breakthrough? (correcting phase error in DAC)

[esldude]

a filter with a notch at 1/2 the sample rate..............now where have i heard that before?
this sounds like another implementation of the miraculous apodizing filter, if i'm not mistaken.

Yes not so different. :b

 

[cat6man]

it all seems to come down to the impulse response and the ringing
lots of ringing is bad.
brick wall filters have lots of ringing.
higher bit rate sampling permits more gradual filters which have less ringing.
lots of ringing and jitter together are very bad since the time offset due to jitter extends far out in time if there is lots of ringing in the impulse response.

imho, jitter and filtering are so intertwined that even pico-second clocks by themselves cannot eliminate the problem. wadia was on to something important about the impulse response long before others (as far as i can tell).

 

[MRJAZZ]

Yes, I have had an eye-opening experience in this regard when, courtesy of Goodwin's High End, I was able to compare at home the highly acclaimed NAD M51 DAC to my Wadia 12 DAC. I sonically preferred the 20 year-old (!) Wadia 12 to the NAD DAC, except in the bass performance (well, the Wadia 12 had been upgraded in 1997 with the Wadia 860 opamp, but that was also 17 years ago). So much for the hype of ever higher performance of digital over the years. Don't get me wrong, the NAD is a great DAC when judged on its own, but folks, let's be real. Paul at Goodwin's High End wasn't surprised about the findings.

I ended up buying the Berkeley Alpha DAC 2. Now that one did live up to the hype of ever higher performance of digital over the years ... It was simply in a different league than both the NAD M51 and the Wadia 12, with a stunning price/performance ratio. Paul was like, I told you so... (yes, it is 2.5 x more expensive than the NAD M51, but in my view ridiculously cheap relative to its performance).

But the NAD M51 is still a reference of comparison for John Atkinson from Stereophile when it comes to the best of today's DACs. This DAC comes up again, for example, in his recent review of the Auralic Vega DAC. To be honest, his comparisons left me completely in the dark about the Auralic Vega's performance, which may be great, but I really don't know. I like to read JA, but that was one hell of a worthless review, in my opinion.

Well, in general most reviews of audio gear are worthless, again just in my opinion (there are a few writers that do make me pay attention, such as Roy Gregory, now at Audio Beat).

Atkinson actually closes his review of the Vega saying it is a bargain at 3499 compared to the 3 piece Vivaldi cd stack (not the full stack). He does admit that he is going off memory here as the Vivaldi stack was not there during the Vega review..

Cheers......T

 

[Orb]

it all seems to come down to the impulse response and the ringing
lots of ringing is bad.
brick wall filters have lots of ringing.
higher bit rate sampling permits more gradual filters which have less ringing.
lots of ringing and jitter together are very bad since the time offset due to jitter extends far out in time if there is lots of ringing in the impulse response.

imho, jitter and filtering are so intertwined that even pico-second clocks by themselves cannot eliminate the problem. wadia was on to something important about the impulse response long before others (as far as i can tell).

Yeah,
definitely interesting approach with looking at combining both minimum and linear phase.

Would be interesting in knowing how Chord Electronics do theirs as well, but they are very sneaky in that any impulse response analysis is picked up by the internal software and so does not use their filter/co-efficients.
Benefit being solution remains only with them.

Cheers
Orb

 

[Ken Newton]

I don't think Tony Taddeo actually said there was a phase lead...

I took the following paragraph from Taddeo's US patent #5436882, titled 'Method and device for improving digital audio sound':

"It should be noted also that the sampling frequency utilized in the production of digital recordings also has the effect of introducing some slight, undesirable phase shift in the reconstructed analog output signal which is produced by a D/A converter from a digital recording. This phase lead can result in some slight distortion of the reconstructed audio signal."

Taddeo's patent is based on what's known as a 'comb' filter. Comb filters not only feature a response notch, they feature a repeating pattern of response notches versus frequency. A broadband bode plot of such filters resembles a hair comb with it's teeth, hence, the name. Although, it should be pointed out that Taddeo's comb filter does not produce a notch anywhere near the Nyquist frequency, as do so-called 'Apodizing' digital filters.

Comb filters are old hat, and are easily produced by splitting a signal in two, delaying one of those split signals, then recombining the split signals back in to one. The key effect this has on an SINC function interpolation oversampling filter based DAC output is to partially suppress the digital filter's ringing, which takes place at the Nyquist frequency. The performance trade-off is an in-band frequency response roll-off of the treble. Taddeo's solution produces much the same in-band results as slow slope digital filters, but did so before slow slope filters were a common feature option.

 

[Don Hills]

... Would be interesting in knowing how Chord Electronics do theirs as well, but they are very sneaky in that any impulse response analysis is picked up by the internal software and so does not use their filter/co-efficients. ...

... so we only have their word that their special filters exist.

 

[Orb]

... so we only have their word that their special filters exist.

hehe that is one way to look at it
Cheers
Orb

 

[mep]

We should start a new thread called "Digital Distortion of the Day" because every day seems to bring a new digital distortion to light with a 'must-have' solution right around the corner.

 

[esldude]

And if those Chord filters exist, how would they differentiate between transient/impulse testing and transients/impulses in music?

 

[esldude]

I took the following paragraph from Taddeo's US patent #5436882, titled 'Method and device for improving digital audio sound':

"It should be noted also that the sampling frequency utilized in the production of digital recordings also has the effect of introducing some slight, undesirable phase shift in the reconstructed analog output signal which is produced by a D/A converter from a digital recording. This phase lead can result in some slight distortion of the reconstructed audio signal."

Taddeo's patent is based on what's known as a 'comb' filter. Comb filters not only feature a response notch, they feature a repeating pattern of response notches versus frequency. A broadband bode plot of such filters resembles a hair comb with it's teeth, hence, the name. Although, it should be pointed out that Taddeo's comb filter does not produce a notch anywhere near the Nyquist frequency, as do so-called 'Apodizing' digital filters.

Comb filters are old hat, and are easily produced by splitting a signal in two, delaying one of those split signals, then recombining the split signals back in to one. The key effect this has on an SINC function interpolation oversampling filter based DAC output is to partially suppress the digital filter's ringing, which takes place at the Nyquist frequency. The performance trade-off is an in-band frequency response roll-off of the treble. Taddeo's solution produces much the same in-band results as slow slope digital filters, but did so before slow slope filters were a common feature option.

Well I had read a condensed version of that patent which didn't include that statement though it described what the filtering did. And yes it was a comb filter. Of course the procedure I described was also a comb filter, but when everything above the first comb has been removed (as happens with digital sampling) I simply called it a notch. The other difference is I delayed the second signal by one bit while he delayed it equivalent to one half a bit. Using one bit delay actually does a better job to suppress ringing than half that much delay. It does alter frequency response more as a result. His comb filter produced a notch at the sampling frequency. The procedure I described produced one at the nyquist frequency (which is the frequency of the ringing). So my described procedure is closer to apodizing filters though still not same as those. One could produce closer to the Taddeo result by upsampling by a factor of two and lopping off one bit before combining the signals. Then down sample to the original rate.

But unless Soulution describes what they are doing we don't know if it resembles any of these.

 

[Don Hills]

And if those Chord filters exist, how would they differentiate between transient/impulse testing and transients/impulses in music?

Indeed. You wouldn't want it switching in the middle of a music passage. And you could embed impulse and IM test pulses in the middle of a music track...

 

[wisnon]

Excerpt from Hifi News review:

"Robert Watts’ WTA interpolating filter and Pulse Array DAC
technology has evolved over the years and is largely responsible
for the very fine performance realised here. Via S/PDIF, it enjoys
a high 3V maximum output, a low 65ohm source impedance,
wide 111dB A-wtd S/N ratio and impressively low 0.0007%
distortion through its midrange [red/black traces, Graph 1].

At high frequencies the analogue stage is slightly ‘stressed’
[blue trace, Graph 1] and so distortion is higher at peak output
(0.017%, 20kHz) than at lower levels (0.002%, 20kHz/–20dBFs).
Jitter is exquisitely low at <10psec for all sample rates from
44.1kHz-192kHz via S/PDIF and only slightly higher at 135psec
via USB thanks to a series of ±750Hz sidebands.

In practice its USB performance is identical to that via S/PDIF except for its S/N
which falls back to a ‘16-bit’ 96dB. We’ve reported this before,
as the QuteHD allegedly uses the same Italian-sourced USB PC
drivers implemented by the North Star Essensio and M2Tech
DACs [see HFN July ’11, May and June ’12].

The responses are exceptionally flat (within ±0.01dB)
with 44.1/48kHz media, dropping by just –0.3dB/45kHz and
–4.5dB/90kHz with 96kHz and 192kHz files, respectively.
Using an impulse to measure response [black trace, Graph 2]
reveals Watts’ ‘protection’ of his IP: impulse data is detected
and passed directly out, preventing engineers from extracting
his proprietary filter coefficients from the recovered time
domain data! Readers can download full QC Suite test reports
detailing the Chord QuteHD DAC’s S/PDIF and USB performance
by navigating to www.hifinews.co.uk and clicking on the red
‘download’ button.
PM
ABOVE: Frequency response, 20Hz-20kHz, with
steady-state data (red trace). The filter allows
impulses (black trace) to pass through unprocessed
ABOVE: Distortion vs. 24-bit/48kHz digital signal
level over a 120dB dynamic range. S/PDIF input
(1kHz, red) and USB input (1kHz, black; 20kHz, blue)
C"

 

[Phelonious Ponk]

...I compared the sound both ways with a variety of CDs and SACDs, and the results were eye opening. Engaging the phase cancellation function consistently led to greater spaciousness and the elimination of note “smearing” that I had assumed was endemic to the recordings. Based on what I heard, the Soulution phase shift cancellation circuit could be a watershed development in digital sound evolution...

Those of you who have not acquired this new "breakthrough...." Are your notes smearing? Really?

Tim

 

[wisnon]

Is this just phase flipping (inversion)? If so, a much cheaper way to do the same thing would be this: The Silk power conditioner with Phase Flipper function.

Phase aligns up to 4 components in your audio chain and does gentle Power conditioning too.

http://www.lampizator.eu/NEWDAC/Lampizator/SILK_AC_filter.html
€700-

 

[Ken Newton]

Is this just phase flipping (inversion)?...

We don't know for certain, but don't believe so.

 

[cat6man]

Excerpt from Hifi News review:

"Robert Watts’ WTA interpolating filter and Pulse Array DAC
technology has evolved over the years and is largely responsible
for the very fine performance realised here. Via S/PDIF, it enjoys
a high 3V maximum output, a low 65ohm source impedance,
wide 111dB A-wtd S/N ratio and impressively low 0.0007%
distortion through its midrange [red/black traces, Graph 1].

At high frequencies the analogue stage is slightly ‘stressed’
[blue trace, Graph 1] and so distortion is higher at peak output
(0.017%, 20kHz) than at lower levels (0.002%, 20kHz/–20dBFs).
Jitter is exquisitely low at <10psec for all sample rates from
44.1kHz-192kHz via S/PDIF and only slightly higher at 135psec
via USB thanks to a series of ±750Hz sidebands.

In practice its USB performance is identical to that via S/PDIF except for its S/N
which falls back to a ‘16-bit’ 96dB. We’ve reported this before,
as the QuteHD allegedly uses the same Italian-sourced USB PC
drivers implemented by the North Star Essensio and M2Tech
DACs [see HFN July ’11, May and June ’12].

The responses are exceptionally flat (within ±0.01dB)
with 44.1/48kHz media, dropping by just –0.3dB/45kHz and
–4.5dB/90kHz with 96kHz and 192kHz files, respectively.
Using an impulse to measure response [black trace, Graph 2]
reveals Watts’ ‘protection’ of his IP: impulse data is detected
and passed directly out, preventing engineers from extracting
his proprietary filter coefficients from the recovered time
domain data! Readers can download full QC Suite test reports
detailing the Chord QuteHD DAC’s S/PDIF and USB performance
by navigating to www.hifinews.co.uk and clicking on the red
‘download’ button.
PM
ABOVE: Frequency response, 20Hz-20kHz, with
steady-state data (red trace). The filter allows
impulses (black trace) to pass through unprocessed
ABOVE: Distortion vs. 24-bit/48kHz digital signal
level over a 120dB dynamic range. S/PDIF input
(1kHz, red) and USB input (1kHz, black; 20kHz, blue)
C"

I love it!

Let's hide the impulse response!

If I had one of these, I would take that as a challenge...........do you think they also hide the step and square wave responses?
Come on, someone out there should be able to do this, right?

 

[Ken Newton]

Here's a link to a Chord slide presentation on DAC design issues and technology, including their so-called Watts-Transient-Aligned (WTA) digital filter. However, it doesn't provide enough technical details to determine exactly how the WTA filter works.

http://www.chordelectronics.co.uk/files/Hugo DAC Technical Master Class (1).pptx

 

[wisnon]

We don't know for certain, but don't believe so.

So, is it like the BSG Qol signal completion stage device?

 

[Ken Newton]

So, is it like the BSG Qol signal completion stage device?

As I said, we don't know for certain. We don't have the key technical details.

 

[wisnon]

I was asking for a best guess.