[Eric_C]

Quote:

 
Did you not read the thread title ?
 
It is impossible to get measurements on the amp you mention and several of their earlier products measure abysmally !

 
Quote:

If you want a real good headphone amp, get a PS Audio GCHA. Dead silent, detailed and very affordable nowadays...
And if interested in something even better, I have one too many amps and would not mind to sell or trade my PS Audio GCHA with Cullen mods...
 
No matter what was said about it, this thing is gooooooood..

 
Haha it would seem karmanfamily doesn't even own the amp he's recommending:
http://www.head-fi.org/t/589416/cullen-gcha-ps-audio#post_8574347

 

[kwkarth]

An objectively "good" amp, as I would describe it, is the one in the LME49600 application note. Quite a few commercial and kit designs use variants of it, and the original uses a top-spec (LME49200) op-amp and has a DC servo. Powerful, low distortion, low output impedance,  and devoid of the headaches inherent in discrete semiconductors (which only outperform monolithic chips if you've done some very careful hand-matching.) And not too expensive, either - I think it specs out at $30-$40 in components, and quite a bit less in quantity. 

[size=medium][COLOR=1122CC]LME49600 Headphone Amplifier Evaluation Board User's Guide[/color][/size]

Same thing with the ODAC. Low jitter, low distortion, low noise, not particularly complicated. 

As far as I can tell, the difference between the two is a Corvette -vs.- Porsche thing. The Porsche has better looks, better fit and finish, better feel, better seats, better noise insulation, and a much more storied history than the plastic car from Kentucky. But put them on the Nurburgring, and a $48,000 Corvette can keep pace with Stuttgart toys at twice the price. 

Nuforce and Schiit appear to operate under the old-school "fiddle with it until it sounds good" school of audio engineering. This often doesn't produce ideal results, especially with inclement operating conditions, and often omits things like protection circuitry. 

I'm not anti-luxury. Far from it, in fact - I'm sorely tempted by some secondhand Audeze LCD2s, which are the first headphones I've ever seen to post results that can seriously compare to high-end loudspeakers. However, while fancy engineering tweaks are still required to make a transducer produce a respectable square wave, a $3 op-amp does it just fine. 

In any case, your post is your purely subjective opinion, to which you're entitled. Your opinion of the design philosophy of the companys you mentioned is vastly different from what they publish and deliver.

 

[mikeaj]

Quote:

In any case, your post is your purely subjective opinion, to which you're entitled. Your opinion of the design philosophy of the companys you mentioned is vastly different from what they publish and deliver.

 
I mean this sincerely:  can you describe their design processes then?  (I don't mean to single anybody out, but these companies are well known and were already brought up.)
 
Do they just not do enough testing before releasing a product?  How do they evaluate a design or prototype in order to iterate and improve it?  Both companies ultimately do not seem that interested in "meter reading" and such, as even is mentioned in one's marketing materials, but how much does that play a role in development?  What kind of listening do they do in development, when do they do it, and do they use the usual experimental controls (blinding protocols, multiple trials, baseline comparisons, and so on)?

 

[Spasticteapot]

Quote:

In any case, your post is your purely subjective opinion, to which you're entitled. Your opinion of the design philosophy of the companys you mentioned is vastly different from what they publish and deliver.

 
The problem here is that these companies use a subjective design philosophy. This thread is about objective design, in which the entire signal path should ideally be a straight wire with gain.
 
There are some esoteric bugbears out there like intermodulation distortion on which I'm not hugely qualified to comment, but the solutions often involve extremely expensive high-linearity JFETs. (Nelson Pass, AFAIK, actually has his custom-made to spec.) It's possible to build a low-distortion low-noise low-TIM amp with very little feedback and high stability into both inductive and capacitative loads, but this ain't it. 
 
Mr. Pass sells amps that cost as much as a car. I personally think it's in the realm of severe overkill, but I won't fault him for it, because Mr. Pass does flawless work and is extremely free with his knowledge. You pay through the nose for improvements of dubious audible benefit, but the amp does exactly what it says on the box. 

Mickey-mouse engineering on stuff this expensive is not acceptable. 
 
EDIT: 
If you'd like to look more into Mr. Pass' design philosophy, he's written a great deal on the subject, including a lot of DIY designs. Nice guy. 

 

[adydula]

Hello...
 
The corvette that runs around with the euro cars costs a wee bit more than $48K....I have one its more like...$77 - 100k....Corvette that is....

 
I have several Schiit amps...and have an O2 Objective amp. I have 50 hrs now at least on the O2 and hundreds on hours on the Lyr and Asgard with Audeze LCD2's, AKG 701's, Grado 325is and AudioTechnica 50's...
 
Its very, very refreshing to see an objective minimalistic approach to designing a headphone amp...the O2 provides a very inexpensive amp that does measure well, and its design has been published for all the world to see. You dont see that from many vendors.
 
This doesnt mean that other vendors make less than stellar stuff. You can pay lots of dollars for stuff and get good measurements and sound...and you can pay lots of dollars for stuff and get less than stellar results but the stuff might weigh alot or look good.
 
I like using the least amount of stuff on the straight wire with gain challenge.....the less there is the less there is to get in the way...and objective measurements go a long way in level setting.
 
I have purposly not mentioned anything about how the O2 amp compares to the other amps....dont want to run this thread in the wrong way. If you want that information PM I will be glad to share my experiences with anyone.
 
That said for $144 the O2 Ojective amp is a "steal" for what it does and does not do..
 
All the best
Alex

 

[kwkarth]

In any case, your post is your purely subjective opinion, to which you're entitled. Your opinion of the design philosophy of the companys you mentioned is vastly different from what they publish and deliver.

I mean this sincerely:  can you describe their design processes then?  (I don't mean to single anybody out, but these companies are well known and were already brought up.)

Do they just not do enough testing before releasing a product?  How do they evaluate a design or prototype in order to iterate and improve it?  Both companies ultimately do not seem that interested in "meter reading" and such, as even is mentioned in one's marketing materials, but how much does that play a role in development?  What kind of listening do they do in development, when do they do it, and do they use the usual experimental controls (blinding protocols, multiple trials, baseline comparisons, and so on)?

Read each company's web pages. Look at the design career experience of the designers in the companies. What the companies espouse, and what the historical facts document is quite the opposite of the slanderous post that was put up earlier.

 

[kwkarth]

While it's certainly not my job to defend these companies, I think it should be evident that their designs are products of carefully considered trade offs and compromises. Certainly not the slap dash, shoot from the hip process that they're being accused of. The only thing cavelier about the design process we've been discussing are the foolish assertions that have been made in this thread. If one happens not to agree with the overall net design philosophy of a given company, that's one thing and one is free to spend one's money elsewhere. To me, the height of foolishness is to choose purely by published spec, particularly since many of those throwing about buzzwords do not even understand the correlation between the measurement considered and the resultant sound signature.

To reiterate, I don't recommend anyone make foolish or ignorant charges or make decisions based upon half baked and misunderstood data. Such foolish slander is not cool at all.

 

[kwkarth]

In any case, your post is your purely subjective opinion, to which you're entitled. Your opinion of the design philosophy of the companys you mentioned is vastly different from what they publish and deliver.

The problem here is that these companies use a subjective design philosophy. This thread is about objective design, in which the entire signal path should ideally be a straight wire with gain.

There are some esoteric bugbears out there like intermodulation distortion on which I'm not hugely qualified to comment, but the solutions often involve extremely expensive high-linearity JFETs. (Nelson Pass, AFAIK, actually has his custom-made to spec.) It's possible to build a low-distortion low-noise low-TIM amp with very little feedback and high stability into both inductive and capacitative loads, but this ain't it. 

Mr. Pass sells amps that cost as much as a car. I personally think it's in the realm of severe overkill, but I won't fault him for it, because Mr. Pass does flawless work and is extremely free with his knowledge. You pay through the nose for improvements of dubious audible benefit, but the amp does exactly what it says on the box. 


Mickey-mouse engineering on stuff this expensive is not acceptable. 

EDIT: 
If you'd like to look more into Mr. Pass' design philosophy, he's written a great deal on the subject, including a lot of DIY designs. Nice guy. 

Let's be honest here. These companies use objective data to make subjective design decisions. That is what the design process is all about, O2 amp design included. Design is a subjective decision tree process. Design iteration is a process used to gather more objective data about considered trade offs, in order to make subjective decisions about what trad-offs to employ in the final design.

Sometimes, the wrong decisions are made in terms of meeting final design criteria, and the mark of good design is to incorporate these new data in future design iteration.

 

[xnor]

Quote:

Let's be honest here. These companies use objective data to make subjective design decisions.

 
Objective data as in the specs of some component or real-world measurements? The former are often used in a copy-paste fashion in the specs of the final product..
What about companies that design by ear (only), there doesn't seem to be much objectivity involved if at all, which explains the often just mediocre measurement results done by random guys on the internet, which is actually job of the manufacturer?

 

[mikeaj]

re: using objective data to make subjective design decisions
 
This actually may be a less charitable explanation than that offered by others, saying that it's all about fiddling around until it works.  Given the results, this maybe implies some level of heavy cost-cutting, incompetence, and/or laziness—or it could just be some kind of cost/benefit analysis or philosophy I don't understand, which could count as incompetence, depending on your perspective.  Actually, I don't really have any objections to heavy cost-cutting measures, but that doesn't seem to jive with the manufacturing, marketing, pricing, and so on.
 
 
On a side note, we have the classic contrast between those here who say "minimalistic" and mean not reinventing the wheel, doing what it takes to reach the goal, lower parts counts if something more complicated is not necessary; and the blurb on Schiit's site about "simple", meaning class A, no op amps (circuit inside is too complicated? I guess), no feedback on power amps, etc.

 

[kwkarth]

Let's be honest here. These companies use objective data to make subjective design decisions.

Objective data as in the specs of some component or real-world measurements? The former are often used in a copy-paste fashion in the specs of the final product..
What about companies that design by ear (only), there doesn't seem to be much objectivity involved if at all, which explains the often just mediocre measurement results done by random guys on the internet, which is actually job of the manufacturer?

I'm not aware of ANY company that designs by ear only.

 

[xnor]

Quote:

I'm not aware of ANY company that designs by ear only.

So?
Btw, have you noticed that I put the only in brackets?

 

[kwkarth]

re: using objective data to make subjective design decisions

This actually may be a less charitable explanation than that offered by others, saying that it's all about fiddling around until it works.  Given the results, this maybe implies some level of heavy cost-cutting, incompetence, and/or laziness—or it could just be some kind of cost/benefit analysis or philosophy I don't understand, which could count as incompetence, depending on your perspective.  Actually, I don't really have any objections to heavy cost-cutting measures, but that doesn't seem to jive with the manufacturing, marketing, pricing, and so on.


On a side note, we have the classic contrast between those here who say "minimalistic" and mean not reinventing the wheel, doing what it takes to reach the goal, lower parts counts if something more complicated is not necessary; and the blurb on Schiit's site about "simple", meaning class A, no op amps (circuit inside is too complicated? I guess), no feedback on power amps, etc.

Hmm, again I am not aware of ANY professional company that designs by trial and error. That would be a certain recipe for disaster. Having been involved in and responsible for product designs for professional companies, I can tell you that we always started by creating clearly stated design objectives (a target) and moved forward from there. Even product definition was an iterative process. I've seen some projects (not mine) fail because they never had a very clearly defined product definition before they started flailing and failing.

 

[kwkarth]

I'm not aware of ANY company that designs by ear only.

So?
Btw, have you noticed that I put the only in brackets?

Sure, but how do I know what you had in mind when you did that? So why don't you tell us!.

 

[xnor]

Don't you remember the nuforce udac2 drama?
 
"better sounding wins over better measuring"

and then they admitted the measurements of their own product surprised them! How did they miss the distortion products only ~40 dB down? Oh yeah, by listening.