FORGETTING the Digital to Analog conversion part, what is BEST Digital source?

[Ryan Berry]

4 hours ago, barrows said:

Perhaps you could share an example of this poor engineering?

Take for example Ayre products.  These are exceptional audio performers by anyone's standards, but, Ayre chooses not to use global negative feedback in their circuits.  Because they make the subjective choice to not use GNF, their products exhibit slightly higher distortion and lower damping factor (in amplifiers), the distortion levels are still low enough to be below what is considered to be audible.  They have these somewhat higher distortion measurements ON PURPOSE because it is their feeling that GNF sounds worse.  This is not poor engineering, in fact it is superb engineering as they have been able to produce innovative solid state circuits without resorting to GNF, which still have low distortion.  The engineers at Ayre are highly capable, and could produce much lower distortion components by adding some GNF, but they choose not to specifically to produce a sound they feel is more accurate.

 

I can do you one better and point out that Ayre actually introduced some feedback in our designs in 2014.  After designing another product that had negative feedback out of necessity and knowing that the product sounded good, Charley had thought that it would certainly sound better than our existing design.  So they put it in, but didn't really listen to it as much as we normally would as they were under a time crunch and I was swamped with other work at the time and could put zero time into it myself.  The first models went out and were almost universally complained about.  They sounded better in some ways and certainly improved a few other factors in the units (not to mention they were way easier to set up in the shop), but the overall sound quality was definitively worse in other ways and uncharacteristic for Ayre, preventing them from standing out in the market.  So we recalled every unit in the field and re-released the redesigned units. 

 

We, like many other companies, make sure to subject our products to a large number of tests when designing and building our products.  In production, this can quickly tell us if something is wrong with a unit, while in development it can spot some pretty obvious faults early on that helps us wasting time later.  For example, if an amplifier that's designed to put out 300 watts starts showing 5% distortion at 20 watts through a 4 ohm load, SOMETHING'S not right and we may as well figure that out before bothering to hear how good or bad it may sound.  As you rightly point out, there are limits to how low this number gets without giving up far more important parts of the circuit design, only one of which is the lack of feedback; but measuring the distortion of a static signal and then expecting it to tell the story of how dynamic music is going to sound through the unit is an attempt to oversimplify things...something we were were suckered into a few years ago and understand that it's very easy to fall into the trap from the outside.

 

Distortion's a very convenient number to look at for people that do not have the resources to test theory after theory as they design a product themselves, or for the casual shopper that cannot spend months auditioning gear, but it's really not a good premise to base your design work or purchasing around as it ignores other factors that are not expressed in simple measurements.  Not to mention the fact that we're all human and thus have our own preferences and biases.  In our experience, hitting low distortion numbers is something that's nice on paper, but we don't buy the products to stare at the spec sheet.  In fact, one of the reason for our circuit design is an attempt to minimize coloration, not introduce it.

Of course, I'm speaking as a manufacturer, so I know my comments will be taken with a grain of salt.  But hopefully it also helps a few people.

[Ryan Berry]

23 minutes ago, Ralf11 said:

I just want to thank barrows for "Ayre chooses not to use global negative feedback in their circuits" - my emphasis - it is all too rare that people distinguish global vs. all or local neg. feedback

 

 

moving on:  Benchmark is using a feed forward circuit in their AH2B

A fair distinction in general.   However, we do not use only zero global feedback. 

[Ryan Berry]

1 hour ago, mansr said:

In other words, it lacked the Ayre house distortion. If an amp stands out, it's doing something wrong, IMO.

I disagree with your assessment. The goal is to be as transparent as possible, not to try and "improve" the music with what the designer may happen to like. 

[Ryan Berry]

17 minutes ago, mansr said:

I agree with that. But then, how does an amp "stand out"? The goal should be for all amps to sound exactly alike: transparent.

That's the million dollar question that we haven't had to answer yet.  If a perfectly transparent amplifier did exist, then I imagine you'd see a lot of companies scratching their heads and trying to figure out how to differentiate themselves amongst the competition.  I believe that with the many designers in the world, most of them have the same goal -- to create a product that is as revealing as possible.  But like Charles Hansen would remind people when discussing design options: there's no such thing as a free lunch.  So designers spend hundreds of hours listening to various ideas and ultimately have to choose what they believe gets them to that goal better than the other options.  Everything is a compromise of some degree.  Through the many innovators in the world using their uniquely talented minds to determine what they feel is the best choice, I believe is why we wind up with varying opinions on the subject of what makes the most important differences in sound quality; and thus products in the market that sound different from each other.  Some go by measurements.  Some go by our very subjective ears.  Some use scientific methods of changing single factors at a time and double-blind testing, others scoff at the notion.  Almost all of us do all of these things to various degrees, I'd imagine.

 

That's not to say that I'm naïve enough to believe that there's not some level of putting a company's "signature sound" in the product, though that's often a result of the overall design the inventors chose.  And of course, other companies may embrace having a "flavor" to the sound: I know my vehicle's audio system is no where near what I would consider clear and transparent...but man is it fun to listen to when I'm on the road.  I'd never let a system like that in my house to relax for the evening, though.  There are also inconvenient things like budgets and needing to make a product that people would by and that you can afford to sell in order to stay in business that factor in, of course; but even removing those silly notions and sticking with straight theory, you're going to see different designers trying to reach the same so far seemingly unattainable point through different paths.  Please note that I'm not trying to speak for any other companies here.  You'd have to get their feelings on the topic from them, but these are my beliefs.

[Ryan Berry]

10 hours ago, Speedskater said:

I want to apologize to Ayre, They are not one of the poorly designed products that I was referring to. (had to dig out my old magazines)

But for an example of a poor design, see the Feb. 2018 Stereophile magazine. (and two good designs)

Thank you for the follow-up, I like to think we work hard to do things right. We often wish right didn't mean expensive too!

[Ryan Berry]

4 hours ago, barrows said:

These things are not debatable, it easy just to look and see what the clock is, then you will know.  A separate transformer for the Ethernet interface is highly advisable, other wise the noise from the Ethernet processing will couple through the transformer to everything else.  The new(wish) Ayre DAC has 4 transformers internally, it is likely that one of these is dedicated to the (built in) Ethernet interface.

Hey Barrows,

 

That's correct.  One of the transformers in the QX-5 is dedicated to the Ethernet PCB, which runs the Ethernet connection and the two USB Host ports for flash drives or WiFi antenna if you plug it in.  We also have separate windings on the other 2 board-mounted transformers for the S/PDIF inputs, the USB section, the Logic, and the FPGA - all powered linearly; and of course the separate stand-alone Mercury transformer dedicated to the Audio portion of the unit.  It's definitely not an inexpensive way to do things, but the hit on sound quality for any of the alternatives was not an option for us.

[Ryan Berry]

12 hours ago, beerandmusic said:

 

I just looked up the QX-5.  That looks like an impressive piece of gear like what I am looking for.  It may be out of my budget, but it looks like you really have something here which is what I have been longing to see....It may be worth reaching for....

 

Can you tell me does it cross off all my checkboxes?

 

DSD256+ on both usb and enet

galvanic isolation on usb

ES9038 in mono mono design

ultra low phase noise clocks

LPS or optional LPS

WIFI

Ability to turn off circuits not in use

jitter reduction circuitry

 

And did i see the price right?  $6K?  That seems very low for AYRE?

Did you skiimp on the dac or somewhere, where separates would offer more SQ?

 

thanks

 

Would be happy to talk to you about the features sometime, but I think we hit most of what you'd be looking for and a few things besides.  I don't want to sound like I'm advertising on the forum (not sure of the rules on that), but can talk in private message or over the phone any time.  I know the price is out of the range, but keep an eye out, you never know what we may be coming out with next...