DCS Vivaldi
- Thread starter MylesBAstor
- Start date
As far as I know yes. The smaller chips I think are latches and I guess they feed the MELF resistors. It could be something like Malcolm Hawksford envisaged in one of his letters to the AES journal - a transversal array of single-bit DACs.
What's fascinating from reading some of the blurb is that on the previous generation they used quad latches and on this one they went over to single bit ones. So they must have discovered ground bounce in the meantime
Opus
What do you think? Always appreciate the benefit of your technical assessment based on what you have read. Thanks!
Lloyd
I'm of the 'old school' of engineering, by which I mean I concur with the dictum 'An engineer is a guy who can make for $1 what any idiot can do for $2'. Here my gut tells me its the other way around - there's an attempt to reach a (stratospheric) price point and unnecessary fluff has been added to get there. As an example I cite a comment made by Thorsten Loesch (designer at AMR) that AMR's dealers were asking the manufacturer to create products at higher price points because there was a market. So I'm skeptical of the whole thing and the marketing materials do nothing to dispel my skepticism. Would you like a deconstruction of their technical highlights document?
I'm of the 'old school' of engineering, by which I mean I concur with the dictum 'An engineer is a guy who can make for $1 what any idiot can do for $2'. Here my gut tells me its the other way around - there's an attempt to reach a (stratospheric) price point and unnecessary fluff has been added to get there. As an example I cite a comment made by Thorsten Loesch (designer at AMR) that AMR's dealers were asking the manufacturer to create products at higher price points because there was a market. So I'm skeptical of the whole thing and the marketing materials do nothing to dispel my skepticism. Would you like a deconstruction of their technical highlights document?
As far as I know yes. The smaller chips I think are latches and I guess they feed the MELF resistors. It could be something like Malcolm Hawksford envisaged in one of his letters to the AES journal - a transversal array of single-bit DACs.
What's fascinating from reading some of the blurb is that on the previous generation they used quad latches and on this one they went over to single bit ones. So they must have discovered ground bounce in the meantime
Thanks, and hmmm... It is very difficult to achieve the performance of an integrated circuit with a discrete approach. For example, in my designs the switches required a lot of care to reduce glitches, and a lot of that was careful sizing and matching, matching well beyond what an array of discrete switches is likely to provide. Besides ground bounce, clocking/time alignment and jitter, IR drop, and a host of other pitfalls are waiting a discrete (or integrated) design. Seems like a lot of work and I wonder if the results will bear it out. I am sure it will for some. A good integrated DAC including error correction and compensation (of the q-levels) with a whole lot of magic in the output buffer woiuld be way cheaper and probably provide most if not all the performance of the DAC box. I do not know, of course (nor am I likely to ever see or hear one, alas).
Wasn't the original Accuphase DAC discrete (at $8k back in the 80's it was quite the beast in its day)?
I need to rejoin the AES but am looking at the latest bills for my kids' college and HS semesters and thinking "maybe next year"....
No surprise to those who know me, but I strongly disagree with that old saw. Most idiots wouldn't know how to start doing what many engineers do daily. Not that some idiots don't make it into engineering, but eventually they end up moving over to marketing...
Thanks Opus. i appreciate your Executive Summary in language [even] i can understand. If you wish to do a deconstruction of their technical highlights for those on the Forum who would understand it...i am sure people would very much enjoy it. I cannot promise i will be able to follow it...and, as you have already said, your executive summary of skepticism is well understood. Thanks as always!
So if I understood you correctly, the Don variant would go like this 'An engineer is one who can do for $1 what any other idiot wouldn't be able to shake a stick at, even for $50' ?
Not that some idiots don't make it into engineering, but eventually they end up moving over to marketing...
Ah - therein lies the rub - in the words 'moving over'. The problem as I see it is that there are almost always two camps - engineers, and marketers. They don't speak each other's language and have a tendency to despise one another's skill sets. I'm an engineer who's taught himself marketing so in my designs I hope to unify the two camps.
Here are a few initial responses to what they've written. In the intro they say its a 'no compromise' design. In fact all real engineering (as opposed to marketing-disguised-as-engineering) is no-compromise, its applied optimisation. Optimising is quite different from compromising - we get compromises when marketing insists on a particular price point or a particular margin on a product. We emgineers can optimize though when marketing realize that price is a constraint, not a hard-and-fast rule. So when marketing give engineering some flexibility over the price and feature set that's when no-compromise engineering can take place. Such flexibility is very unlikely to be given though when there's the normal undercurrent of animosity between the two camps. Speaking of feature sets, it rather looks here there's compromise :
Features standard digital inputs in addition to an asynchronous USB interface. The enhanced
digital volume control allows direct connection to a power amplifier so that in the majority of
systems there is no need for a separate preamplifier. Maximum output can be set at either
two or six volts to suit different amplifier and speaker combinations.
So there are just two output level settings - 2V and 6V. I have to say this beggars belief in
a no-compromise design. Amps and preamps come in more sensitivities than two.
Vivaldi Transport - Extracts revelatory levels of detail from both CD and SACD.
This is total nonsense - there is no 'detail' to extract from digital disks. Either
you've got all the bits right or you haven't. That's the nature of digital - I assume
the marketing guys are ignorant of this. Perhaps engineering have tried to enlighten
them but failed.
Enough or shall I go on?
It's not a hard concept for the marketing guys to get. I suspect it sunk in if the engineering team explained it slowly enough. But it seems to be an impossible concept for many audiophiles to grasp. They want to hear this, very much. And the marketing guys want to tell them what they want to hear. Very much.
-- a marketing guy
Tim
PS: I'm not sure the gap can be breached between Marketing and Engineering. Marketing gets to tell Engineering how to do things they (Marketing) don't even understand, because they know things that Engineering refuses to accept. It's not a gap, really. It's a chasm.
Oh, and please go on!
OK then because Tim's egging me on, here's another installment.
Vivaldi Master Clock
A powerful yet simple to use Grade 1 master clock. Featuring two banks of clock outputs
capable of outputting different frequencies, the new auto clocking mode in Vivaldi improves
ease of use and minimises jitter.
Used as part of the Vivaldi digital audio playback system it improves on an already
spectacular sound and takes it into an entirely new domain. Images snap into sharper focus
and the music displays a substantially greater sense of authority and power as well as, most
importantly, offering noticeably higher resolution of detail.
Practically every digital audio engineer I've read or spoken to would say that the best place
for the clock is right next to the DAC chip itself. Any distance at all increases the chance of picking
up noise and hence increases jitter. I can't fault this logic, so I have to say that for the consumer
having an external clock is very hard indeed to justify. Studios though are different, they do
need synchromisation between various boxes, so they have 'house clock'. Even then though, the
purpose of external synchronisation is to ensure long term (meaning sample-sample) sync -
the short term performance is what matters for the sound and that's dependent on the local oscillator.
Not the external sync signal, which has to be passed through a cable with its inherent band-limiting
effects and added noise.
So given that I can see no engineering justification at all for having an external clock, only downsides,
let''s take a peek at what marketing justifications are on offer.
'Powerful... master clock' - does this mean its Class A? Heavy on the mains supply?
'Ease of use' .. I can only assume they were at a loss for the words to justify the existence of this piece.
'it improves on an already spectacular sound and takes it into an entirely new domain.' - pass me the sick bag please.
I looked for the insipid green face smiley for the last comment but seems this forum isn't so well equipped as others. We need a sick face!
If this last effort at a marketing emetic is to be believed, I can't see how to escape the conclusion that
marketing must have asked engineering to introduce a 'clock knobbling circuit' which applies extra jitter
when the unsuspecting punter hasn't seen the wisdom of purchasing the external clock. But I could be
wrong - what am I missing? Anyone?
Vivaldi Master Clock
A powerful yet simple to use Grade 1 master clock. Featuring two banks of clock outputs
capable of outputting different frequencies, the new auto clocking mode in Vivaldi improves
ease of use and minimises jitter.
Used as part of the Vivaldi digital audio playback system it improves on an already
spectacular sound and takes it into an entirely new domain. Images snap into sharper focus
and the music displays a substantially greater sense of authority and power as well as, most
importantly, offering noticeably higher resolution of detail.
Practically every digital audio engineer I've read or spoken to would say that the best place
for the clock is right next to the DAC chip itself. Any distance at all increases the chance of picking
up noise and hence increases jitter. I can't fault this logic, so I have to say that for the consumer
having an external clock is very hard indeed to justify. Studios though are different, they do
need synchromisation between various boxes, so they have 'house clock'. Even then though, the
purpose of external synchronisation is to ensure long term (meaning sample-sample) sync -
the short term performance is what matters for the sound and that's dependent on the local oscillator.
Not the external sync signal, which has to be passed through a cable with its inherent band-limiting
effects and added noise.
So given that I can see no engineering justification at all for having an external clock, only downsides,
let''s take a peek at what marketing justifications are on offer.
'Powerful... master clock' - does this mean its Class A? Heavy on the mains supply?
'Ease of use' .. I can only assume they were at a loss for the words to justify the existence of this piece.
'it improves on an already spectacular sound and takes it into an entirely new domain.' - pass me the sick bag please.
I looked for the insipid green face smiley for the last comment but seems this forum isn't so well equipped as others. We need a sick face!
If this last effort at a marketing emetic is to be believed, I can't see how to escape the conclusion that
marketing must have asked engineering to introduce a 'clock knobbling circuit' which applies extra jitter
when the unsuspecting punter hasn't seen the wisdom of purchasing the external clock. But I could be
wrong - what am I missing? Anyone?
So if I understood you correctly, the Don variant would go like this 'An engineer is one who can do for $1 what any other idiot wouldn't be able to shake a stick at, even for $50' ?
Ah - therein lies the rub - in the words 'moving over'. The problem as I see it is that there are almost always two camps - engineers, and marketers. They don't speak each other's language and have a tendency to despise one another's skill sets. I'm an engineer who's taught himself marketing so in my designs I hope to unify the two camps.
1) Close enough, but of course I am biased... And, most of my career has been bleeding-edge chips that engineers do for a gob of money and laymen couldn't do at all. However, those "laymen" are experts in other things I know nothing about, so it all evens out in the end.
2) I have known some great engineers who moved into marketing, and some marketing folk who have real understanding of marketing and engineering. Unfortunately, they seem to be in the minority... I have done a ton of marketing myself in support of various proposals and programs through the years, and my bosses thought I was good at it, but my heart was and is in the engineering...
On the transport marketing: The only claims I have read that make any sort of sense in a round-about way is that, if the transport extracts the bits better with lower error rate and less output jitter, it can be reflected in the final product. In this case, when the data is resampled/resynched, and assuming error correction works, there can be no real benefit. At least IMO.
p.s. What exactly is the definition of a "Grade 1 Master Clock"? Is that in the AES standard? I do not know, am just curious as to what the grades of clock circuits are in the audio world.
Google found me this result in Google books (after I'd filtered out all the hits to dCS marketing materials) : http://books.google.com/books?id=y0...onepage&q=aes-11 grade 1 master clock&f=false
The short answer is its about long term stability, which is so not an issue in audio as crystals are all way better than mains frequency stability over the short term. Grade 1 is 1ppm.
@Tim - now you're just getting greedy
Thank you! I thought it was an AES thing but I couldn't remember what. 1 ppm long-term stability seems impressive, but what is "long-term"? Seems pretty tight given things like SAS clocks only require 50 ppm...
Last edited:
Similar threads
Members online
- kmyl
- bonzo75
- Emilzebira
- Juko
- Young Skywalker
- Achilles
- TLi
- Strad
- shawnf
- sapirg
- carolus
- skinnyfla
- Opabin
- jeromelang
- austinpop
- Stef
- Ferris010
- lolo77
- nonesup
- StefanF
- christensenleif@msn.com
- Michael S.
- JayT
- leyenda
- kimurastanley
- DW101
- StereoLogic
- VGAustin
- Azjya
- MarcelNL
- hopkins
- Rotatoubib
- audioblazer
- Azh123
- ICUToo
- naim
- Pichler
- nirodha
- rocco1113
 | Steve Williams Site Founder | Site Owner | Administrator |  | Ron Resnick Site Co-Owner | Administrator |  | Julian (The Fixer) Website Build | Marketing Managersing |