  The filters you describe are both digital signal filters.

 

[judmarc]

  @ judmarc:
Yes, but what is analog interpolation filtering?  That part confuses me.  The filters you describe are both digital signal filters.

 
We have to distinguish between the way a filter is implemented, and what is being filtered.
 
Here's Wikipedia:
 

Digital filters are not subject to the component non-linearities that greatly complicate the design of analog filters. Analog filters consist of imperfect electronic components, whose values are specified to a limit tolerance (e.g. resistor values often have a tolerance of ±5%) and which may also change with temperature and drift with time. As the order of an analog filter increases, and thus its component count, the effect of variable component errors is greatly magnified. In digital filters, the coefficient values are stored in computer memory, making them far more stable and predictable.

 
So analog filters are implemented with analog electronics; digital filters are implemented with coefficient values stored in computer or chip memory.
 
What is being filtered in a DAC is the digital bitstream; but what is doing that filtering - the filter itself - can be analog or digital.  Usually the interpolation filter and (I believe) sigma-delta modulator are implemented digitally; usually (I believe) the reconstruction filter is a relatively simple analog design.

 

[judmarc]

 
I am confused as well, interpolation means between discrete points, can only be done in the digital domain?!
 
My personal experience:
- I've heard nasty artifacts with a NOS/filterless DAC I tried a few years back so typically running away from filterless designs.
- I've heard no better than NOS / R2R DAC (with a brickwall filter).
- I've never ever preferred a resampled / up sampled version of a 44.1k, be it software or hardware based upsampling
- My take is that oversampling is thus more detrimental than a well designed NOS reconstruction filter and that is the direction I am headed to.
 
cheers,
arnaud

 
Hi arnaud, I agree with most of what you have written, especially your references to the mathematics and why brick-wall filters are necessary for NOS DACs.
 
However, I think in the piece above you have mixed up non-sigma-delta and NOS DACs.
 
R2R DACs are not usually NOS.  You will still have the 8x upsampling (unless, as with the Phasure, it is done in the computer beforehand, or you feed the DAC recordings done in 352.8/384kHz resolution).  What you will not have is the sigma-delta modulation step.
 
What R2R DACs have you heard and liked, and we can look at whether they are NOS?

 

[jacal01]

  What is being filtered in a DAC is the digital bitstream; but what is doing that filtering - the filter itself - can be analog or digital.  Usually the interpolation filter and (I believe) sigma-delta modulator are implemented digitally; usually (I believe) the reconstruction filter is a relatively simple analog design.

 
OK den. But I harken back to the DaVinci DAC quote and your interpretation of it as the source of my confusion.  Surely an analog interpolation filter is implied. if not overtly stated as such. Marketing doublespeak?
 

 
Quote:

The Duet Engine utilizes analog interpolation with parallel output modules utilizing highly precise clocks to double the sample rate for CD 44.1/16 files.

 
In other words, they use an analog rather than a digital oversampling filter. Still an upsampling/oversampling/interpolation filter. (Interpolation is the more technically proper term.) This is, as they mention, to avoid the negative effects of "brick-wall" filtering, i.e., harmonic distortion, that would be required if the DAC truly did no interpolation of 44.1 rates.

 
So an "analog" (digital) filter are those that filter the digital signal via discrete electronic components rather than programmed algorithm software, right?