[Freo-1]

They are a priori snobs. What their particular thoughts on this particular subject makes them is simply ill-informed. But being ill-informed is pretty much the "audiophile's" lot, so really nothing new there at all.

I think you can put yourself at the top of that list. 

All equipment falls short of live music.  So, we put systems together that get one the closest based on their preferences.  Since preferences, differ, the solution differ.

[DS-21]

All equipment falls short of live music.  So, we put systems together that get one the closest based on their preferences.  Since preferences, differ, the solution differ.

There are preferences grounded in reality (i.e. actual differences, such as between speakers with consistent midrange directivity and speakers that throw a mushroom cloud of excess midrange energy into the room at the bottom of the tweeter's passband) and there are preferences that lack any foundation whatsoever in the real world (i.e. most of what so called audiophiles blather about). I focus on optimizing the former to my tastes, which tends to give better results than wasting one's time with nonexistent idiocies.

PS: Have you yet bothered to check out any of the Powercube results of real amps published by The Audio Critic, or did that topic become uninteresting when the existence of actual data constrained your ability to render unfounded speculation on what the device might "prove" about your mythology?

[medium jim]

The Hypex amp modules (UcD as well as Ncore), and the newer (ASX2, System 7) Icepower modules both seem to fix most of the traditional problems of Class D, though.

But not all of them.

Jim

[Diamond Dog]

I, for one, would like to see some scientific research done into whatever aspect of Class D amplification apparently causes some people to lose any sense of civility when they are in contact with it or involved in discussion regarding it...

Let's work towards a cure.

D.D.

[Rclark]

The Hypex amp modules (UcD as well as Ncore), and the newer (ASX2, System 7) Icepower modules both seem to fix most of the traditional problems of Class D, though.

But not all of them.

Jim

what does this post even  mean?  You've never had any of these in your system.

[neekomax]

I, for one, would like to see some scientific research done into whatever aspect of Class D amplification apparently causes some people to lose any sense of civility when they are in contact with it or involved in discussion regarding it...

Let's work towards a cure.

D.D.

So much win has this comment.

Interesting discussion, please carry on. I am considering amplification, and trying to learn as much as possible. So far, the most salient point, I think, is DS-21's comments about impedance loads affecting frequency response with some amps (tube and class D, from what I can tell, perhaps other topologies as well?). This seems huge to me.

[medium jim]

what does this post even  mean?  You've never had any of these in your system.

Or have you ever had tubes in yours.  There are intristic flaws with class D that may not be curable.  This doesn't mean that they can't sound good.  Many like the way they sound, but all the comparisons to tube amps is ludicrous as it truly is apples and oranges. 

Then there is the hubris of some that class d makes all other amps inferior.  This mentality leads nowhere.

Jim

[Rclark]

Again, what are you rambling about? Lmao.

[milford3]

A quick explanation of how a class D amp works.

http://www.axiomaudio.com/1500-amplifier#t-5

[medium jim]

A quick explanation of how a class D amp works.

http://www.axiomaudio.com/1500-amplifier#t-5

Try

http://lenardaudio.com/education/12_amps_6.html

or

http://www.nuforce.com/index.php?option=com_k2&view=item&id=620:q-what-are-the-problems-with-class-d-digital-switching-amplifiers&Itemid=405

When a Class D amp fails, who is going to be able to repair it?

Jim

[Russell Dawkins]

Antony Michaelson of Musical Fidelity had this to say (large missing segments indicated by "....."  and italics his):

"Dear Fellow Music Lover,
Many customers ask me about the design aims of my products. They are particularly puzzled by how Musical Fidelity can strongly
promote the merits of high powered amplifiers such as the 550K or 750K whilst on the other hand producing low powered amplifiers such
as the A1. I’d like to try and explain my ideas.
.....
Accuracy and realism.
I strongly believe that with an ordinary high quality loudspeaker of about 87dB sensitivity you’ll need about 500 watts to avoid any
clipping or limitation. I cannot stress strongly enough that this assumes that you are using the system at a healthy, quite loud, listening level.
If you listen quietly or are in a small room or sit close to the loudspeakers, the high power figure is irrelevant. My point is that if you want
accuracy and realism the amplifier should never clip or limit.
.....
Keith Howard is an experienced technical journalist. His research has confirmed that 500 watts power is a good level for a properly neutral
system. He also suggests that if you really want no clipping, that you would need even more power. His position certainly appears to agree
with ours but there is more to his research. After finding out how much power he thought the system should have, he then set out to find
out how much clipping would be subjectively acceptable without much apparent loss of neutrality.
After a lengthy experiment, Keith found that about 3dB of clipping did not cause much upset. This equates to 250 watts.
To summarise the above: if you want to be sure of almost complete accuracy and neutrality you need an amplifier of 500 wpc (or more) but
250 wpc will be acceptable for most people most of the time. A little inaccuracy never did anybody any harm.
.....
Once you are below 100 watts clipping (with my previous points taken into account) is a regular feature. It follows that the dynamic
reproduction is no longer neutral, accurate or low distortion. Once you’ve reached this area subjectivity becomes far more relevant. Each
amplifier has a unique clipping/limiting characteristic.
.....
Once you’re down to below 100 watts amplifiers will be clipping regularly. At this point reproduction is inaccurate. Now the amplifier
design is rather like an impressionist picture. It cannot be a photographically accurate representation of the recording but somehow has to
create the impression of the musical performance. At this level the limitations of the amplifier have to be accepted as a constant and the
design has to be created in order to give the most convincing illusion of musical accuracy.
.....
Now to the A1. Yes it doesn’t have much power and yes it’s clipping regularly (but not much more than any other amplifier below 70 or
100 watts) but we have accepted its power limitations and have decided to engineer it to produce a convincing, beautiful, musical
experience.
When I listen to the A1, I am completely blown away by it. It has an apparently sweet, subtle, delicate sound which seems to make
something of a nonsense of my high powered ideals.
However, as soon as a you switch to an amplifier which has enough power not to be clipping, it is immediately obvious what you have
been missing. The point is that whilst listening to the A1, its ability to project and create musical illusion is so strong that I don’t object to
its power limitations. In my world, that’s job done.
.....
Over the years I have had endless frustrating run-ins with reviewers and hi-fi salesmen about the accuracy of what they’re hearing. When I
have taken them a proper, high powered amplifier to use, their immediate reaction was that it was “dull” or “undynamic”. What they were
hearing was completely faithful and accurate but what they wanted (or perhaps expected) was the usual clipping and limitation. Very, very,
very frustrating."

Excuse the length of this post, but I thought some would benefit from the clarity of this, from an honest designer. I post this because it seems not to be on MF's website any more.

[medium jim]

There just isn't a magic bullet answer to the Subject question.  There are too many variables, such as speaker impedance curve, the efficiency of the speakers, to name a couple.  Then there is the way different topologies, not to mention how tubes vs. SS react to the varying impedance's of the speaker.  The general rule of thumb is that a tube amp will require about half the wattage of a comparable SS amp to do the same thing. 

Having both SS and Tube amps, the SS being around 425 per channel (4ohm) and the tube about 110, I find the tube amp to flesh out all the detail and has just as much dynamics.  But that is in my system and YMMV.  Some speakers mate better with tubes, some better with SS.   Then there is personal preference, I like the warmth of tubes and the soundstaging they produce.  Many prefer the slam and dynamics of SS.   Modern day amp builders are bridging the gap between the species, but are not there just yet, unless you plan to spend obscene amounts to get to the promised land. 

Jim

[Freo-1]

There are preferences grounded in reality (i.e. actual differences, such as between speakers with consistent midrange directivity and speakers that throw a mushroom cloud of excess midrange energy into the room at the bottom of the tweeter's passband) and there are preferences that lack any foundation whatsoever in the real world (i.e. most of what so called audiophiles blather about). I focus on optimizing the former to my tastes, which tends to give better results than wasting one's time with nonexistent idiocies.

PS: Have you yet bothered to check out any of the Powercube results of real amps published by The Audio Critic, or did that topic become uninteresting when the existence of actual data constrained your ability to render unfounded speculation on what the device might "prove" about your mythology?

Look mate,if you can throttle back the attitude a bit, we can have a decent conversation, OK?  You no doubt have a fair knowledge level,  so there is no need to skewer someone if they fail to agree with your positions.   I get the fact you place a large value on measurements, and I  agree with that premisis (to some extent). 

Regarding the power cube issue,  I have looked into the subject some more.  One rather significant change I wound up making was swapping out my Pass Labs XA 30.5 for a INT-150.   The overall sound from the INT-150 (with my 88db/w speakers) is just better.  There is more power when it is called for, only two gain stages, outstanding noise floor, etc.  So, it's all good.
Where (I think) we may part company is on the subject of tube gear.   I have tube DIY preamp and power amp setup.  The power amps output roughly 55 WPC at 8 or 4 ohms.  They sound different to the Pass Labs gear.  On some music, the sound from the tube gear is preferable (more realistic/alive sounding) than  the solid state setup.  On other recordings (digital with a lot of dynamics), the extra clean power from the solid state setup is desirable.  I'm not sure "when playing music" there is a full set of measurements that can capture the perceived sonic differences adequately.   There simply are not a full set of  "facts" to explain why there is a audible difference.  We have a lot of knowledge about how all this gear works, but let's face it:  The transfer function of amp technologies are not the same.

Lastly, Class D amps certainly have improved overall.  Sony actually made  decent sounding Class D gear some years ago (S-Master Pro), so they have and are a viable option for many audiophiles.  However, many other audiophiles continue to be less than enamored with their overall sonic qualities.  The fact that they can produce individual frequencies well does not translate into reproducing music flawlessly.   You can't point to a set of measurements to prove or disprove the sonic qualities of Class D.  The human ear is a very sensitive device, and the typical measurements do not adequately capture the time-domain aspects of reproduction that people pick up on.
 
 

[DS-21]

The Hypex amp modules (UcD as well as Ncore), and the newer (ASX2, System 7) Icepower modules both seem to fix most of the traditional problems of Class D, though.

But not all of them.

Jim

Such as? I mean real things, as opposed to the baseless and silly innuendo one can find in reply 52, supra.

Now, if one prefers tubes, one will also tend to prefer a less-advanced Class D amp than the best of the current breed, simply because less advanced Class D amps they have the same basic flaw that amps powered by overpriced lightbulbs have: high source impedance. ("source impedance" and "output impedance" are synonymous)

For example, I have an old Panasonic SA-XR55 AVR ca. 2004. That AVR became a bit of a cult hit, because it was cheap and had some cool technology: it did direct PCM->PWM conversion, which is to say D/A conversion and amplification were effected in a single stage. Go back to "audiophile" commentary at the time, and you'll see people were talking about how "tube-like" those Panasonic DAC-amp receivers were. Here's an example. And they were right. The XR55 has a source impedance of around 1Ω. That's downright SET like, about twice as high as the source impedance of a McIntosh MC275, and ~8x higher than the source impedance of a typical modern high-fidelity amp.

Interesting discussion, please carry on. I am considering amplification, and trying to learn as much as possible. So far, the most salient point, I think, is DS-21's comments about impedance loads affecting frequency response with some amps (tube and class D, from what I can tell, perhaps other topologies as well?).

Most Class AB amps are designed with sufficiently low output impedance to be basically flat into any loudspeaker. There are exceptions, such as "Sideshow Bob" Carver's Sunfire amps, which offer a tap with a resistor behind it to emulate the FR errors caused by the high output impedance of a typical amp, or a less-than-cutting-edge Class D amp.

Having both SS and Tube amps, the SS being around 425 per channel (4ohm) and the tube about 110, I find the tube amp to flesh out all the detail and has just as much dynamics.  But that is in my system and YMMV.

"Flesh out the detail" is usually audiophile-speak for "elevated midrange." And that's one thing tube amps often do: elevate the midrange because of their high source impedance, as discussed above.

To flesh that out, let's look at some measurements from Stereophile. In each case, focus on the black curve. (They call it grey, but it looks black to me.)

Here is the frequency response into a simulated loudspeaker load of the most recent tube amp Stereophile tested, the "Octave Audio RE 290 power amplifier," which has a source impedance greater than 2Ω:



By comparison, here is a high-fidelity amplifier with source impedance measured at ≤0.16Ω (Mark Levinson 532) into the same simulated speaker load:


And another high-fidelity amp with output impedance measured at ≤0.17Ω (Bryston 7B-SST), into the same simulated speaker load:


(NOTE: just because those two are expensive, does not mean lots of money is required to get a high fidelity amp. Stereophile just hasn't tested a reasonably-priced AB amp recently, and I didn't want to scroll through their tests forever.)

As for class D, I'm not aware of a Hypex-based amp that was measured into Stereophile's simulated load, but here is how the Icepower 125ASX2 module, as found in a recent Bel Canto integrated amp, measures into the same simulated load:


Will a tube amp sound different from the three high-fidelity amps? (Those three amps won't sound different from one another to people who listen with their ears rather than their eyes or wallets, so long as the 125ASX2 can reach sufficient levels at a given listening position in a given room with a given pair of loudspeakers.)  Considering that there are FR differences in excess of 1dB in the 1-2 kHz octave even into the relatively forgiving Stereophile/Kantor simulated load, a attentive listener should hear a difference between them even level matched.

Now, given there is an actual, non-illusory difference between the high-fidelity amps and the lightbulb thingy, one could of course prefer the lower-fidelity reproduction. But...I suspect if someone replaced a tube amp with a commodity AB amp used miniDSP with parametric filters in front of the AB amp to mimic the effects of its source impedance on the set of loudspeakers used for the comparison, there would be no differences to hear.

The bottom line is simply that there are a lot of things that "audiophiles" think matter, but actually don't. Or, at least, the differences they imagine are rather more exotic and esoteric than the actual mechanisms at play.

But as for the topic at hand, power required to reach a given SPL, that's something easily calculated if one can fill in all of the relevant variables. One needs to know the speakers' broadband sensitivity (as measured into 2.83V at 1m), and then measure the actual falloff with distance in one's actual room in the midrange/treble. Then add a little power for fudge factor, to compensate for things like thermal compression in the loudspeaker.

[andy_c]

As for class D, I'm not aware of a Hypex-based amp that was measured into Stereophile's simulated load, but here is how the Icepower 125ASX2 module, as found in a recent Bel Canto integrated amp, measures into the same simulated load:

Stereophile measured the Channel Islands D-100, which is Hypex UcD-based, in 2005. The curve with the slight wiggles is the response plot into the simulated loudspeaker load. Looks very good!

[Pete Schumacher]

When we see speakers with a specified frequency response of +-1.5dB, we think of them as pretty darn nice.

An amp that stays within that spec should be just fine too.  And what if the speaker being driven has a "dip" in the midrange where the tube amp has an extra dB to give? 

The only way to really know how the amp will "sound" is to hook it up to the speakers and let 'er rip.

[medium jim]

When we see speakers with a specified frequency response of +-1.5dB, we think of them as pretty darn nice.

An amp that stays within that spec should be just fine too.  And what if the speaker being driven has a "dip" in the midrange where the tube amp has an extra dB to give? 

The only way to really know how the amp will "sound" is to hook it up to the speakers and let 'er rip.

+1

DS-21:

Flesh out Detail has nothing to do with midrange....in fact, the first time I heard the expression was a review about a SS amp, go figure....I guess you have a different meaning to suit your needs. 

Moreover, A/B topologies are also employed by tube amps, you can even have Class B/D, go figure.   I agree with Freo-1.....much of this is born by each individual's ears and cannot be quantified by science or graphed.   It would be nice if we could enjoy the differences each of has and allow all to have fun instead of crapping on others likes and dislikes.  But that would be too damn civil for some.   

Jim

[roscoeiii]

When we see speakers with a specified frequency response of +-1.5dB, we think of them as pretty darn nice.

An amp that stays within that spec should be just fine too.  And what if the speaker being driven has a "dip" in the midrange where the tube amp has an extra dB to give? 

The only way to really know how the amp will "sound" is to hook it up to the speakers and let 'er rip.

+2.

And probably time for someone to provide a link to Atma-sphere's paper on power vs. voltage paradigms. So I will channel Ralph and give this:

http://www.atma-sphere.com/Resources/Paradigms_in_Amplifier_Design.php

[medium jim]

+2.

And probably time for someone to provide a link to Atma-sphere's paper on power vs. voltage paradigms. So I will channel Ralph and give this:

http://www.atma-sphere.com/Resources/Paradigms_in_Amplifier_Design.php

+1   Excellent link


Jim

[andy_c]

It is an amazing link, but not for the reasons you might think.  I'd like to cut the writer some slack and refer to him as merely a brazen BS artist.  Anybody who knows anything at all about amplifier design will immediately recognize what's written there as complete nonsense.  I wonder if the writer actually believes his own BS, because if he does, he is utterly incompetent and unlikely to have the ability to design a decent amp to begin with.  So I think this is conscious BS artistry and not an attempt to describe legitimate engineering.

Unfortunately, this situation has become the norm in the so-called "high end" of audio.  Many of the people designing this stuff don't have a clue WTF they're doing, but they really know how to do the marketing thing.  It's sad, really, because so many people not only fall for it, but actually think that people who do not fall for it are in the wrong.