For the musical selection I'd try dead silence with lots of gain, or if that's "not allowed" I'd try some notably quiet passage.
Some "Good Engineering Practices" can not be ABX tested. When Mr. Brown's papers or Mr. Ott's book write about ways to reduce susceptibility to EMI/RFI no test can be performed that covers all conditions. You would first have to have a system that is being interfered with (or is your system interfering with something else?) before starting the test. But the possible causes, the possible cures and the possible system configurations is limitless, so our ABX tests will go on forever.
Quote from: mzil on 08 November, 2012, 10:38:01 AMFor the musical selection I'd try dead silence with lots of gain, or if that's "not allowed" I'd try some notably quiet passage.Changing the gain will have no effect on speaker wires. In fact, perhaps the way to test is with the amplifier turned off. Otherwise you could be detecting noise pickup before the amplifier.
@all:Would there be anything wrong with recording the output of an amplifier at the speaker terminals with either a load box or with the speakers connected?
Quote from: Speedskater on 08 November, 2012, 12:18:29 PMSome "Good Engineering Practices" can not be ABX tested. When Mr. Brown's papers or Mr. Ott's book write about ways to reduce susceptibility to EMI/RFI no test can be performed that covers all conditions. You would first have to have a system that is being interfered with (or is your system interfering with something else?) before starting the test. But the possible causes, the possible cures and the possible system configurations is limitless, so our ABX tests will go on forever.That is the usual cop-out which gets rejected time and again on this forum.If you wish to make claims you need to provide proof. The burden is not on me to prove that such claims could never be right.
What is happening is the interference entries the amp via the speaker wires, then sneaks back to the input stage though the negative feedback network and gets demodulated and amplified.Mr. Brown explains it much better than that.
@mzil:Do you mean similar to cranking the volume well past normal loud listening levels just to hear fade-outs?I think that's cheating, though running the amplifier without feeding it content is perfectly OK in my book.
I have owned quite a few different speaker cables in my time and have always noticed the copper strands of wire within the cable run straight and parallel to each other I.e. they are not twisted. In fact the same is true of mains cable.
Can anyone tell me why this is I was told the electricity has trouble flowing through twisted strands, if each strand was enameled you might get an inductance effect but obviously it is not so the electricity should take the straightest path, so if the strands are twisted or not should have no overall effect.
Can anyone tell me why the copper strands are never twisted.
Really all cables be they power, signal interconnects or speaker should be twisted! (save for co-ax)Jim Brown writes:Twisting Cable pairs are twisted together for two very important reasons. First, bringing themmore tightly together reduces the coupling of external magnetic fields (while increasing thecoupling between the conductors) by reducing the loop area between them. Second, twistingthem together in a very symmetrical fashion causes any noise coupled onto one conductor tobe more perfectly cancelled (in the receiver) by noise coupled onto the other conductor.Twisting reduces both magnetic (inductive) and electric (capacitive) coupling.To understand how twisting does this, consider a magnetic field from a source that is closer toone side of the cable than the other. At any point along the cable, one conductor will becloser to the source than the other, so the induced voltage will be greater in that conductorthan in the other. But one half twist along the cable in each direction, the other conductorwill be closer to the source, and so will have the greater induced voltage, but the polarity willbe opposite. The more symmetrical the twisting, and the "tighter" the twisting, the more perfectly the two induced voltages will match each other over the length of the cable, and thusbe better cancelled by the receiver. The number of twists per unit length is called the "lay" ofthe cable.Twisting also reduces capacitive coupling onto the cable, and for the same reasons. The abilityof twisting to reduce coupling extends to very high frequencies. Ethernet networks run onhigh quality, unshielded, twisted pairs at frequencies in the hundreds of MHz, and requiregood crosstalk rejection to function well.Cables, Transmission Lines, and Shielding for Audio and Video Systemsby Jim Brown - Audio Systems Group, Inc.http://audiosystemsgroup.comhttp://audiosystemsgroup.com/TransLines.pdf
From page 2 of the above paper:Output Wiring is Important Too! It is well known, for example, that RF interference is often coupledinto the output stage of audio equipment – for example, the power amplifiers that feed loudspeakersor headphones. There is always feedback around that output stage, so RF present at theoutput will follow the feedback network to the input of a gain stage, where it will be detected andamplified.
But my question is how do you time align the two recordings during the playback test? Manually? Restarting from the beginning still may have a "tell", a giveaway, in that the listener would be subconsciously picking up on the different delays before the music starts.
^ Some consider silence as a test signal "cheating".
Having "no problem at all" now doesn't mean adding a light dimmer in one's wall won't be a big problem a year from now, all because one concluded, "UTP in-wall speaker wire is for the voodoo believing nuts and costs more. I'm going to save money and buy the parallel runs instead".
The above could possibly be true in a situation like near a high-powered radio transmitter. I've seen some crazy stuff happen in radio stations that had their transmitters in the same room that their announcers worked in. It was probable that a radio station in this dire of a circumstance had some errr, expedient engineering.Back in a typical residential situation, it doesn't happen.
Quote from: mzil on 08 November, 2012, 04:35:46 PM^ Some consider silence as a test signal "cheating".In the event that you used a caret to refer to a previous post, no one claimed using silence as a test signal was cheating.Rather, it was cranking the volume beyond what you would ever use for listening to real content that was suggested (by me) as cheating.
It can happen in a non-typical situation. At one time I lived in a house that was almost under the antenna of a 2 KW AM broadcast transmitter... ...I had to wind about 20 turns of the speaker leads through a large toroid to eliminate it. (The RF field was quite strong, if I hovered the palms of my hands just above the keyboard of my PC it would type random characters..)