Are audio companies all involved in a huge conspiracy? - Page 43

Thanks arnky for all your help! Very knowledgeable chap! I really do wish to learn more about human hearing and understand why we hear what we do, because what you are saying makes a lot of sense re perception. I think some people are under the impression that human ears are like precision audio analysers. We can hear everything ... but what little I know is that our ears are very insensitive to these tiny measurable changes. Very interesting stuff but I need to learn more.

Do you know of any reading material I could look at that goes into this in more depth? Thanks again for your knowledgeable contributions.

Piling on what Roger said, and related to the point you just made, listening tests show that we get less and less sensitive to fidelity problems in our systems as the number of channels increases. Best discrimination occurs with mono, i.e. one speaker and we progressively lose our ability to hear artifacts as we climb up to 5 channels. For this reason, stereo reproduction tends to be a more critical scenario than multi-channel.
By far the best text on this is Sound Reproduction: The Acoustics and Psychoacoustics of Loudspeakers and Rooms by Dr. Floyd Toole: http://www.amazon.com/Sound-Reproduction-Acoustics-Psychoacoustics-Loudspeakers/dp/0240520092/ref=sr_1_1?ie=UTF8&qid=1359752885&sr=8-1&keywords=floyd+toole

It will teach you everything from effect of speaker wire to advanced topics related to speakers and acoustics including the point above. Can't recommend it highly enough. You will learn more in a week of reading that text than spending months here .

That caught my eye. Has it been tested if it is number of channels or number of speakers? Is there a difference in sensitivity with one surround speaker per side vs. (as in my case) three? Or is it just bound to channels?

If you ever try to do an ABX test for some small but audible difference with both a stereo system and a surround system, you usually pretty quickly figure out which one (stereo) you want for best sensitivity to small differences. There may be some exceptions, but I don't know what they are.

The test ("Comparison of Loudspeaker-Room Equalization Preference for Multichannel, Stereo, and Mono Reproductions: Are Listeners More Discriminating in Mono? " by Sean E. Olive, Sean M. Hess, and Allan Devantier" )was using the ITU recommended configuration for surround system, i.e. 30 degrees for for the fronts and 110 degrees for the rears. The listener was in the center of the 3.9 meter circle. They did not go beyond 5.1 so I can't answer your question directly. But the trends were pretty clear. Here is the actual text from the paper:

"There was a significant interaction between the
different loudspeaker equalizations and the
number of playback channels. The effect was
largely isolated to the unequalized loudspeaker;
its preference rating decreased monotonically as
the playback channels were reduced from five
(surround), to two (stereo), to one (mono)."

In other words, the system with no equalization and hence lowest scores, had its score go up as we added more channels. And in doing so, its gap relative to the equalized systems became smaller and smaller. The only change was the addition of more channels/speakers. They go on to say it more clearly:

"The third conclusion that listeners are more
discriminating of spectral defects in mono versus
stereo and multichannel playback modes is both
thought-provoking and reassuring."

I think the reason for this is that our illusion of what is being played gets more and more real as we add more channels. And with it, we look past any distortions that might be there. In that sense then going to 7.1 and 9.1 will keep adding to that effect.

Isn't it typical... when I do need to ask something, it's not known... same as when I ask questions in stores...

I'll venture a guess. It's not the number of speakers, but the amount of information. Do two speakers driven in mono score any differently than one speaker?

Having said that, when you have 2 or 3 surrounds carrying a single channel, they become a form of decorrelator, and that means more complex arrivals and different spatial properties that could be counted as "more information" by the brain, thus contributing in some degree to the benefit Harman described.

So the official answer is: It depends.

Yeah, that's what Ingvar Öhman (Ino audio/Guru Audio) very often starts his posts with too.

Is there such a thing as quality power? I'd rather have quality power over quantity? Isn't all power clean when used in it's performance envelope? I've never understood that either. Quantity would help with headroom, but quality? Like better quality watts ... over quantity of watts. Is that just outdated thinking or is there such a thing?

No. Some people who use this term, however, mean something real but quantitative: The ability to drive low loads. An amp rated 50 watts into 8 ohms might be able to pump out 90 watts into 4 ohms—or it might shut down altogether. IOW, "quality" is a stand-in for quantity—i.e., number of watts into lower loads.

In most cases, however, I'd guess the definition they're using is, quality watts are the watts you get from quality amps. And quality amps are the ones with the right nameplates.

I had another thought with regard to resolving these issues of differences or lack thereof in the sound of high quality electronic gear. ( What fun would resolving it be!!!!???) I harken back to the Bob Carver experiment per Sterrophile with a Conrad Johnson power amp. Mr Carver alleged he could make his new M1.0 amp (with 200 clean watts per channel) sound like anything the editors chose. They chose a very well respected Conrad Johnson tube amp that cost 5 to 10 times Mr. Carvers asking price of $400 (or something like that.) Mr. Carver accomplished this at least to the satisfaction of the head guy by using a null test (i.e. combing out of phase versions of the same signal until they zeroed out via tweeks in his amp.) He then produced this version as the M1.0t.

This test seems very powerful in retrospect and could theoretically be applied to any electronic gear. Suppose Mr. Carver would answer the challenge today with one of those $50k or more superamps and succeeded. This would be particularly interesting because it would step around the issue of whether or not the differences were audible. If the pieces of equipment could be made to null, then the original differences whether audible or not would have been due to relatively minor and inexpensive artifacts of circuitry, proving the megabuck construction to be superfluous. Or not if they couldn't be nulled. But even in this case we'd have valuable information about what the superamp does that a good cheaper amp doesn't co. Amps would be particularly appropriate for this test since there is much that can be subtly adjusted. Wires are harder but maybe something could be done. DAC's would seem highly appropriate as well.

The Carver challenge has become a sort of fish story with everyone having their own version of it . It is useful to read about it from the people who were actually there. Click on he first link in this google search: https://www.google.com/search?hl=en&tbo=d&rlz=1C1SNNT_enUS374US375&q=site%3Awhatsbestforum.com+page+12++Tube+vs+Solid+State+Is+the+War&oq=site%3Awhatsbestforum.com+page+12++Tube+vs+Solid+State+Is+the+War&gs_l=serp.3...15306.15306.0.16345.1.1.0.0.0.0.123.123.0j1.1.0.les%3B..0.0...1c.1.2.serp.N4am0zw4lv8. Then go and read John Atkinson's response a few replies down.

The link to the original article does not work. So use this one: https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&ved=0CDcQFjAA&url=http%3A%2F%2Fwww.computeraudiophile.com%2Fattachments%2Ff10-music-servers%2F531d1337842321-opus-melody-carver-stereophile-challenge.pdf&ei=cG8NUa6RCYLY2AX-94CoCA&usg=AFQjCNEJORBoleJ7jDBHcwz7T8Rml2qj5w

Net, net, it is not the type of data that can be used by either side to make an argument. To accept it you would have to accept there are audible differences between amps. And that design of the amp impacts that. And importantly, subjective sighted tests are right and can be backed by the right set of measurements and differential tests. People in the so called objectivity camp don't accept any of this. On the other side, the effort did bring different designs closer together and hence lends support to the fact as you say that the sound of other amps can potentially be duplicated. As I said, neither side wins in this regard .

There are a number of articles on my web site that explain these issues in detail:

Converter Loop-Back Tests
Artifact Audibility Report
The Truth About Record Levels
Dither Report
Converter Comparison
A common-sense explanation of audiophile beliefs
Audiophoolery
Perception - the Final Frontier

My Audio Expert book summarizes this and goes even deeper and offers even more example audio files.

--Ethan

Of course there is. If all forms of nonlinear distortion are down 100 dB or more, frequency response is +/- 0.1 dB over the audible range, and the phase response is related to frequency response by the relationship known as mnimum phase and the amp's performance is not materially degraded by driving a loudspeaker load, then you have quality power. This is by the way a very conservative standard. Performance 5 times worse is very good.

The rule about quantity is called sufficiency. If you don't have enough power then you have clipping, then all pretenses of quality are lost. If you have enough power so that clipping is avoided, then you have sufficient powr. Having more than sufficient power has no audible benefits.
True if that is how you define "performance envelope".

Headroom you never use is a waste.

I guess what made me question all of this is the assumption that not all watts are created equal. If I'm understanding you correctly you are saying that one really only needs sufficient power for their listening levels and distortion products should be below audibility. The point where the amp is still well within it's performance envelope. So quantity is quality, in other words. But there isn't really a better quality watt per se. Watts are watts, assuming I'm understanding this correctly.

That's what a salesman at a hi-fi shop that mostly carries 2ch gear will tell you. The hi-fi industry is trying to limit the damage of AVR's being readily available to the public in all the big brand chain stores. So they have created this *idea* that they are inferior for 2ch music.


What could give AVR's a bad rap is the fact they have a lot more adjustability in the terms of bass management and EQ and speaker distances and such and therefore a lot more scope to get things wrong with a bad setup. A salesman from a conventional hi-fi store would only need to hear one badly set up AVR 15 or 20 years ago and use that as an excuse to look down his nose at them for the rest of his career and tell every potential customer that walks through his door so.

Again, something a hi-fi shop salesman who is trying to move his old stock of 30 watt NAD 2ch amps will tell people when he knows that down the road at the big chain store they are selling 120 watt AVR's like hot cakes.

That they are different is not an assumption, it is a fact. I've bench tested dozens if not hundreds of receivers and power amps. They all measured different, have different schematic diagrams, different output stages, different power transformers, etc. They are not created equal and they are not all the same.

However, audiophiles and music lovers tell me they care nothing about all that. They tell me that they are worried about sound quality. I'm an audiophile and even though I understand almost all of the finest points of amplifier design, when it comes to the AVR I'm listening to I only care about what it sounds like.

So, in 1975 some friends and I devised a way to compare amplifiers where all that mattered was sound quality. We found that some amplifiers sounded different, but they also measured decidedly inferior, even with the relatively insensitive test equipment of the day. We discovered all of the science about there being thresholds of distortion and noise below which nobody heard nothing no matter what. It all fits together.

Exactly.

I would prefer to say that if the quality is right or good enough, then having the right quantity is the remaining significant parameter.

Watts are watts, given that certain reasonable purity levels are maintained. There is no audible benefit to having way too many watts.

When you say quality is right, are you referring to low THD, IMD etc? If I had enough power that it would mean distortion products are sufficiently low enough at the levels I enjoy then ..I would assume that that is the benchmark for quality power. But I'm a slow learner I could very well be misunderstanding.

Yes. BTW both IM and THD have a common cause which is nonlinearity in the amplitude domain. It is scientically called nonlinear distortion. Variations in frequency and phase response are correspondingly called linear distortion. The latter term may seem counterintuitive, but it is proper science.


Distortion and power are in general separate things. As long as you have sufficient power to avoid clipping, nonlinear distortion and power are independent of each other. You seem to keep trying to put them together and I'm getting tired of trying to come up with new ways to say the same thing.

One more time:

An amplifier can have 10 watts of output or 1,000 watts of output, and lets say 0.01% distortion and 10% distortion. All 4 permutations of these are possible. You can have a 10 watt amp with 0.01% distortion and another 10 watt amp with 10% distortion. You can have a 1,000 watt amp with 0.01% distortion and another 1,000 watt amp with 10% distortion.

Just to really confuse you, if I call the 10 watt 0.01% distortion amp a 15 watt amp, it just might have 10% distortion.

I'm sorry. I didn't mean to say that they are the same thing. Just that ... if the amp is kept well below clipping the distortion should be low enough not to be relevant. An amp with lots of headroom may be coasting along at more sensible levels, so what I tried to say and failed was that more headroom = less distortion so cleaner result.

Actually that does make sense to me. I've seen on Audioholics and other sites when amps are asked to produce more power than normal and distortion skyrockets. Thanks for all your help in this thread, arny. I am sorry for frustrating you and it was not my intention.

I think you need to re-read Arnyk's reply once more. It might seem logical to draw the conclusion you did, but it's real life merits is limited. There does not have to be a very strong correlation between headroom and low distortion in an amp. Now apply the same thing to a speaker element and you'll have a stronger correlation.

I read these and wish I could have participated at the time of this later discussion. A careful reading seems to suggest that at that time of the original test, and remember it was 1985, Bob succeeded in his null goal by adding distortion and unflattening frequency response of his m1.0 to match the CJ. John Atkinson reports that he tested production M1.0t's and got much lower nulls and that the differences occurred at the frequency extremes. Had I been participating at the time, I would have asked whether the situation could simply be explained by the fact that Bob, thinking the midrange was what he could usefully copy, was loath to make his amp roll off at the extremes and increase distortion.

Today we have high end solid state amps way in the megabuck range and very good cheaper solid state. I'd think it would be very revealing as to what the null of such would be. Each is clearly intended to be flat. For practical purposes, the choice of a tube amp doesn't reveal very useful information. We know that some people like tube sound and a likely explanation is that its the "defects" in reproduction that is attractive. Without denigrating that, I'd be interested in what differences if any would be revealed by nulling equipment with the same design goals: i.e. very low distortion and ruler flat response.

The megabuck justification is that there are differences that would generally be perceived as better and than they can only be achieved by megabuck type build. If in fact, less costly equipment can be nulled with inexpensive mods, then the megabuck hypothesis is out the window, and what we know is that such differences which may in fact be heard by some, are a matter of taste or confirmation bias (i.e. I can distinguish the megabuck sound and therefore it must be better.) Or not. In other words, IMO the most important question is whether any differences can be attributed to the high cost build.

Quote:
Originally Posted by brucest 

I had another thought with regard to resolving these issues of differences or lack thereof in the sound of high quality electronic gear. ( What fun would resolving it be!!!!???) I harken back to the Bob Carver experiment per Sterrophile with a Conrad Johnson power amp. Mr Carver alleged he could make his new M1.0 amp (with 200 clean watts per channel) sound like anything the editors chose. They chose a very well respected Conrad Johnson tube amp that cost 5 to 10 times Mr. Carvers asking price of $400 (or something like that.) Mr. Carver accomplished this at least to the satisfaction of the head guy by using a null test (i.e. combing out of phase versions of the same signal until they zeroed out via tweeks in his amp.) He then produced this version as the M1.0t.

This test seems very powerful in retrospect and could theoretically be applied to any electronic gear. Suppose Mr. Carver would answer the challenge today with one of those $50k or more superamps and succeeded. This would be particularly interesting because it would step around the issue of whether or not the differences were audible. If the pieces of equipment could be made to null, then the original differences whether audible or not would have been due to relatively minor and inexpensive artifacts of circuitry, proving the megabuck construction to be superfluous. Or not if they couldn't be nulled. But even in this case we'd have valuable information about what the superamp does that a good cheaper amp doesn't co. Amps would be particularly appropriate for this test since there is much that can be subtly adjusted. Wires are harder but maybe something could be done. DAC's would seem highly appropriate as well.

Thanks, I read these and wish I could have participated at the time of this later discussion. A careful reading seems to suggest that at that time of the original test, and remember it was 1985, Bob succeeded in his null goal by adding distortion and unflattening frequency response of his m1.0 to match the CJ. John Atkinson reports that he tested production M1.0t's and got much lower nulls and that the differences occurred at the frequency extremes. Had I been participating at the time, I would have asked whether the situation could simply be explained by the fact that Bob, thinking the midrange was what he could usefully copy, was loath to make his amp roll off at the extremes and increase distortion.

Today we have high end solid state amps way in the megabuck range and very good cheaper solid state. I'd think it would be very revealing as to what the null of such would be. Each is clearly intended to be flat. For practical purposes, the choice of a tube amp doesn't reveal very useful information. We know that some people like tube sound and a likely explanation is that its the "defects" in reproduction that is attractive. Without denigrating that, I'd be interested in what differences if any would be revealed by nulling equipment with the same design goals: i.e. very low distortion and ruler flat response.

The megabuck justification is that there are differences that would generally be perceived as better and than they can only be achieved by megabuck type build. If in fact, less costly equipment can be nulled with inexpensive mods, then the megabuck hypothesis is out the window, and what we know is that such differences which may in fact be heard by some, are a matter of taste or confirmation bias (i.e. I can distinguish the megabuck sound and therefore it must be better.) Or not. In other words, IMO the most important question is whether any differences can be attributed to the high cost build.

Not only that even reasonably priced SS amps can be very linear, have a lot of reactive load handling power, and work well with very low impedance loads.

You don't have to do null testing, all you need to do is make proper use of modern test equpment.


Hats off to John Atkinson as a leading collector of information relating to this by the means I described above. IME the most audible undesirable property of tubed amps is their relatively high source impedance.


Furthermore such differences, if they existed would have to be unmeasurable using modern test equipment, including nulling. It is well known that all instances of non-measurable "audible" differences are vaporized by DBTs. They were artifacts of sighted evaluations all along.

I experienced one of the best examples of differences due to a high cost build that I ever have when I had a Threshold SA 4E in my possession to test a couple of years back. The amp was far better than it seemed in basic testing if you did some non-traditional but fairly conventional testing. When operating at low frequencies (< 100 Hz) and with low impedance loads (< 4 ohms) it would put out as much power as you could suck power from the wall to support. At 1 KHz and an 8 ohm load it looked like a fairly conventional 125 wpc power amp. Its output under non-traditional conditions was like an order of magnitude greater.

Pick the right Martin Logan speaker or a certain legacy Infiinity speaker and this puppy would probably be the death of a great many very fine modern power amps. With normal (well-designed) speakers you would be looking at a null test. What matters to you?
Edited by arnyk - 2/5/13 at 7:31am

I didn't say any of the stuff above quoted under my name.

How did the quote business go wrong.

If you didn't do post 1273 who did?