Some mysteries

If RF interference is a problem (I imagine it could be with some low level signals) I'm sure there are techniques to suppress it that should be part of the circuitry of the player itself. It seems rather less likely that RF is going to add a signal to mains voltage.

But even if it does, it seems to me that all those miles of wire dangling from pylons and bathed in the glow of mobile phone radiation, TV signals, radio broadcasts and serving millions of people their electricity (including industrial users with big motors to power) and passing through busbars, transformers, multiple fuses and hundreds of less than ideal junctions and joints are something to worry a great deal more about than the last few feet.

Not to forget that that lovely expensive custom cable is going to be screwed to a terminal which is wired to a PCB with the cheapest wire the manufacturer could find and then routed as signals along a minuscule thin copper pcb track radiating merrily to its neighbours before supplying cuurent to a bunch of components whose skinny little legs are bits of tin that a paper clip would be embarrassed to be made of... 

Even ignoring the lack of convincing hard evidence in support of this hypothesis, there is a  lack of any internal logic to the argument - which means everyone should be sceptical and should be looking for other explanations for any perceived impact on the sound.

Oberserver expectency is a powerful placebo effect. Studies on cable "sound" have revealed some interesting facts.  Double blind tests of speaker cables have been done where the participants were tricked. They were told they hearing two cables but in fact the cables were not swapped at all. But typically, users scores indicate that in between 60-62% of the tests, participants report hearing differences between cables even when they were listening to exactly the same cable. The sheer effort of listening allows the imagination to create differences that don't exists.  Properly conducted tests need to callibrate for this effect to avoid mis-leading false positives.

I know - I'm the pround owner of a pair of Linn interconnects that cost me 10x what I would normally pay for a bit of wire.  I got them from Grahams Hifi after 2 hours of careful evaluation. Of course they don't make the slightest difference to the sound but I would swear they did during the demo. 

ServerBaboon wrote: Last few metres I would have thought that the way that the last few few metres of cable can make a difference is the amount RF from all the electronic crap in your home, if shielded ic's can prevent unwanted RF noise entering your system can this not be the same for shielded mains cables? (I am speaking from the viewpoint of electronic ignorence). Also the aftermarket mains cables may have better quality connectors which is where some differences could be heard. The main problem I have with the aftermarket mains cables is price... I have managed to get hold some Lapp shielded mains cable (the copper shielded version) so I am going to give it a try on the 'cheap' with some fleabay mains connectors.

And with all that outside interference, presumably the time of day one listens to one's hifi will affect the output. And of course, what time you ate, and how much (hearing being the only sense affected by hunger or satiation). And... well, all the other factors and variables.

But there comes a moment when we stop measuring signal, and start listening to music. It *is* subjective. It affects mood, expectations et al and are affected by them. There may never be a good hardcore scientific reason for the change I heard. Or there might be one now. Or in twenty years time. But i'll be listening to better sounding CDs while others are toiling to find those reasons. Bless you.

Incidentally, to add a little more perspective (subjective as always), we replaced the power cables on the CD player, phono stage, power amp and sub. If i remember rightly, I could hear a difference on the power amp, slightly more on the phono stage and thought i could barely discern a difference on the sub, but not enough that I'd want to put my hand on my heart and swear it wasn't placebo. The reason I mentioned the CD player in the first place, is that along with the sub it was the component I least expected to hear a difference on, and the scale of difference compared to the other components was notable.

Hence my earlier comment that it might not make a difference. But then so much of this is trial and error. What works for one system may make no difference to another. Personally, I believe that science advances through hunches, inspiration and sheer dumb refusal to believe what everyone else tells them is hard fact. Then if something seems to work, the hard work begins of finding out why, and challenging existing theory to that end. Mankind seems to need to be secure and unassailable in its various rationales, but it's in the bits that don't quite fit that the exciting stuff happens.
 
 

From a manufacturers point of view it is important to be able to establish the cause in hard facts, otherwise that manufacturer could not reliably repeat the circuit several thousand times. That is not to say that the customer must prove what he/she does, that would be absurd. Audio is a hobby and what makes people happy is OK by me. However, "tuning in the wrong direction" is a concern for me if it adversely affects the performance of products I make. I have spent more hours than I like to remember sorting out problems for customers that were due to "tuning in the wrong direction".

Recently, after many years of proving designs in the "old fashioned way" which includes some rather expensive test gear and the experience required in using it, I decided to repent in my stubborn ways and invest in a very good Spice Simulator. Running numerous new not yet released designs on it including some that aren't mine, I was surprised to see many of the wisdoms I'd been semi-influenced by being exposed as a complete and utter mess. I'd always found by my experienced ear that the good old JLH wisdom of bypassing electrolytic capacitors by a 100nF film capacitor didn't work - the simulator showed up exactly why: some really (not) nice big spikes in the upper frequency response - for the non-technical: distortion!

Recreating these circuits as befores and afters, the difference could be clearly heard. As I said "semi-influenced", which means I'd been tempted but had not put these things into products on the market. But with the dreaded legislation (which I'll not go into again here) there are several previously unheard of problems to be solved before the autumn in my case.

I just mention the above in passing, but with the right (mathematical) model, many myths can be laid bare by simulation, and many more can be given false credence through not modeling correctly, or dare I say it - deliberately modeling falsely.

And now for something completely different (well almost): Dave, you wrote...

"Not to forget that that lovely expensive custom cable is going to be screwed to a terminal which is wired to a PCB with the cheapest wire the manufacturer could find and then routed as signals along a minuscule thin copper pcb track radiating merrily to its neighbours before supplying cuurent to a bunch of components whose skinny little legs are bits of tin that a paper clip would be embarrassed to be made of... "

Dave, please educate yourself before making wild accusations that are blatantly wrong and misleading - I find the above to be highly insulting. Thank you (on behalf of Graham Slee Projects Limited and numerous other properly educated manufacturers).


Edited by Graham Slee - 21 Jul 2008 at 1:23am

Apologies , Graham.

Perhaps I should have said: "some manufacturers".  I've not examined the insides of a GS product. 

But the comment certainly applies to every pair of loudspeakers I have owned that I have looked inside (KEF, B&W, Royd, TDL). These have all been wired with thin cheap wire. If the last 3 metres of speaker cable are important, it's rather underdone by the last 12 inches of internal wiring....

Also most have crossovers clearly built down to a price. I had to re-do the crossovers on my TDLS recently because the capacitors had burst open.  They were grossly underspecified. The guy at Wilmslow told me that they are asked to fix dozens of these - the faults are always the same - the specified caps were specified to a barely adequate voltage. This seems typical - at least for low cost speakers.

Graham Slee wrote: From a manufacturers point of view it is important to be able to establish the cause in hard facts, otherwise that manufacturer could not reliably repeat the circuit several thousand times. That is not to say that the customer must prove what he/she does, that would be absurd. Audio is a hobby and what makes people happy is OK by me. However, "tuning in the wrong direction" is a concern for me if it adversely affects the performance of products I make. I have spent more hours than I like to remember sorting out problems for customers that were due to "tuning in the wrong direction". Recently, after many years of proving designs in the "old fashioned way" which includes some rather expensive test gear and the experience required in using it, I decided to repent in my stubborn ways and invest in a very good Spice Simulator. Running numerous new not yet released designs on it including some that aren't mine, I was surprised to see many of the wisdoms I'd been semi-influenced by being exposed as a complete and utter mess. I'd always found by my experienced ear that the good old JLH wisdom of bypassing electrolytic capacitors by a 100nF film capacitor didn't work - the simulator showed up exactly why: some really (not) nice big spikes in the upper frequency response - for the non-technical: distortion! Recreating these circuits as befores and afters, the difference could be clearly heard. As I said "semi-influenced", which means I'd been tempted but had not put these things into products on the market. But with the dreaded legislation (which I'll not go into again here) there are several previously unheard of problems to be solved before the autumn in my case. I just mention the above in passing, but with the right (mathematical) model, many myths can be laid bare by simulation, and many more can be given false credence through not modeling correctly, or dare I say it - deliberately modeling falsely. And now for something completely different (well almost): Dave, you wrote... "Not to forget that that lovely expensive custom cable is going to be screwed to a terminal which is wired to a PCB with the cheapest wire the manufacturer could find and then routed as signals along a minuscule thin copper pcb track radiating merrily to its neighbours before supplying cuurent to a bunch of components whose skinny little legs are bits of tin that a paper clip would be embarrassed to be made of... " Dave, please educate yourself before making wild accusations that are blatantly wrong and misleading - I find the above to be highly insulting. Thank you (on behalf of Graham Slee Projects Limited and numerous other properly educated manufacturers).