2. speakers
  3. 1217550410

The Richard Clark "all amps sounds the same" test


Okay, I know there has been tons of commentary on this issue, but I have a specific question. And it will make it clear why I'm posting this amp question in the speaker threads.

I'm curious if anyone knows if Maggie 20.1s or something equally hard to drive and equally transparent has been tested? I know planars have been used on his test, but I don't know any details.

Oh - for those who don't have any idea what I'm talking about see:

http://www.tom-morrow-land.com/tests/ampchall/rcrules.htm
and
http://www.tom-morrow-land.com/tests/ampchall/index.htm

and if you google it there is almost endless commentary on it.

Okay - but I want to test the following:

Magnepan 20.1s
Pass Labs X350.5 or XA160.5
Pass Labs Pre? (Don't care as much)
EmmLabs CD Player

Then, we need a low-cost amp. Now, the trouble is, he has a reasonable request in his test, each amp has to be used within its thresholds, so no using them at 300 watts when one is rated at 30 watts. Obviously with one clipping and the other one not clipping you will hear a difference.

This also applies to a 4-ohm speaker. So, assuming someone hasn't done an extremely similar test and can just tell us the difference, the next question is what is the worst amplifier that is rated at 4 ohms? While the X350.5 is high-power, the test could be done at 85 db, so you don't need too many watts to make that work.

This would effectively answer all the "maggies need high power to by dynamic" and lots of other similar questions. Because the test is at one db level, does one amp really push more bass out of them than another?

Hey - actually wouldn't Tympani IVs be harder to drive? Maybe we should use Tympani's :).

What do people think, is this issue still alive or has someone resolved these issues? I have to think I could hear the difference and may have my wife run some singly-blind tests for me - I don't have any of the equipment above, but do have 3.6s and an Aleph 5. See how that Aleph sounds compared to some sort of $100 amp rated at 4 ohms.

Might there be a 4-ohm rated amp in a boombox or bookshelf system? I'll poke around. Sure a single op-amp chip in a bookshelf system (often what $100 system amps consist of, just a few chips) would sound worse than a Pass Labs Aleph, which Stereophile said compared to the Levinson 300 lb amps?

Oh - and the essence of my idea with this test is that perhaps the sound is 'more similar' on speakers that are easier to drive, but with 20.1s - and this is just as important - with a highly resolving ribbon speaker - the difference might become more apparent.

Oh - also, I'm not sure if he allows me to choose the music, but I have found over the years certain parts of certain passages that show the differences of components more than others. I think that would also be important - what passages are played, as on some I would believe the differences would be impossible/difficult to detect.

If I'm just repeating stuff that can be found elsewhere let me know... Just seems like we should be able to bust this test.
lightminer
shadorne's avatar
Ag insider logo xs@2x
shadorne
7,467 posts
 
My guess is all these tests were done transistor vs transistor- I'm pretty sure I'd be hard pressed to accurately tell the difference.

I agree - I would suspect the majority of high end car amplifiers are soild state - I would imagine microphonics from vibration might pose a problem with a tube application in that context.

Indeed I was careful to mention "solid state" in my previous remark, as I think you can hear a tube versus SS more easily, especially those with an output transformer - so it would be incorrect to say everything sounds the same. You would indeed be more likely to find slight variations in tube amps due to the voltage paradigm and choice of output transformer (as well as topologies). However, frankly speaking, the test (as described) seems to try to minimize all these aspects too. In short, by limiting amplifier operating range to extremely modest levels (which avoids where most large differences might show up), careful volume level matching, and by trying to compensate for the voltage paradigm through the addition of resistors - it is hardly suprising that remaining differences would pose a challenge to even a discerning listener to score so highly (as Richard requires) when listening to music.
mrtennis
3,295 posts
 
hi drubin:

tou posed a (hypothetical) challenge on 8/01/08. i apologize for not answering:

here is my answer;

you can not prove the null hypothesis. this means that you cannot prove the hypothesis :

all amplifiers sound the same

ignoring the practicality of all combinations of pairs of amplifiers, induction is not definitive or absolute proof.

the exception can disprove the rule.

after doing n blind tests and, say, hearing no differences between two amplifiers, you may hear a difference in test n+1.

lightminer:

what i have stated attests to the futility of testing amplifiers. as many amps as you test, it may be possible to hear differences between a pair of amps, yet to be tested.
jult52
155 posts
 
I just added a B&K power amp to a Rotel integrated (so the Rotel went from functioning as an integrated to just as a pre-) and I found differences in key areas, using the same CDs at the same volume (I am careful about listening to music at pretty much only two gain levels). Many positive changes but one I found was that the Rotel integrated set-up actually produced a smoother midrange string sound. Post the change-up, the string sound was "coarser", without any changes to volume, range and dynamic peaks. I am very suspicious about Clark's test, even accepting his parameters, given my experience.

BTW, the Shostakovich 8th describes the awful siege of Leningrad, not the battle of Stalingrad.
rcprince
3,059 posts
 
Jult52, actually it's the Shostakovich 7th that is the Lenningrad Symphony. The 8th as I recall was generally about the horror of that war; the government called it the Stalingrad Symphony, but I don't think Shostakovich had any one battle of the war in mind.

As far as the thread goes, I'm not sure given Clark's requirements I could detect a difference, certainly not 24 out of 24.
rpfef
110 posts
 
Personally, I would love to be able to take part in a fair double blind comparison. I do believe double blind tests are by far the best way to get at the truth of our own perceptions.
But to be effective in approaching that truth, I would want the comparison to be between amps I am already familiar with and which I already believe sound different. What I want to find out is if my perceptions are to be trusted.
Even though I hear, for instance, my Plinius amp as sounding identifiably different from, for instance, the Bedini of identical power it replaced, I am never certain that the differences I believe I hear are not mental constructs of the non-blinded situation combined with my own mental processes.
The program material must be familiar, also. All this is because the hypothesis being tested is: The Plinius 100 sounds identifiably unlike the Bedini Classic 100. That is the premise on which I am basing my purchase choices.
As a sworn skeptic, I want some better evidence of these phenomena than personal testimony or my own perceptions.

Using familiar gear, rather than strangers, means I have a ready yardstick by which to measure my results. I don't care what Richard Clark thinks. I just want to come closer to knowing whether I am fooling myself. After all, a lot of money is at stake: the thousands I spend on "better" equipment.

Setting up a double blind test in one's own room would be the ideal for this kind of experiment. Now that i have thought of this, I will perhaps try somehow to set one up. Obviously, if I do not score well, I will feel the test has exposed my prior self-deception. But, if I do well, I really will not be sure of anything because a really solid double blind test is pretty much impossible in my house.
Program material, also, would need to be familiar, for the same reasons.
More to discover