|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
165.214.11.81
In Reply to: RE: Recent and future amplifier measurements posted by Frihed89 on November 06, 2015 at 11:35:37
What is/are some important factors for class-D amp(s) ?
Follow Ups:
...here are JA's measurements of the Devialet D-Premier D/A integrated amp at $15,995 from 2012.
by the meter readers who profess the *superiority* of performance from switchers - when they exhibit absolutely dreadful square wave response. As Dan A'Agostino observes, they really need to get rid of all the "fur".
JA cannot even get his distortion measurement gear to work - unless he first filters out all the HF garbage! Even at that, you get exceptionally poor results. Hmmm. Yeah, give me more of that! :)
E-Stat, you're a smart guy ( as is Tomservo, who hopefully will chime in here), but studies have shown that square wave output doesn't equate to sound quality to the human ear.
:)
but studies have shown...
Reminds me of the title of my college course, "How to Lie with Statistics" . :)
We don't know everything we don't know.
when they exhibit absolutely dreadful square wave response
Got an example? The result posted for the Devialet looks fine. And I can't say I've read every Stereophile issue, but I can't recall any recent Class D measurements with a particularly bad square wave response. And one of the best I've seen came from the NAD M2.
I understand the concern about ultrasonic switching noise, but like SACD's ultrasonic noise I don't think it's really an issue in practice.
Got an example?That would be every one I've seen.
The result posted for the Devialet looks fine.
We must be either looking at different figures or have very different points of reference. Are you referring to figure 2 found here ?
Did you follow JA's post to the Bel Canto measurements? Look at that one. Tons of fuzz with the filter off. Even then the 10k one isn't exactly pretty with the filter engaged. :)
edit: Here's another beauty.
Edits: 11/09/15
What's wrong with figure 2 from the Devialet measurements? Seriously, it looks fine to me.
The Bel Canto and Anthem examples are pretty fuzzy, I'll give you that. But given that this noise doesn't go anywhere since it's way out of the tweeter's pass band, what is the harm?
we definitely have different points of reference. :)
Are you familiar with the concept of a square ? It's a pretty ordinary shape. Ideally, square wave response would actually look like a square!
Like this!
You do realize that the Devialet is working at a 96 KHz sample rate, right? That's exactly what a square wave SHOULD look like with a brick wall filter response between 40-48 KHz.
And the square wave out of the Ayre would look exactly the same if the source of the square wave was a DAC playing 96 KHz data.
Every audio amplifier is going to exhibit a low pass filter response, but the corner frequency and slope will vary from amplifier to amplifier. These difference will make the square waves look different but as long as the amplifier bandwidth is sufficiently large the difference between a slow or steep LPF characteristic doesn't really matter.
A bad square wave is one which correlate with an in-band performance issue, such as sloping tops which indicate the amp can't handle very low frequencies, anomalies caused by peaky FR, or worse yet, asymmetry.
Here is one example I would consider somewhat problematic (Sharp SM-SX100):
And another (Manley 250 Neo-Classic):
Still waiting to see a good one, yep all bad for sure and still cant understand why no 20 hz SW from JA..
Go Rossi ......
You do realize that the Devialet is working at a 96 KHz sample rate, right?So, JA is using different criteria for this test vs. that of most other amplifiers. I would agree that comparing apples to oranges is problematic.
Edits: 11/09/15
I thought you didn't believe in measurements? Now you're looking at them and making conclusions from them? Ha!!
Dave.
Kindly quote where I've made that unqualified comment.
Now you're making stuff up and laughing at at! LOL for sure. ;)
Ah, so you do believe measurements are valid? At least some of them? :) I knew there was a little objectivist DNA in there....somewhere. :)
I'm laughing because it's funny. Highlighting hypocrisy always is. :)
Dave.
...it's not like believing in Santa Claus because measurements serve a good purpose - equipment designers couldn't work without them.
And sometimes they correlate with what you hear when playing music through a system.
Would you buy an audio component only on specs and measurements without listening?
Maybe you don't believe in them either ;-)
comments about my observation about a hilarious cable *test" whereby two cables are connected together via a ground sharing Y-adapter (this summing the metrics) and used to *compare* each one's individual performance. :)
Do you believe 74, 510 and 626 represent the same value?
He *theorizes* that sources with near zero output impedance (not found in my world) ignore vast differences in MEASURED results.
Sorry. You're missing the context of a previous discussion with E-Stat regarding this. Kind of an inside joke and not really relevant. He's just yanking my chain and I'm just yanking his.
Cheers,
Dave.
Ah, so you do believe measurements are valid?
That's what you find when you aren't making $hit up. Do a search if you have any interest in facts.
Highlighting hypocrisy always is.
As is those who must create their own fictions in an attempt to make some point...
Others, like inmate Ugly actually ask questions and get actual answers.
Good Squarewave responses from class=D ... ! you are kidding ,right ..?
Go Rossi ......
It doesn't get any more textbook than this:
From Stereophile's measurements of the NAD M2.
Per JA's observations in the Stereophile test:
"Although it's convenient to refer to the NAD M2 as an integrated amplifier, it's actually something rather different: the M2 is a multiple-input D/A converter with an output stage that can drive a loudspeaker. "
Perhaps that is the direction others should go. Even then, note that the pretty wave has been filtered from high frequency noise.
I hadn't noticed that caveat on the graph, but JA says there is 100 mV of ultrasonic noise on the unfiltered output which would have fuzzed up the graph. Interestingly, I just re-read his comments on the Devialet and he said there is no ultrasonic noise on the output, which makes me wonder whether that is a benefit of the hybrid design or Devialet just includes aggressive filtering on the output.
@Dave K
The Devialet is the only class-D device i have heard not exhibiting the typical fuzzy high frequency ping usually encountered with other designs . Audible or not something is going on there ...@JA,
John thanks for the response and link
Regards
Go Rossi ......
Edits: 11/09/15
Stereophile's measurements of the NAD M2, whats causing the ringing ..?
Go Rossi ......
A square wave has an infinite number of harmonics. The ringing is what you get when you remove all the harmonics above a certain frequency. The NAD M2 is a digital device, and you're looking at a 1 KHz square wave at 96 KHz sampling, so there is a brick-wall filter response just under 48 KHz. The measurement I linked looks exactly the way a 1 KHz square wave is supposed to look when all harmonics above 48 KHz are abruptly cut off.
Similarly, the Devialet measurement looks exactly the way a 10 KHz square wave is supposed to look given the same digital filter response.
I have no idea. Never heard one. Never owned one. The only ones I ever read about are a new line from Blue Circle Audio and they don't have enough "real" dealers for Stereophile to review their equipment.
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: