Help me pick a new DAC!

[bdiament]

Hi autoformer,

 

Granted the MH gear requires adapters (I made my own but they can be ordered online too) and it requires climbing a short learning curve. In return, it provides performance and capabilities far, far beyond devices that are seen as its "competition". (Like a Geo is "competition" for a Lambourghini Countach ;-})

 

Personally, I would consider the quality of the drivers the issue and not whether they are 3rd party or not. (Metric Halo's drivers, as everything else from them, have been, in my experience, impeccable.)

 

I would not at all compare any MH product with those from Benchmark. In my experience, they are as different as can be and have nothing in common.

 

If "neutral" is the goal, to my ears, the MH is all alone. It will very quickly show the difference between neutral and what other DACs do. (Which is "better" depends very much on what the listener seeks. I've seen many comments using the word "neutral" to describe devices which, to my ears, impart a quite distinct character on everything that passes through them... quite the opposite of neutral, as I hear it.)

 

Whether MH will fit the bill for "delicious" will depend on how the listener defines that term. I want to hear past the gear to the recording. In that respect, for me, the MH boxes are delicious like no other.

 

As to "accurate and truthful", the MH is the only DAC in my experience I can say this about without reservation. I would not use the expression to describe DACs that provide "enhanced detail", i.e. spurious harmonics not contained in the source. (This includes just about every DAC I've heard that does on-the-fly sample rate conversion.) Neither would I use it to describe those that make everything sound "silky smooth" -even when the source isn't.

 

Best regards,

Barry

www.soundkeeperrecordings.com

www.barrydiamentaudio.com

 

 

[Paul R]

That's excellent news Barrows, thanks for sharing.

 

For whatever reason, almost every component in our systems was designed in the U.S., the U.K., or Canada. About 3/4 of them were manufactured in China though, including the Macs that are at their heart.

 

-Paul

 

 

[autoformer]

At this point, with all your input, suggestions and discussion, I am leaning toward the Ayre QB-9. It meets all the criteria, priced so that I can afford some nice cables to go with it, and has the singularity of purpose that I always aspire to but can rarely achieve in my software development work. It does one thing and one thing only.

 

But before I place the order, I decided to spend some time today doing a little shootout between my W4S DAC2, Benchmark DAC1 USB and Wavelength Brick.

 

As I've said before, I love the Brick on my solid state + electrostat system as it smooths out that system's inherent sharpness in a way that I enjoy. But for the tube system in question here, it's a little too much of a good thing.

 

That left the Benchmark and the W4S DAC2. Through their USB ports, they are vey different. Vocals seem thicker and more chesty sounding with the Benchmark, much less so with the DAC2. Switching A/B between the two was like swapping a set of resonant-cabinet speakers (Benchmark) for a set of very well-damped speakers (DAC2). The Benchmark also pushed the vocalist closer to me. I felt as if the soundstage was deeper with the Benchmark. Maybe this is the "forward" presentation that people talk about with the DAC1.

 

I have a V-Link floating around the house, so I decided to try it with the Benchmark and am glad I did. The deep soundstage and forward presentation that I enjoy was still there, but there was now more air and detail. And as a further bonus, vocals had lost that thick, chesty character. The Benchmark had retained all its dynamism and forward presentation style but it was now somehow easier to listen to. I've never heard the Benchmark sound this good!

 

The V-Link/V-DAC has seen duty only in my more casual systems, like the hobby room, for example. But it seems that the V-Link has something to higher end systems as well. A very useful device that I believe is selling at a steep discount these days with the advent of the V-DAC mkII. I hope MF doesn't phase out the V-Link because I can see how a lot of people will find it quite useful.

 

Now I'm starting to wonder if a Benchmark DAC1 HDR with a Wavelength Link would fit the bill. Total cost would be about the same as a QB-9. The HDR's output stage is supposed to be improved from my DAC1 USB. Its S/PDIF coax inputs will allow 24/192 and the Wavelength Link will also do 24/192.

 

[Mark Powell]

Barrows,

 

I totally agree with what you and 'Autoformer' say, and thought he made his point very well, including his comments about the UK having given up. My 'its all Chinese' was somewhat exaggerated just to make a point. But nevertheless it is largely true. I am in the market for a 'near' high end amp. More than half the ones I have seen are made in China, though not all of them were notionally American. Arcam here just bought its production back from China. They of course are a mid-market producer.

 

I am not against the Chinese. But you went to the moon. We engineers and 'scientists' respected that enormously and still do. I don't think there were many Chinese components on board.

 

My amp choice has narrowed to Naim or American. But it has got to be 'real' American. Pass or the lower-cost section of Audio Research appeal. McIntosh are a bit expensive for me. In the UK they have a reputation of being for wealthy and cautious American businessmen looking for a safe choice. Probably unfair.

 

[autoformer]

Our global economy means that no one country can do everything needed to produce every product at every price point. Just because a product has only 80 or 90% local content doesn't mean you abandon the idea of buying locally altogether. If we all took that point of view in the US, we'd be done. You do the best you can.

 

McIntosh is still building products with very high US content. They are vertically integrated to the point that they even wind their own transformers, mill and cut their own steel and aluminum parts, make their own circuit boards and even etch the glass for the faceplates - all in house in New York. And though they are owned by a Japanese holding company, they are still providing jobs for Americans.

 

As to the computer, well...at least the Mac is designed in Cupertino. And the most important component in that computer - the software is a true US product.

 

And yeah, the Harbeths are made in the UK, but again, that doesn't mean I should throw up my hands and not bother trying to keep my system as locally-made as possible.

 

Whenever I go the UK for business or to visit friends, I am always surprised at the cynicism and hopelessness that I see there. It is as if people have just given up on being a world power and have just decided to accept their fate. We in the US are in the same position as a global power in decline. But my hope is that the US will not take it lying down. I still have faith that the American people will realize what's happening, get off their fat lazy rear ends and start building things again. But for that to happen requires that consumers are willing to buy those products. My focus on buying my hi fi gear from US sources is just my small way of pulling for the home team.

 

[autoformer]

At RMAF last weekend, I visited the Peachtree (based here in Bellevue, Washington) room and learned that production of the new high end "Grand" line of products (currently a preamp and an integrated amp) is being done on the USA. Why? Quality control and less frustration and headaches in general.

 

Many fine American hi fi products are made in the USA:

Martin Logan speakers are made in Lawrence, Kansas

Bel Canto is made in Minneapolis, Minnesota

VPI turntables are made in New Jersey

Zu Audio speakers are made in Ogden, Utah

PS Audio has moved production of the Power Plant to the USA

Magico, Studio Electric, Decware, Benchmark, Wyred4Sound and the list goes on...

 

In my last job at a major mobile phone manufacturer, we attempted to do development of software for mobile devices and the back end services in a global matrixed organization. Our attempts to work with developers in India and China were an absolute disaster. We ran into two main problems:

 

1) Consumer software development is complex. Agility and speed often mean that you start with a spec that evolves in response to unanticipated problems and market conditions. Doing this with a team that is physically co-located works very well. But in an environment where interlocking parts of a project are being developed by teams in different countries, you can't operate that way. For that model to work, you need very firm specifications and that slows you down, makes you less agile and adversely affects team morale.

 

2) We found that the teams in China and India really needed those firm specs. Individual developers in the US team would run into problems and improvise on their own initiative whereas the Chinese and Indian teams would generally stick to the spec and whack their heads into a wall, even though clearly it was the spec that needed to change. So, in meetings, you would look at a piece of work by one of the overseas teams that was an attempt to solve a really hard problem presented by the spec and see that they went to a lot of trouble and ended up with a crazy, inelegant solution. On the other hand, the US engineers would simply come back to you and tell you that we shouldn't do this or that, the reasons why and suggest an alternative. The US engineers challenge management's plans and assumptions continually, which always results in better products. It is the difference between doing the "right thing" and doing what you are told. As a technology manager, I much prefer the former in an engineer.

 

The best way I ever found to work with overseas development teams was to give them discreet units of very well-defined work to do, driven by a stable very well-defined specification.

 

I can only assume that it is much the same way with hardware production, only more so. I talked to several US manufacturers at RMAF who were either moving production to the US or who had considered production in Asia but had decided against it for similar reasons to those I describe above.

 

 

[jamesg11]

Barry,

 

re MH - you might not have convinced autoformer, but you've won me over!

 

Jim.

 

[Audio_ELF]

Mark, other (high end) UK HiFi manufacturers would include Chord, Leema, Sugdem and AMR. All produce kit at the high end.

 

Also Rega, Roksan, Creek and Cyrus all sneak into the "High End" category IMO.

 

I believe all these companies manufacture primarily in the UK. And that's not mentioning Linn, Naim, Meridian (former category) and Arcam (latter).

 

If you add purely speaker manufacturers this list about doubles.

 

[PeterSt]

 

 

[bdiament]

Hi Jim,

 

As always, I suggest a personal audition.

I report what I hear and it is probably clear I've become a huge fan of MH gear.

I've owned or used everything they make and have compared it (over and over again) to all the "contenders" I can find. It has been an education, every single time (such as the one unit widely reviewed to have the superb soundstage - a soundstage that sounded collapsed (!) when compared to the ULN-8).

 

I've been using an '8 now for over 5 years - since it was first released to beta testers. This is a long time in the world of digital electronics, however, I still haven't heard another device that is as good as just getting out of the way.

 

For me, it is in a class by itself but that is because it fits what I seek; it gets out of the way and gives me access to the recording itself.

 

But not everyone wants this. Some want "enhanced detail", some want a converter that makes everything - even rough recordings - sound "silky smooth", some want "delicious". Nothing wrong with any of those but if that is the goal, for all recordings, the ULN/LIO-8 is not the answer. The MH boxes don't "enhance" anything and don't make things "smooth". They simply provide the recording, for better or worse. For my work, this is a necessity; for my listening pleasure, it is what I seek.

 

Be aware of the learning curve and non-"standard" (in hifi gear) connectors as well. Above all, a personal audition will tell you everything you need to know.

 

Have fun!

 

Best regards,

Barry

www.soundkeeperrecordings.com

www.barrydiamentaudio.com

 

 

 

 

 

[autoformer]

Well, this will be no surprise to any of you, but I am down to two DACs: The Ayre QB-9 and Wavelength Cosecant. Barry makes a good case and I know more than one person who is switching from a hi-fi DAC (an ARC DAC8, in one case) to a Metric Halo.

 

I spent some time looking at John Atkinson's measurements for these two DACs as well as the Benchmark DAC1 USB, which I own and for which JA seems to have a real soft spot. By the numbers, the Benchmark is at least the equal of either the Ayre or Wavelength, with the exception of USB jitter. I'm not an expert at interpreting the measurements and would love to hear from someone with the training to understand them more fully.

 

http://www.stereophile.com/content/benchmark-dac1-usb-da-processor-headphone-amplifier-measurements

 

http://www.stereophile.com/content/ayre-acoustics-qb-9-usb-dac-measurements

 

http://www.stereophile.com/content/wavelength-cosecant-v3-usb-digitalanalog-converter-measurements

 

So why is it, then, that so many people find the QB-9 and Cosecant so far superior to the humble DAC1? I keep reading comments from people who hear a certain "magic" in the Cosecant and I wonder where that comes from? Look at the measurements for the Cosecant in the link above. The frequency response curve is alarming, with bass response starting to roll off below 100Hz, and down by over 1db at 10Hz. High frequency response starts to roll off around 10kHz and is down almost 1db at 20Hz. Compare this to the Benchmark's virtually flat line and you start to wonder. Where the Cosecant really shines, though, is in its low jitter. So if jitter reduction is really the Wavelength's strong suit, I still wonder if it doesn't make sense to go with the Benchmark, a DAC with "superb" measurements, as JA puts it, and a Wavelength WaveLink in front of it handling the USB interface.

 

Boy would I love to hear a Cosecant in action. Unfortunately, my local dealer for them just closed its doors.

 

 

 

[bdiament]

Hi autoformer,

 

I think this is a good example of specs not telling anything at all about how a component sounds. I find the numbers can be useful in identifying certain flaws and ultimately, completely useless in identifying assets.

 

By reclocking the signal, it isn't hard to show what may appear to be impressive jitter numbers. (I remain completely convinced, as I have been for years, there is more to the digital picture than has as yet been identified, much less quantified.)

 

Great jitter numbers and a flat-line frequency response won't alter the fact that some converters add enough spurious harmonics to every signal that passes through them to give the impression of "enhanced detail". A lot of folks like this sound. Personally, it makes me run for the medicine chest in search of aspirin. To my ears, Music doesn't sound like that. Decent recordings don't sound like that. Yet that is what happens to them when played through such a device.

 

But that's just me. Apparently a lot of folks like this sort of sound. Some reviewers pronounce it the cat's meow. (Makes me wonder if they've ever heard a cat either. ;-}) I've read rave reviews about more than one that I would classify as an "ear ripper".

 

Whether its the ear rip---- er, I mean "enhanced detail" converters or their counterparts, the "silky smooth" converters, I'm always wary when any audio device applies a common character to different recordings. Every master I've ever heard sounds very different from every other master. When gear applies a character, the differences are diminished. With those differences, goes (in my experience) much of the magic contained in the better recordings.

 

In the end, it is in the ear of the behearer.

Only your ears will tell you what the spec sheets (and reviewers) can't.

 

Have fun!

 

Best regards,

Barry

www.soundkeeperrecordings.com

www.barrydiamentaudio.com

 

 

 

[Priaptor]

http://www.stereomojo.com/Lampizator%20Level%204%20DAC%20Review.htm/LampizatorLevel4DAC.htm

 

This is the second time I have heard about this. They also have a 2 box unit, their Level 5

 

[Mark Powell]

Going right back to the original question.

 

As I see it, from what I have gained mainly from this site, plus one or two others I have figured this out:

 

1) Almost all DACs consist of the manufacturer buying a whole bunch of off-the-shelf chips from a semiconductor manufacturer, shaking them around as the manufacturers literature suggests and putting them in a box. This includes the USB input and the digital volume control where fitted. The DAC manufacturer does have some freedom when power supplies are considered and power supplies are (fairly) important. However, even a good one need not be expensive as the power requirements are low.

 

2) There are two sorts of USB chip. A cheaper variety, such as the C-Media type and its equivalents and a more expensive and higher performing variety.

 

3) Digital volume controls work partially by dropping off the least significant bits. No doubt this is explained in the chip manufacturers literature so no great expertise is needed to incorporate one in your box. The more bits you start with the better.

 

3) DAC chips are all 'much of a muchness'. It is doubtful that one chip manufacturers engineers are top experts and another chip manufacturers are incompetent. The Sabre chips tend to be more expensive than most (tens of dollars rather than a dollar or two) and are thought to be better. Not in widespread use as most DAC manufacturers ('assemblers' really) would rather pay two dollars than ten.

 

4) Because of the above the main difference between a 500 dollar DAC and a 5000 dollar DAC is the price.

 

5) There are two DAC manufacturers who do it differently. One of them uses no DAC chips whatsoever. Neither meet the 'Made in USA' requirement.

 

It is different with, for example, amplifiers. While it is possible to 'assemble' an amplifier from bought in modules, by 'Audiophile' standards the result is no more than adequate. By everyone else's standards they are fine. Most electronic engineers and electronic manufacturers consider us to be a misguided fringe and a very small fringe at that. So they don't bother with us or our opinions. This leaves a space for relatively small suppliers to operate. Good. It allows them to make money and it pleases us. Thomas J Watson, the founder of IBM, once said "Good business is a business that satisfies both parties". How right he was. And the best of the Audiophile amplifiers are very very good indeed. Stunningly so.

 

But paying thousands of dollars for a bunch of mass market chips in a DAC? No thanks. Not even if they stick a couple of mainly decorative cheap Chinese tubes on the end to try and convince me it's 'Audiophile'.

 

 

[bdiament]

Hi Mark,

 

"...Because of the above the main difference between a 500 dollar DAC and a 5000 dollar DAC is the price..."

 

In my experience, "the above" misses enough to make the statements quite a bit more than a little off base. If the designer seeks Quality, building a DAC (or any other component) is a whole lot more than putting third-party parts inside a box. This is what separates the finest DACs from the rest (whether the rest cost much less or much more).

 

To say the main difference is the price suggests sonic differences are minimal or insignificant. From such, I would conclude you've heard many DACs at many price points and found them all to sound substantially the same to your ears. If that is not the case, I would conclude that you have not heard many DACs.

 

I've never heard two that sound the same. This, even though I do find there are general "schools" of DAC sound. As I've written elsewhere, to my ears, there are the "detail enhancers" (including many that use the current chip-of-the-month) and there are the "silky smooth" sort. Within each group there is great variation but members of both generally apply the same sound to whatever passes through them. Some make everything sound sound really, really good -- and that is precisely what I find to be wrong with them; in doing so, they obscure the true sound of the source.

 

And there are those very rare DACs that don't fall into either of the two primary schools and tend to pass on the sound of the input signal, without editorializing.

 

Which a given listener prefers will depend on what that listener seeks.

But they all differ in many, many ways beyond the price tag.

Just my perspective of course.

 

Best regards,

Barry

www.soundkeeperrecordings.com

www.barrydiamentaudio.com

 

 

[autoformer]

My sincere thanks to everyone who contributed to this thread. In the end, I decided to go with the QB-9, since I think for this particular system I need a more linear and truthful source. Everyone who owns a Cosecant just loves them, but what they seem to like most about it is its sonic signature. If you look at John Atkinson's test results for the Cosecant, it doesn't measure very well, except for its remarkably low jitter. The low end starts to roll off around 100Hz and the high end starts to roll off around 10kHz. In an all solid state system, this probably sounds really good, for the same reason that my Brick sounds brilliant when feeding my large theater system (McIntosh C50 preamp, MC452 power amp and Martin Logan Summit X speakers.) The Summits are very revealing speakers, and they show the Brick's nuanced approach to good effect. But with the laid-back Harbeths, all that nuance and delicacy is just lost. Actually, the best sounding thing I could come up with on my Harbeth system was a Benchmark DAC1 with a V-Link handling the async USB interface. To me, the Harbeth/McIntosh tube setup just seems to sound better with a more hard-edged source. As you can guess, the MC275 and C22 already have a very warm, rounded sound, probably more rolled off on the frquency extremes than modern electronics, and I was concerned that the Cosecant would be too much of a good thing.

 

If I had been buying this DAC for my big system with the Logan ESLs, I would have certainly gone for the Cosecant. And in fact, I may do just that in the next few months, once I hear the C50s DAC with the new firmware updates in place. But for now, in my small McIntosh/Harbeth system, I believe that the more linear QB-9 will be the better choice.

 

And Barry, once my plans for doing some field recording get better defined, I'm certainly going to revisit the ULN-8.

 

It should arrive in a couple weeks and I'll report back here.

 

Thanks again!

 

 

[barrows]

It appears you are showing a level of ignorance in your responses, or perhaps you are ignoring some of the available converters out there... Autoformer, you will be interested in this as well, considering you have decided to go with the QB-9

 

Here are some facts about the QB-9, which show that it is not engineered and built the way Mark claims for US made DACs:

 

1. The Ayre QB-9 uses a proprietary oversampler and digital filter, running two filter choices which Ayre engineers spent an entire year developing-this is not an off the shelf chip based soultion which anyone can purchase.

2. The Ayre QB-9 uses an off the shelf DAC chip from TI, but it bypasses the onboard oversampling and filters, in favor of the proprietary Ayre filters described above.

3. The Ayre QB-9 has no off the shelf chips (IC opamps) in its output stage, instead, it has a proprietary Ayre designed output stage, implemented with discrete transistors and resistors.

4. The Ayre QB-9 power supply does not use off the shelf chips for voltage regulation (except the regulator used only for the control circuitry), instead, it uses proprietary Ayre designed voltage regulators, built with discrete parts.

 

Autoformer: To your previous question, as to what would be responsible for the sound of the Cosecant/Ayre vs. something like the Benchmark, my response is all of the above features-the Benchmark uses off the shelf, IC solutions for all of the above functions: Ayre and Wavelength prefer to engineer their own, proprietary solutions rather thna relying on chip based soultions sourced from TI, AD, National Semi, etc...

 

[autoformer]

@Barrows: Thanks for the info on the QB-9!

 

Perhaps Benchmark's off-the-shelf component approach isn't such a bad thing. If you were to make your decision based solely on JA's measurements, you would go for the DAC1, which seems to best the QB-9 in every category, yet most people seem to prefer the sound of the QB-9. I wonder what factors that are not being measured could account for that. Perhaps, as the Benchmark fans say, we just "can't handle the truth".

 

Ayre QB-9 measurements:

http://www.stereophile.com/content/ayre-acoustics-qb-9-usb-dac-measurements

 

Benchmark DAC1 measurements:

http://www.stereophile.com/content/benchmark-dac1-usb-da-processor-headphone-amplifier-measurements

-- and --

http://www.stereophile.com/content/benchmark-dac1-usb-da-processor-headphone-amplifier-follow-july-2008

 

 

[barrows]

Well... personally I cannot stand the benchmark, and it sure does not sound like any music I have ever heard, except perhaps the absolute the worst sounding rock concert in a basketball arena!

Note, that any measurement results at/below -120 dB will be inaudible, so differences there are moot. Also note that measurements never tell the whole story, as steady state single frequency test tones do not tell much about what will happen in circuit with actual music signals. My suspicion is that time domain effects (not really tested in the traditional measurement set) are responsible for much of the difference in sound.

Note that IC opamps, as an example, are designed specifically to measure well, and are generally not even listened to during development. On the other hand, a company like Ayre, both measures, and listens to, its circuit designs during development, and designs its circuits to both have good measurements, and to sound accurate with music.

 

[Paul R]

To be fair, I think you will find a company like Benchmark also measures and listens to their products. I find the Benchmark DACs to be, if anything, brutally honest to the recording. And yes, that means a lot of music sounds artificial and in some cases, downright awful.

 

But it sounds pretty much exactly like the recording, as it was meant to.

 

-Paul

 

 

[bdiament]

Hi Paul,

 

I hear it the way barrows does.

To my ears, almost no DACs are "brutally honest to the recording".

Certainly those that "enhance detail" -for me at least- don't get remotely close to even telling the truth, much less brutal honesty.

 

As I hear it, nothing that applies the same character to every sound passing through it has a shot at being honest.

 

Best regards,

Barry

www.soundkeeperrecordings.com

www.barrydiamentaudio.com

 

 

[Paul R]

Well, if both you guys agree, then it pretty much has to be so. (grin)

 

I find the Benchmark is at times, unpleasant, but I assumed it was because of a poor recording.

 

-Paul

 

 

[AudioDoctor]

that thinks the Benchmark is truly atrocious. I listened and listened to that thing, wondering what all the hype was about, waiting for it to "break in" before I finally gave up. Been happy ever since now that the damned thing is gone...

 

[autoformer]

Surprising comments about the DAC1 here. You guys really hate that thing!

 

I must say that it doesn't mirror my experience. Then again, the Harbeths seem to make everything sound great.

 

 

[4est]

The Benchmark DAC has always had polarized opinions.