Let's clarify this a bit. What you are saying is this, "I, Ralph Karsten, claim that someone who happened to work at GE in the 1950s made a claim during that time period about the relative audibility of even- vs odd-ordered harmonic distortion components. I won't tell you the name of that person, nor give you a reference to the article. I will, however, state that this claim is attributable to GE as a whole, rather than the unspecified person who wrote the unspecified article, and that the claim can be considered as gospel truth, and applies once and for all time. Trust me."
Nope- didn't say that at all. What I am saying is that GE did a study in the 1960s wherein they found that people would tolerate a large amount of 2nd order- up to 30%, but would not tolerated a 'small' amount of odd orders- 0.5% was highly objectionable. I read that in a book, FWIW, a long time ago and don't recall which one and there are a lot of books. GE does not appear to have published the study online. However I can provide a simple procedure that shows a lot of that this is about- its not hard- simple test equipment can be used for the demonstration.
Ralph-You need to go back and reread what you said and I will summarize it as follows:
1. If a speaker requires an amplifier that uses negative feedback to sound its best, it's a flawed speaker design and will never sound real.
2. If an amplifier uses negative feedback, it is a flawed design.
I don't like the effects of negative feedback. It makes whatever is using it to sound brighter than the original signal (if two circuits have the same bandwidth, the one with negative feedback will sound brighter). But to be clear, I think that is more of a flawed approach rather than
flawed designs. The reason for this is that the designer had a goal, and if they were successful with that goal, then design isn't be flawed, even though the overall
approach is. And the Power Paradigm is flawed too- its tricky to get flat frequency response out of the speaker (but if what I see in the measurements is real, than the same can be said of the Voltage Paradigm too- flat loudspeaker response simply does not exist at all).
The
main issue here is that the ear/brain system translates distortion to tonality. And it will favor that over actual frequency response. The Voltage Paradigm for the most part has you stuck with brightness and harshness as a coloration, the Power Paradigm allows you some flexibility.
Here is the “thing”: I’m sure that the Atma-Sphere amps when used with speakers that have complimentary impedance curves (two wrongs make a right?) will sound good. I think what some people take exception to is the attitude that only Atma-Sphere amps get it “right” and all other amps be they SS or tube amps with transformers are “wrong” if they use negative feedback. Pack that onto the statement that any speaker that won’t sing with the Atma-Sphere amps (or amps that don’t use any negative feedback) are speakers that will never sound like “real” music.
A person does not have to use our amps. But you are right, I do personally think my amps are better if not the best
but FWIW its my opinion that a designer should want his/her project to perform as best it possibly can. So many designers are likely to say that they make the 'best'. But let's leave the 'what's best?' conversation out of this thread for now...
And I really agree with that last statement- restated in this way: ' Pack that onto the statement that any speaker that won’t sing with amps that don’t use any negative feedback are speakers that will never sound like “real” music'.
And then you couple the above beliefs with the thought process that has been stated and/or implied that people who don’t “get” what Ralph is saying and believe what he is saying is true are stuck in a false paradigm. If we want to have an amplifier that sounds like “real” music, we all need to have Class A OTL amps that use triodes, no negative feedback, lots of second harmonic distortion, very low dampening factor, and have a high output impedance (although Ralph likes to refer to his output impedance as “moderate”). That is a lot of Kool-Aid to swallow.
"people who don’t “get” what Ralph is saying and believe what he is saying is true are stuck in a paradigm" seems more correct to me.
Actually I am not a fan of lots of 2nd harmonic. And our amps really don't have any on account of their being fully differential- the 2nd is cancelled out at each stage in the amp. So this is a case where the topology can be used to control the distortion signature- we tend to get mostly the 3rd harmonic, and do as much as we can to keep it down, as we do with the output impedance. Is that a Kool-Aid? Dunno. It is certainly an attempt to keep known-distortion sources (pentodes, transistors, transformers) to a minimum, in order to reduce distortion. Some designers don't worry about distortion (SETs are typically 10% at full power, our amps are closer to 1%).
I will be at RMAF in October and I plan on visiting every room that is using Atma-Sphere amps so I can hear the amps with different speakers and form my own opinions. It will be interesting to see what speakers are going to be used with Atma-Sphere amps.
Make sure you introduce yourself! There will be two rooms to the best of my knowledge.
Confidence becomes hubris.
Agreed!
If you are serious digital/SS guy on a high-end forum you get this sort of thing all the time from the valve/vinyl crowd. You get used to it. Ralph came in and applied exactly the same argument, which goes something like "I don't have the data to back it up, in fact the data is not in my corner, but my approach is more like real music than yours," and he applied it very narrowly. He pushed most of the inner circle out into the cold and unmusical, and left a very small elite. The old elite won't suffer their banishment gladly.
FWIW I have said nothing of the sort... I would not even characterize the above statement as paraphrasing. Its simply outright wrong. There is plenty of data in my corner, some presented here, some of it historic, which I have presented (follow the links and references). I have pointed out that many will regard my position as blasphemous and will tend to discount the data as it threatens their world view.
Applying exactly the same systematic, hard-wired, unalterable signal processing to everything from crude early Charlie Parker recordings to modern digital studio recordings, and expecting it to have the universal effect of bringing the listener closer to real music is not even rational, much less correct. You may love the way it sounds, but correct is certainly not a word I'd apply. I don't doubt this stuff can sound really good, I'd just like to see one of you admit, just once, that what you're doing is altering the recordings to taste, not bringing them closer to "real music," or some other euphemism for the signal integrity you cannot claim. And I think if all that energy and passion was going into developing great signal processing tools that could be adjusted to recordings and rooms, it would do a lot more good.
Tim
I agree. I prefer to go without signal processing of any sort.
This topic comes up surprisingly often, and you will find components that adhere to this so-called "power paradigm."
But to answer your question about how many are designed for CONSTANT VOLTAGE, the answer is: almost all. The reason I know this has to do with speaker measurements and how they're done. The standard is to applied a CONSTANT 2.83V across the frequency band, and with that the speaker comes out as flat as its intended to. In other words, it's made for a constant voltage. Although I'm sure they exist somewhere, I've personally never seen a speaker designed for constant power.
I first encountered this issue with an Orpheus Labs amp I reviewed VERY long ago -- it was a solid-state "constant power" amplifier. It had a special circuit so that its power remained the same regardless of load.
Doug Schneider
www.soundstage.com
Doug, you have been in my rooms a number of times over the years and shows. So you have seen some Power Paradigm loudspeakers because I tend to show with them- it gets better results. I have been doing this a long time- before Marc reveiwed the old MA-1 on your site years ago.
Another anti-two channel rant brought to you by Tom. If everyone felt the way you do Tom about how deficient two channel stereo sounds, it would have died out a long time ago. We would all be stuck with some sort of multi-channel sound. The majority of people on this forum have high-end two channel systems and love the illusion their systems and rooms provide. Your constant beating on the stereo illusion with both the statements you make in your posts and your tag line that is on everyone of your posts hasn't swayed anyone that their stereo doesn't sound as good as they think it does.
mep, you and I are on the same page in this regard. I might add further that I am involved in the local music scene here in the Twin Cities. Every band that makes a recording in this town does it on a two-channel format. Two channel is here to stay for a long time.
Fransisco-Go back and read what Ralph said about the bad tubes and the way Bascom King took the measurements. I *believe* that Ralph said the reason the MA-1 didn't meet the specified power output was because of the defective tubes and the way Bascom set the variac. Ralph went on to say in another post that his newer amps have 90% less distortion than the MA-1 did which means that his current amps should measure 1% THD.
Correct. From the readings it also appears that Bascomb had a test instrument connected to the output terminals of the amp that shorted one of the speaker terminals to ground. The output section floats relative to ground (ala ARC, Classe and others); if one side is shorted the distortion will go up dramatically and output power will be reduced as the resulting drive to the power tubes becomes assymettrical.
Ralph has touched on it........a bit. in this discussion this thread, the amplitude changes are significant enough already without analyzing output spectrum even though it also behaves differntly feedback vs non feedback. there is no magic here in the world of electronics. the magic is in the ear/brain interface.
a SET amp is quite something, we got dynamic gain with frequency and dynamic harmonic spray, and it helps plain old stereo sound more alive (IMO), and that aint no jive! but if there are only a few instruments playing, say jazz or simple rock, OK, however if playing full orchestra...faagettt it, IMD gets crazy bad, sound goes to mudd.
For the most part I heartily agree- except that it should be noted that if you plan to use an SET, the speaker efficiency should be such that the amplifier will not be asked to use more than about 20-25% of full power. This will keep the distortion so low most of the time that it can't be measured. But when it
does appear, it will appear on the peak transients. So odd-ordered harmonic distortion will appear on those transients, meaning that the loudness cues to the ear will be on those transients too. If you have ever wondered why SETs have a universal reputation for being far more dynamic than their power level would suggest, this is the reason why.
IMO/IME 90% of the time when audiophiles use the term 'dynamic' you can safely change the word to 'distortion' without changing the meaning of the conversation at all!
they are all tone controls. for example, our solid state voltage feedback amps, when the second harmonic comes back via feedback (global, or local) and mixes with the original signal we develop a third harmonic, thus third is dependent also on second. i did some tests with the amp shown at left when preparing it for publication but I lost those pages of my original manuscript, but it corroborates with the power paradigm anyway (its a super simple set built to reveal SET attributes at an inexpensive price point and prove something about frequency range vs satisfying sound).
I beleive I have been on topic with this quite a lot in this thread where I have pointed out that the ear/brain system translates distortion into tonality. Anything that makes distortion is also going to have a tonal signature. IOW on this topic we seem to agree.
