Big Sound 2015: Pascal's Wager Seems to be Paying Off

Blaise Pascal was a pretty scientific dude. From his Wiki page:

[He] made important contributions to the study of fluids, and clarified the concepts of pressure and vacuum by generalizing the work of Evangelista Torricelli. Pascal also wrote in defense of the scientific method.

In 1642, while still a teenager, he started some pioneering work on calculating machines. After three years of effort and fifty prototypes,[2] he built 20 finished machines (called Pascal's calculators and later Pascalines) over the following ten years,[3] establishing him as one of the first two inventors of the mechanical calculator.[4][5]

Pascal was an important mathematician, helping create two major new areas of research: he wrote a significant treatise on the subject of projective geometry at the age of 16, and later corresponded with Pierre de Fermat on probability theory, strongly influencing the development of modern economics and social science.

And yet, when it came to God he famously came up with his idea now known as "Pascal's Wager." From the Pascal's Wager wiki page:

[Pascal's Wager] posits that humans all bet with their lives either that God exists or not. Given the possibility that God actually does exist and assuming an infinite gain or loss associated with belief or unbelief in said God (as represented by an eternity in heaven or hell), a rational person should live as though God exists and seek to believe in God. If God does not actually exist, such a person will have only a finite loss (some pleasures, luxury, etc.).[2]

I reckon cables and burn-in are kind of like that. I don't regularly hear any significant difference with these types of thing, but I do sometimes hear stuff that seems to be a slight improvement. So, as long as I don't spend extravagantly, I think it's reasonable to make the bet that an accumulation of small improvements in these areas will pay off with a better listening experience in the long run.

The problem, of course, is where to draw the line. For example, this is ridiculous:

A $5000 plug strip!? C'mon. He makes an argument in the video that the unit provides a common ground, but you don't need to spend $5000 to do that. I shan't dwell on it further.

On the other hand, there are things about shitty power that can get in the way of great audio reproduction—noise; voltage sags; grounding issues; etc. Cleaning that kind of stuff up may indeed improve listening. So, when I started putting together this idea for Big Sound 2015 I decided to give the folks at PS Audio a call and get some power conditioning from them. Well, a lot more than power conditioning, really. Power regeneration, in fact.


PS Audio sent me two P10 Power Plants ($4999 ea., a dollar less than that silly marble plug-strip above) to include in my experiment. Now, I wasn't really expecting much; I just unboxed them (no small task at 73 lbs.); set them up in the room; and went about my business of getting things hooked up and running. It was sometime the next day as I was listening that I said to myself, "Damn, this system is sounding really good." I had been struggling a bit with volume matching the various combinations of headphone and amps (more on my solution for this Wednesday), so I kind of had my head buried in the details and seem to have missed the big picture improvement. But the more I listened, the more I became convinced that the power regenerators were doing good things for the overall sound I was hearing.

Basically, the P10 Power Plant is a giant 60 Hz signal generator. It takes power from the wall; turns it into DC; generates a clean 60 Hz waveform synced with the wall AC; and then runs it through a giant amplifier to power the sockets. The important factors here are:

  • The AC is completely separate from the wall AC so it has none of the noise or voltage sag of the incoming AC lines.
  • The regenerated AC sine wave has none of the distortions (flat topping, harmonics) of the normal AC line that can introduce noise in power supplies.
  • The power amplifier has an extremely low output impedance. At 0.015 Ohms, PS Audio claims it's 100 times lower than your normal wall outlet. I did a bit of Googling and it looks to me that most mains seem to hoover around 0.15 Ohm (source, source, source), so 10 times better may be a more likely number. (By the way, read through those sources and you'll become a bit more concerned with the quality of AC power you're getting from your wall sockets.)


The performance logging page for one of the P10 Power Plants he shows me I'm drawing around 230 Watts of the 1400 Watts the P10 is capable of, and that my wall voltage varies between 113VAC and 121VAC.

In addition to providing clean power, the P10 also has some cool features. If you hook it up to your home network, you can use an internet tool through the PS Audio website to monitor power usage, and incoming and out-going line voltage and distortion. Through this app you can also program system turn-on and turn-off sequences, and schedule system up and down times. Go to the P10 Power Plant page for more info.

I dont think you've heard the last from me on this subject. I'm going to have to break out some big plug-strips and revert to wall power to investigate how big a difference I'm hearing. There's some issues about gear warming up and the long delay in the time it takes to switch over, but I've got to give it a try as it seems the difference is significant.

At any rate, if you asked me now, I'd say the PS Audio Power Plants payed off in my Pascal's Wager on AC power cleanliness. I'm digging what I'm hearing.

xnor's picture

is one of the worst arguments ever. It already assumes that there is only one god (based on one religion), who conveniently happens to be the one you already believe in and this god has to punish people.
Besides, it can easily be reversed.

Applying this bet to audio equipment is not without its problems either.
More expensive components do not need to be better. That's the problem with betting: you can and will lose.
Expensive stuff can be a complete waste of money - money that could be spent on more fruitful things.
Weird or really expensive stuff (like in the video) will make you look like a madman. :D
The belief that everything makes a difference will lead you on a long journey of worrying about upgrades instead of enjoying music free from worry.

Tyll Hertsens's picture
I essentially made the bet that putting together a system with good (but not outrageous) cabling, a solid power supply; high-performance amps; and the current crop of world's best headphones would make for a system that sounds both really good and will let me fairly report and the experience, both good and not so good, of these headphones paired with particular amps.

So far, it seems like it's delivering sound from this gear as good or better than I've ever heard before in my home. Frankly, I'm stoked. I am enjoying the listening very much, thank you. Whether this system is going to make it easier or harder to tell differences is another story. Things do sound very good, and that can make subjective evaluation difficult.

Point is, the bet on really good AC power seems to be paying off.

Oh, and Pascal's actual wager...I don't buy it. I think there are ways to somewhat determine the nature of consciousness in the universe so as to answer sufficiently the question of God and your relationship with it...but that's for a different place and time.

xnor's picture

I don't see the bet here. Even if the technical performance of these new shiny cables and amp and DAC was identical I would expect you to enjoy them more - that appears to be just simple psychology.
That's also a reason why people love the objective side of your reviews.

If you need to clean up mains to significantly improve the performance of an amp, then maybe it's not such a good amp after all, right?
This significant difference should also be measurable - and hey, I wouldn't even be surprised to see such differences in certain products.
I'm open to the possibility of differences, including so significant ones that they cause audible differences, but I can't just accept "sounds better to me now". Btw, did you see my comment about Bayes' at the Toole article?

(Bad) theology indeed doesn't fit here.

Limp's picture

When a sommelier starts to talk about the terroir of various kinds of bottled water, do you tend to trust his expertise on wine less, or more?

sszorin's picture

Quote : "...It already assumes that there is only one god (based on one religion), who conveniently happens to be the one you already believe in..."
The assumption is correct as it was demonstrated by the Greek philosophers, some 2400 years ago, that either there is no God or there is only one God. The 'cause' of the Cosmos cannot be divided.

Quote : "Besides, it can easily be reversed"
You mean the pleasures of catching venereal diseases through "free sex" outweigh experiencing 'beatific vision of Heaven' ?

Let's forget Pascal for a moment - what kind of odds would the betting men of Las Vegas offer on a proposition to build an airplane by piling up a giant mound of metal ores, minerals and various chemicals and then waiting for succeeding hurricanes to thoroughly mix this pile so that, progreassively, a fully functioning airplane would materialize ?

Jim Tavegia's picture

Why assume that it is God that punishes people. Are there not two forces, "possibly" at play? There are things in life that we think are important, that really are not.

Dreyka's picture

Damn. I was betting the next article would be about Neil Young's recent statements. This is however more amusing.

I guess you are able to perform measurements with and without the signal generator.

SleepModezZ's picture

Were the previous amp measurements done with the Power Plant of without it? Will you produce two sets of measurements so that we can see what difference it makes, i.e. what kind of "dirt" AC power directly from the wall outlet is introducing to the amp's signal. There have to be a difference, and at an audible level, I presume?

SleepModezZ's picture

The idea of the wager is an insignificant investment with the possibility of abundant fruitage. $4999 is not insignificant for most people. It might be more than the total amount I have used for audio related equipment. Was the result a real jackpot then? No, not really. It was something Tyll even failed to notice for a time but was then convinced that it made a difference. It is still possible that Tyll will re-evaluate his findings after making some measurements.

If AC power is a significant problem, the cheap solution would then be to use portable gear. Is AC->DC->AC->DC really necessary when you could go for DC only solution?

AstralStorm's picture

This solution won't work, because portable gear very often lacks sufficient internal decoupling even if the battery supply is clean.

Come to think of it, the whole reason for decoupling is 1) providing for peak current requirements 2) cleaning up power sources.

That and portable gear has components crammed close together enough so that they can interfere with each other. Distance is often a good shield, especially between digital and analog parts of the circuit (e.g. between DAC and buffer/amplifier)

True solution would be to instead run everything off a great online (AC into DC battery into either DC or again clean AC) UPS. In fact, suitable UPSes are way cheaper than this snake oil and for $5k you can have one that can run for many hours disconnected.
Alternatively, buy a laboratory grade power supply for $300. That's how much they actually cost, not $5000 snake oil.

SleepModezZ's picture

Ok, if portable amps do have to compromise because of their size, one solution would be to do a bigger amp but make it battery powered. Actually, if I recall correctly, the Objective2 is running on two rechargeable 9V batteries. If the O2 makes compromises because of its size, it would be possible to make a bigger design but keep it battery powered.

Am I right, that for a O2 the Power Plant (or UPS) would be useless? If the amplifiers that Tyll is testing are benefiting from the Power Plant, it would be more logical to blame the amps than to praise the Power Plant. Like xnor wrote above: "If you need to clean up mains to significantly improve the performance of an amp, then maybe it's not such a good amp after all, right?"

AstralStorm's picture

Actually O2 has plenty of decoupling - a necessity for working with batteries. PSRR is pretty high too.

Its troubles stem from the opamp and not the rest of the circuit. The standard tests are completely blind to it - and I bet a good blind test would catch the issues, but nobody bothered, as it's very hard to set up for a $100 piece of gear. This is the essence of the O2 challenge problem - nobody cares enough.

Tyll actually skipped the PSRR tests here by using some fancy shmancy power conditioning hardware. This might make some amplifiers look better than in reality.

xnor's picture

What are you talking about?
What measurement do you suggest that is not blind to "it"?

xnor's picture

This solution won't work, because portable gear very often lacks sufficient internal decoupling even if the battery supply is clean.

What? You mean the reservoir caps are too small?
We can see how well the power supply can keep up by using a worst case low crest factor test signal (e.g. a sine).

As for desktop gear: you shouldn't need a UPS for that. There should be no excuse for lack of a clean power supply.

AstralStorm's picture

Yes, I mean the reservoir caps are often too small for harder loads.
Usually whenever large amounts of current are required - for planars - or voltage swing such as with high impedance headphones. First taxes output buffer bypass, second taxes gain stage bypass.

Which case fails depends on the design of the amplifier - and how gracefully it clips.

THD+N would catch it, if only Tyll didn't use supreme quality power conditioning which can supply anything clean with very fast slew rate - P10 has considerable output capacitance, making amplifier supply bypass much less critical.

I'd recommend actually testing the amplifiers on real power line input and not supremely conditioned power.
Clean power supply is not a given in audio - a good amplifier should have high PSRR, easily in excess of 80 dB, making power supply side noise moot.
Plus it should have more than sufficient decoupling/reservoir to handle loads it is rated for.

Both are generally skimped on in portable amps where you cannot install large filters and large capacitors.

xnor's picture

You were talking especially about battery-powered headphone amps so THD+N would certainly show it.
I don't see the problem here if the portable amp delivers enough power (below clipping) for your headphones.

For desktop amps I don't see how a conditioner would make big differences regarding reservoir caps either.

sszorin's picture

Reading this makes me think of one smart Alec who argued with me that printed circuits of an iPhone can do as good job of filtering THD + Noise as dedicated components of an amplifier.

ab_ba's picture

My favorite quote from the video. Yes indeed. Did Rob Reiner direct it? I think he thinks electrons are like little trains running along tracks inside those wires. Yep, for mecahnical insulation. That's why we use rubber. I think I saw one used on eBay for $1. $500 for shipping though.

AstralStorm's picture

You could use a clamp for any large cap - with an obligatory rubber boot. It costs $1 or $2 a piece.

However, in almost all designs, the capacitors are either large enough that microphonics are very low frequency and mass dampened or mounted stiff enough using SMD/SMT techniques so that they are unimportant because the whole board cannot vibrate enough.
(And who installs an audio amplifier on a washing machine or such?)

It's not like those are phono cartridges - and even those should be already mounted in a way resistant to outside vibrations.

Long time listener's picture

I still use the old PS Audio ultimate outlets with their "bolun" inside to reduce noise. Even that relatively small amount of noise reduction makes a clearly audible difference. I also used to use one of their power regenerators, but gave up on it because it was undependable and kept breaking down. Surprised you've never investigated this side of things before.

Anxiously awaiting Part II of the portable source overview.

SleepModezZ's picture

According to PS Audio, even the Power Plant can benefit from having an Ultimate Outlet before it, doing som 'pre-cleaning' of the AC. In Inner Fidelity's sister publication, Stereophile, Robert Deutsch reports: "Well, I tried this combination, and I have to agree with PS Audio's recommendation. The difference was not night-and-day, but the sound became a bit more open and more dynamic, and there was some reduction of upper-frequency grain. The results were better with the High Current UO than with the Standard." Maybe there could be some benefits of chaining several Power Plants and Ultimate Outlets because clearly just one can't clean the AC wholly? With only one Power Plant, there is probably still quite a bit "dirt" coming through and sullying the perfect audio ecstasy.

Long time listener's picture

I think PS Audio was just trying to sell more products. And if their power regeneration is doing its job, just one power plant should be enough. My Power Plant Premier, while it was working (a very short time) was fabulous.

tony's picture

Our GM Technical Center has plenty of these AC Power devices.

Utility Company Power is too dirty and noisy for consistent research work ( I'm told ) , even the frequency varies ( they create their own 60 hz. ( bet-cha you won't go that far ).

I'm happy you are going to this extent ( as much as it's a major pain and expense to set-up a system like this ), I'm surprised you have the budget, courage and inclinations. This is high level research engineering territory.

I've always felt that my little home systems sonic performance would go up and down, late night listening being the best. I know that when my neighbor runs her Garbage Disposal I could hear it in my system ( or at least I think I could ). However ( I have to add ) my Schiit Asgard 2 doesn't seem to suffer with this problem ( Jason Stoddard claims great power supply design, I probably should believe him ).

In the old days of noisy vinyl and tubes; 120VAC issues were probably masked and well below our equipment's capabilities.

Nowadays people are listening to "Quiet" being shades of "Black", the quoted spec.s on noise being down by huge amounts of db. If I hear any noise on a piece of music I immediately know it's a ADD not a DDD recorded piece.

So, in this Age of Digital, noise floor levels impact our perception.

You Gotta get clean Power!, even if the cost is prohibitive.

I applaud you,

All the rest of us need to embrace this new Century with it's advancing technical requirements and their decreasing Costs.

On the other hand:

We can just buy a Roy Hall turntable, an old Conrad-Johnson Tube system and a box of Garage Sale vinyl Albums.

We can embrace the noise as part of the experience, as we used to do.

Sometimes I kinda miss the musty smell of cardboard Record Sleeves, might be nostalgia. ( and the Nitty Gritty machine, geez I'm old )

Tony in Michigan

Dreyka's picture

I'm happy you are going to this extent ( as much as it's a major pain and expense to set-up a system like this ),

I'm sure the manufacturers were falling over themselves to give Tyll their stuff for free actually because it's free advertising for the 'ultimate' rig. A good word and you'll already be selling more units.

You Gotta get clean Power!, even if the cost is prohibitive.

It's not even known whether the differences are measurable yet alone audible. It's easy enough to get a silent noise floor anyway without $5000 boxes.

tony's picture

I think the difference is measurable, the evidence abounds.

What isn't clear is whether us consumers can realize a cost effective benefit. ( I'm not quite bothered enough and my iMac doesn't seem troubled by my Mains Electricity Quality ) .

About 50% of my gear is powered by rechargeable Battery packs with the percentage on the increase. So I'm kinda using my own "Clean" Mains supply.

But :

If you're doing a research project, you're mandated to eliminate as many variables as possible. So Tyll must clean up his power if he can. And, as you suggest, the manufacturers will loan him "their" solutions. Sure they would like "honorable" mentions about their support, I suppose that would go without saying.

Scientifically ( I think ) Tyll is on the right track, I'm impressed by his doing all this work. I imaging the Manufacturers people responsible for Advertising are also impressed.

What we are seeing here is Data Integrety capabilities.

I love it.

Tony in Michigan

Theriverlethe's picture

Assuming this is actually an issue, isn't power supply design and filtering supposed to be part of the quality of an amp? Perhaps you're giving poorly designed amplifiers and unfair advantage.

Bill Leebens's picture

By way of full disclosure: I'm Director of Marketing at PS Audio, and provided the Power Plants for the shoot-out.

Tyll, I'm gonna sidestep the arguments regarding Pascal...I'm pleased but not surprised by the results you're seeing. We're so used to the noise and junk on our powerlines that to have it disappear can be pretty shocking. --No pun intended.

SleepModeZ: We don't recommend daisy-chaining Power Plants; there might be some improvement in doing so, but not as much as going from no Power Plant to 1 Power Plant. The law of diminishing returns would apply. And a Dectet in series with a Power Plant would produce more obvious results, at a much lower price.

Long Time: Of COURSE we're trying to sell more products--what company ISN'T? But that doesn't mean there isn't truth in what we say. Besides, I just said we don't endorse daisy-chaining Power Plants. Give me some points for THAT. Finally--sorry for the issues with your Premier. We've come a long way since then.

Tony in Michigan: Holler if we can help your colleagues. I worked in racing for many years, and knew a bunch of good guys at the Tech Center.

Theriverlethe: Do you remember what you post? Sorry. To your point, a robust power supply is vital in amp design, but any amp will work harder and sound worse if the AC going into it is noisy and overrun by spuriae.
And I'm sure Tyll is thrilled that you just pointed out yet another possible variable...

Happy listening!

Cheers, Bill

xnor's picture

To your point, a robust power supply is vital in amp design, but any amp will work harder and sound worse if the AC going into it is noisy and overrun by spuriae.

I don't see how you could assert that.

Bill Leebens's picture

Any amp fed with dirty AC containing significant noise and wideband harmonics will cause the power supply to feed trash into the signal, likely causing the amp to operate outside the bandwidth it's designed to handle. It may also cause the amp to draw more current that its power supply is designed for. In addition, dirty power is often associated with low or high voltage, either of which will produce sub-optimal performance.

Your statement implies that quality of AC going into an amp doesn't matter. I don't see how you could assert THAT.

xnor's picture

You said that any amp will work harder and sound worse when the AC is noisy. I doubt you have tested all amplifiers... Have you published measurements and listening test results for even just a single typical (headphone/power) amp?

How much noise passes through a typical PSU transformer alone? How much is left after filtering inside the PSU?

The power supply's job is not to feed trash into the signal. It's job is to provide clean enough power to the amplifier section -- it's not like PSU design is magic.
And the amplifier section has additional PSRR anyway.

Your statement implies that quality of AC going into an amp doesn't matter. I don't see how you could assert THAT.

No, that's a straw man.
If you made the assertion that "all swans are black" and I say "no, you can't assert that", then does that mean I say "all swans are white"?