If I remember correctly , this is a similar argument Jezz Arkeless advanced for the cambridge 640p , that the it takes 12v AC so it matters not you get yourself a fancy regulated linear supply cause the unit itself turns Ac to dc .

No John

I am making no comment about the power supplies James and Martin use.  I don't know them.  I am distinguishing between the fact that the way AC is turned into DC can have a big impact on sound - whereas what you do with the AC does not have any impact on the sound.

A device needs DC to run.  Ideally that should be ruler flat.  If the source of the DC is mains, not a battery, then a process is needed to convert a wiggly sinewave into a ruler flat DC.  That conversion is critical.  How hairy, wiggly, or otherwise that AC is of no consequence - because the power supplies job is to flatten it.

James has pointed out that statement is definitive and contradicts the experience of others.  But those others experience it themselves, and are unable to demonstrate satisfactorily that the phenomenon actually exists.  They are entitled to those experiences - but I am unable to share them, and science is unable to explain them.

I think noise on the mains is all relative to where you live?
There is also another argument that the national grid also acts as a big antenna for RF noise.

A mains regeneration unit or BPS will be the last thing I add to my system. I would like to hear one to see if it made any difference, but that's not likely to happen soon.

With the greatest possible respect, if that statement is true you can readily measure it. No subjectivism needed. Dc is flat, or it isn't. If you are right, it can be measured. 

Martin - that is typical of the blurring of subjectivism and objectivism which I find so frustrating.  The statement is not untrue.  The conclusions you draw from it are.

Noise left on the DC can be heard as well as measured.  That bad DC affects the sound isn't at all disputed by me.  That bad AC makes bad DC is disputed.

So - to clarify, are you saying we can take a good power supply, whether SMPS like the one in your Chord, or the one in Owens ABH2, or a good linear design as made by uncle tom cobbly and all, and measure the difference in the dc output that is caused by changing a kettle lead, or a gold plated fuse, or even using a balanced mains supply?

Because if we can measure the change in the DC that results from that mains foo, there is absolutely no dispute between us.  IF you could perform that measurement, you make the unimpeachable case that Mains affects the DC which nobody disputes affects the sound.

But can you do that measurement? 

Oh , and while we're at it - yes DQ, you are almost certainly right about Jez.  I don't know the detail of how his designs work, but I understand the principle of how features integrated into the design of an amplifier, inherently reject power supply noise, and form part of the power supply itself.

You might think that would make it hard to measure .  You can't just measure the output of a Pip power supply itself to measure how much noise on the input remains on the output.  But that is not true.  That is why, ages back on this forum, I pointed out that the correct place to measure power supply noise in an amplifier is on the signal output of the amplifier.  It's what Owen measures all the time when designing amplifiers, across frequencies extending into megahz, because of the affect that noise there has on audio frequencies too.  Noise on a power supply is - well, NOISE.  The measure is signal to noise ratio, and noise floor. It may not be the only important factor in a design, nor tell you how a design will sound, but it sure as hell measures noise!

(NB - regarding measuring noise on the output, and balanced power supplies, be aware that because of the way that a BPS can affect the noise on the ground which is common to both the supply and the signal, then it can have an impact on the sound quality.  This is one of the main reasons they are used in a professional environment, although as the much published Sound on Sound article points out, there are better ways to skin that particular cat)

My position on this is I do not know cause I have not tried. However pinkie's logic is flawed cause he assummes all that can be heard can be measured . I consider that to be  crap .

And to better explain what issues are involved I will nick 2 posts of Jez Arkless from PFM -
----

This is a vast subject but first of all I'll say I'm more of a subjectivist on this subject than many of the other engineers who post here. I don't believe it is possible to measure and account for all of the subjective phenomenon we observe from amplifiers (specifically power amps. Line amps are difficult to get all that wrong... although some manage it).

I have built prototypes whose measured performance was impeccable (THD too low to measure with equipment at hand at the time at below 0.001%, FR DC - 100KHz+, damping factor 500+ etc) and yet it's sounded below par...
Some parameters are particularly difficult to evaluate with steady state sine wave testing... things like memory distortion, temperature compensation errors causing program dependent under or over biasing (virtually guaranteed to be present to some extent in all class AB bipolar transistor amps), capacitor dielectric absorption effects, non symmetrical program material (yes there can be "DC" present in the music!) effects on the amp etc etc...

Even for amps with seemingly very similar topologies there will be vast differences when you analyse in more detail. Different active devices used in the same places will often have very different Early voltage, transconductance and inter junction capacitances (the last two will change with signal level and DC supply rail voltages, just to throw a few more variables in).

Many things assumed to be critical by laymen are often less important than believed, such as power supplies (this is a big subject in itself and yes they are very important but what an engineer thinks of as the power supply is often quite different from the layman and the areas which make the big differences are often not in the areas that folks suspect )
People obsess about current delivery for example and yes, if you have low impedance and reactive speakers and like to listen bloody loud then it is indeed that important... but for the average speaker and at the average listening levels its importance is grossly exaggerated.... obviously if one insists that the amp should also be able to drive the load from hell, just in case you buy such a speaker, then the amp is going to be vastly bigger, heavier and more expensive than is required for the average speaker, and with no benefit.

I could go on all day but will spare you all...

[2] Not inexplicable but it might as well be in practice because there are just so many factors interacting that full analysis becomes virtually impossible. Consider a large signal moving through a power amp. Device parasitics, non linear modulation of internal capacitances and Early Effect mean that the amplification factor and frequency response of individual stages (esp. the voltage amp) will actually change over one cycle of a sine wave... this will also modulate the phase of the signal (one form of frequency modulation relies on exploiting this effect!) add some dielectric absorption from some caps and a bit of "ground bounce" to the recipe and mix with some program dependent thermal effects with a relatively long time constant.... next feed this back through the amp (negative feedback) with a phase delay and have all the above applied again but to a different harmonically structured version (Scroggie I believe beat Baxandall on bringing this to public attention) and like the film says ANALYSE THIS!
What did Rumsfeld say about known unknowns...

----

This is not to give carte blanche to any subjective claim , it is okay to say I have difficulty believing that is anything but subjective phenomena for you in your room with your hifi with little or no universal applicability , what I balk at are accountants making definitive absolute statements upon the curious and wonderful world of hifi reproduction .

Bollocks.  Surely in a profession such as the one to which you claim to belong  they teach "RTFQ"

I said "{a measurement} might NOT tell you how it will sound".  That surely does NOT assume all that can be heard can be measured.  Quite the opposite.

What I said was, if the discussion is about noise - that can be measured.  The measurement won't tell you the amplifier sounds good, but it will tell you whether noise has been added to the signal.

I might add, that I missed my usual caveat "within sensible limits".  So plugging a 110v device into 240v may well fry it.  Huge spikes on the mains (like a lightening strike) may be audible.  But within tolerances, outside fault conditions, nobody had demonstrated that changing the mains you feed a good power supply with, changes the DC it produces.

I sometimes wonder does how we think effect to how we listen.