• 08-12-2004, 05:44 AM
    kexodusc
    Is this "sound advice" or more "snake-oil" and "voodoo"?
    Had to buy some new speaker wire, got some 12 guage and 16 guage.
    When I was asked how much I looked to my speaker diagram at the measurements I made and gave him the totals
    ...the salesdude (who I know quite well) told me I absolutely have to use the same length of wire for each speaker, especiall the stereo pair in my home theater (though I have a dedicated 2-channel system and rarely use my HT for music)...I politely declined.

    I can't think of any reason why this would be true...so I'm at the mercy of the forums to convince me otherwise.

    Could using different lengths cause "delay"...doesn't electric signal travel pretty damn fast?

    Would this really effect "imaging" balance etc...or was this (as I suspect) a scam to get me to buy way too much speaker wire?

    What could go wrong here by using "different lengths"? I don't think I ever have used equal lengths before....
  • 08-12-2004, 05:55 AM
    markw
    Quote:

    Originally Posted by kexodusc
    What could go wrong here by using "different lengths"?

    Not a thing. But, the thought of neat, equal lenghts does have an undeniable appeal for those who espouse logic and neatness. It's so tidy it must be right.

    Doesn't do a thing for the sound, though.

    You're right about the speed electricity travels thru the wire. It's fast enough to be totally inconsequential for HT uses. Ever wonder how fast sound travels through air as opposed as to how fast electrons travel in the wire? Now there's an interesting thought...
  • 08-12-2004, 06:02 AM
    skeptic
    Snake oil, definitely snake oil.

    Electricity in a wire travels at the speed of light which is about half the speed in copper wire as it is in a vacuum. John Neutron could probably give you the exact speed and anybody could look it up. It is the speed of light in a vacuum times "the index of refraction (or its reciprical but it will be a fraction of one.) This is NOT the speed the electrons themselves travel in the wire which is very slow up to very high current. This IS the speed which determines how fast the change in voltage at the amplifier end of the wire is felt at the speaker end. Considering that the speed would take photons of light seven and a half times around the world at the equater in one second in a vacuum, the time delay for any speaker wire length in your room is a tiny fraction of a millionth of a second. (at 186,000 miles per second times 5280 feet per mile, that's about a billion feet per second in a vacuum. So say about five hundred million feet per second in wire. multiply a five hundred millionth of a second by the number of feet of wire difference and that's the difference in time delay between the two wires.) This is insignificant in any way at audio frequencies but can be very important at radio frequencies depending on the application.

    The other differences you will experience have to do with differences in electrical impedence which is capacitance, inductance, and resistance. However, since these are so small to begin with for the lengths of wire you will be using on both an absolute basis and compared with the electrical properties of your loudspeaker , their differences due to the differences in length between the two channels is also insignificant.
  • 08-12-2004, 06:21 AM
    jneutron
    Quote:

    Originally Posted by kexodusc
    Had to buy some new speaker wire, got some 12 guage and 16 guage.
    When I was asked how much I looked to my speaker diagram at the measurements I made and gave him the totals
    ...the salesdude (who I know quite well) told me I absolutely have to use the same length of wire for each speaker, especiall the stereo pair in my home theater (though I have a dedicated 2-channel system and rarely use my HT for music)...I politely declined.

    I can't think of any reason why this would be true...so I'm at the mercy of the forums to convince me otherwise.

    Could using different lengths cause "delay"...doesn't electric signal travel pretty damn fast?

    Would this really effect "imaging" balance etc...or was this (as I suspect) a scam to get me to buy way too much speaker wire?

    What could go wrong here by using "different lengths"? I don't think I ever have used equal lengths before....

    I run 300wpch with one run 3 feet, the other 45. It's #12. It's totally fine. For that application, it is strictly a dissipation issue..I don't want power lost in the cables.

    For "high resolution systems", you must consider the rlc of the wire and it's effect on the sytem..

    Consider 10 feet of #12 zip...it has about .2uH of inductance per foot..so, 10 feet of it will have 2 uH of inductance..at 12Khz, the upper limit of lateralization studies (that I've seen to date), that has a reactance of 150 milliohms..

    That reactance will cause a phase shift at 12Khz, and the phase shift will vary with frequency, as will the shifting of cues within the music that we use for imaging.

    In addition to being an inductor simply "in the way", it will also affect the damping factor of the system..ten feet of #12 has a resistance of 1.71 milliohm per foot, or 34 millihenries of resistance..

    at 8 ohm, .01 z out amp, damping factor is 800 at the amp..add the R, you get 181....the inductance by itself: add the inductive reactance...you get a damping factor of 50.

    If you consider damping factor as a compex number, and also consider the virtual image math resulting from two sources modelling a virtual third, the math can get a tad tricky...so I cannot presently answer the question about imaging balance...I've been discussing this with Gene and "mudcat" over at AH, and may publish a cleaner, more concise analysis which includes phase and temporal relationships vs wire measurements.

    Bottom line: use #12 or #14, length doesn't matter, polarity of course does..If you wish to play, do an experiment where you twist the wires, and double them up with another set in parallel..you will halve the inductance that way, and can see if it makes a diff..my guess is that you can in a very unscientific test, either fool yourself into thinking it did, or even that it did not make a difference..

    Oh, also...speed of travel within the wires is determined by the dielectric..and it is 40 to 80 % of the speed of light..it no matter..

    Besides, for all the processing that is done for 5.1, you won't see the wire contributing anything..

    Use zip...enjoy the movies..if lowering the inductance really makes a difference, eventually the big boys will jump on the bandwagon, and then the prices will drop tremendously with volume....which will probably plop a lot of the small esoteric wire vendors off the map..

    After all, I can buy an extremely low inductance, #14 awg equivalent cable off the shelf from belden, one that toasts damn near every single esoteric wire made..88232..use the shields, skip the center..If you need even less inductance, those can be easily paralleled, without that cross coax "solution" of JR's, which pays an inductive penalty by design..but I'm told the 88232 is hard to get...

    Cheers, John
    PS..skep got here first...
  • 08-12-2004, 06:36 AM
    kexodusc
    Well, I'd even argue against the neatness point...If they had to be the same length (and obviously the longest wire) you'd have a whole bunch of slack on 1, 4, or in my case 6 other speakers you'd have to worry about...
    I expected the answers I've received so far...better to ask though.
    I suppose his commission check was counting on the extra 40 ft or so...
    Correction:
    I bought 14 awg and 12 awg, the thicker obviously for my surrounds which are further away.
    I have some doubts as to whether or not even 14 awg is needed for the shorter cables...would 16 awg or higher work? That's all my brothers Energy HTIB has right now...is he missing anything? I'd guess the furthest speaker would need 15 ft of cable in his apartment...
    thanks all
  • 08-12-2004, 06:38 AM
    skeptic
    "I run 300wpch with one run 3 feet, the other 45. It's #12."

    This is an amazing difference. How did you accomplish it, have the amplifier next to one speaker and run the wire around the entire perimeter of the room to get to the other? Amazing, simply amazing.
  • 08-12-2004, 07:46 AM
    Swerd
    Quote:

    Originally Posted by kexodusc
    Could using different lengths cause "delay"...doesn't electric signal travel pretty damn fast?

    Would this really effect "imaging" balance etc...or was this (as I suspect) a scam to get me to buy way too much speaker wire?

    After hearing from an EE and a physicist, I thought a biologist should chime in too.

    The Haas Effect is a psychoacoustic phenomenon that helps us, as well as most animals with 2 ears, locate the origins of the sounds. (I am paraphrasing from http://www.sonicmagician.com/article...ss_effect.html )

    The Haas Effect, also called the Principle of First Arrival, is demonstrated by having someone pop a balloon at a distance outside. It is heard by both ears but at slightly different amplitudes and phase. There is a dominating cue by which the brain can pinpoint the origin of a sound within the boundaries that surround the listener. That cue is the difference in arrival times between your ears.

    Sound travels through air at approximately 1080 ft./sec. (340m/sec.) and arrives at each ear at slightly different times. This difference and the successive reflections assist the brain in determining the originating location of the sound. The delay times needed to alter our perception of the position of a sound source are anywhere from just under 1 millisecond to about 50 milliseconds. Once the delay time gets beyond 50 milliseconds the delay turns into an echo. The brain then perceives this as a second sound.

    Considering the difference between the speed of the electrical signal through wire (about 93,000 miles/second) and the speed of sound through air (about 0.2 miles/second) I think you would need an awful long length of wire going to one of your speakers to create at least a 1 millisecond delay time.

    Anyone care to calculate how long that wire would have to be? That would be a useful fact to drop the next time a salesman says you need equal lengths of speaker wire.
  • 08-12-2004, 07:56 AM
    kexodusc
    Quote:

    Originally Posted by Swerd
    Anyone care to calculate how long that wire would have to be? That would be a useful fact to drop the next time a salesman says you need equal lengths of speaker wire.

    This is my favorite answer so far...

    Well, that and a few months back someone said something to the effect "if a salesman says it...do the opposite"...that was "sound advice" too...
  • 08-12-2004, 08:00 AM
    jneutron
    Oh, geeeze...you gave me a good laugh there..I wasn't expecting that...

    HI Skeptic..

    It's in a 450 seat venue. I have a 100 foot snake running power to the balcony, a mike run from the stage center, and an unbalanced stereo line run to the amp on the right side of the stage..then a speaker on each side of the stage..

    A mono signal, equal channel output, gives a good balance when I stand in the center aisle..meaning the left cable doesn't do anything nasty to the power levels or gross things to the frequency response..

    Course, next year, I'm gonna run a 150 foot #12 and a 100 foot #12, with the amp in the balcony..by that time, I may also put a 125 foot #10 in the snake to allow a 1Kw sub in the center front of the stage..(the photog guy got a really good pic of the hydrogen explosion, I'm tryin to get the jpeg...) It'll make the snake lighter also...my 11 year old can barely carry it now.

    For my 5.1 system at home, I use the wire that came with it..It looks to be about #24...even though the system is "rated" 100 wpc, I can't get really excited about attaching garden hose cables to speakers that are about four inches wide..

    Yah, it's not 3Kw into 105 dB systems...but I really enjoy the dvd's..the sound just happens to be there for my viewing pleasure..

    I also can't get excited about sitting alone in a room, in the sweet spot, listening to music, attempting to discern virtual images...my life includes a very nice woman, there is nothing virtual about her...

    Cheers, John.
  • 08-12-2004, 08:04 AM
    jneutron
    2 Attachment(s)
    Quote:

    Originally Posted by Swerd
    After hearing from an EE and a physicist,

    Who's the physicist??

    Here's a graph of L-R time delays from a source ten feet away..

    <img src="http://forums.audioreview.com/attachment.php?attachmentid=580">

    Since the actual distance doesn't change that much near zero degrees, there will be no real amplitude variation.

    This graph is the Nordmark results...it shows what humans are sensitive to w/r to l-r delays..
    <img src="http://forums.audioreview.com/attachment.php?attachmentid=581">

    I'm investigating whether or not the cable inductance can affect the sound timing at the level we can hear.So far, a cursory examination of the numbers via vector analysis has: 8 ohm, #12 zip, 2 uH, 12Khz... 1.07 degree shift, which is 230 nanoseconds..not anywhere near audibility findings..I still seek..

    Cheers, John

    Hmmmm..images disappeared within the text body, so here they are again.
  • 08-12-2004, 08:34 AM
    jneutron
    1 Attachment(s)
    BTW...if you want to see how fast your signal is travelling down the wire, here's a simple graph to use.

    Knowing the L and C of your wire, per foot, plot where it is on the graph.

    It is not possible, by e/m theory (course, I assume I got it correct, but what do I know???), to exceed lightspeed, so anything below the blue line is not possible..

    for the geeks...100% is DC=1, 80% is DC=1.57, 60% is DC=2.75, 40% is DC=6.25

    Cheers, John
  • 08-12-2004, 09:42 AM
    Swerd
    Quote:

    Originally Posted by jneutron
    Who's the physicist??

    OK, an EE and a guy posing as a physicist :rolleyes:.

    Using those four values for speed of electrical signal in wire expressed as % speed of light in a vacuum, the signal would travel this many miles in a millisecond:

    100% = 186 miles/msec
    80% = 149 miles/msec
    60% = 112 miles/msec
    40% = 74 miles/msec
    That means that one speaker wire would have to be 74 to 186 miles longer than the other for a human to hear a noticeable delay time! Of course all bets are off if you use oxygen-free copper or silver wires ;).

    In contrast, sound travels through air about 1100 feet/second, or about a foot in a millisecond. So relocating your speakers about a foot closer or further than the other might have the same effect as 74-186 miles of wire.
  • 08-12-2004, 09:51 AM
    jneutron
    Quote:

    Originally Posted by Swerd
    OK, an EE and a guy posing as a physicist :rolleyes:.

    We're gonna have to take that poser out in the back and flog em...;)
    Quote:

    Originally Posted by Swerd
    That means that one speaker wire would have to be 74 to 186 miles longer than the other for a human to hear a noticeable delay time!

    You make the assumption that the only delays that occur are those due to propagation delays along the transmission line..I am considering inductance of the line as a lumped element.
    Quote:

    Originally Posted by Swerd
    Of course all bets are off if you use oxygen-free copper or silver wires ;).

    Damn...I forgot all about silver..:rolleyes:.

    Cheers, John
  • 08-12-2004, 12:35 PM
    woodman
    I was gonna jump in and answer this one for you, Ken ... but it's already been sufficiently covered by John (the Neutron bomber), skeptic, and Swerd. Nothing more to add.
  • 08-12-2004, 12:42 PM
    kexodusc
    Thanks anyway, Woodman...I expected as much, at least now I have some theoretical grounds that explain why for me.
  • 08-12-2004, 02:49 PM
    mtrycraft
    Quote:

    Originally Posted by kexodusc
    Well, I'd even argue against the neatness point...If they had to be the same length (and obviously the longest wire) you'd have a whole bunch of slack on 1, 4, or in my case 6 other speakers you'd have to worry about...
    I expected the answers I've received so far...better to ask though.
    I suppose his commission check was counting on the extra 40 ft or so...
    Correction:
    I bought 14 awg and 12 awg, the thicker obviously for my surrounds which are further away.
    I have some doubts as to whether or not even 14 awg is needed for the shorter cables...would 16 awg or higher work? That's all my brothers Energy HTIB has right now...is he missing anything? I'd guess the furthest speaker would need 15 ft of cable in his apartment...
    thanks all


    How short is the run witht he 14 ga? You have it in hand, so there is no need to replace it. Yes, if it was short runs, even 16 ga would work but you would not save enough to worry about.
    Your brother is not missing anything. Just use and enjoy what you bought:)

    Oh, the sales advice was snake.
    The speed of signal transmission is pretty fast. But, with good instruments we can still measure it so maybe not fast enough :D
  • 08-12-2004, 03:14 PM
    mtrycraft

    Since the actual distance doesn't change that much near zero degrees, there will be no real amplitude variation.


    So the perception is straight ahead. right? :)

    This graph is the Nordmark results...it shows what humans are sensitive to w/r to l-r delays..

    Of a single sound sources arrival times at each ear and the consequently different transfer function created in the canal and eventually interpreted by the brain, right?


    I'm investigating whether or not the cable inductance can affect the sound timing at the level we can hear.So far, a cursory examination of the numbers via vector analysis has: 8 ohm, #12 zip, 2 uH, 12Khz... 1.07 degree shift, which is 230 nanoseconds..not anywhere near audibility findings..I still seek..

    I am still grappling with this. The wire is delaying the frequency to the single source of sound as a whole. In a mono speaker it would mean nothing, right? In a two speaker setup, you would be interested in the time delay, the delta, between the two runs of wire?



    Hmmmm..images disappeared within the text body, so here they are again.

    The world of magic :)
  • 08-13-2004, 05:20 AM
    jneutron
    Quote:

    Originally Posted by mtrycraft

    Since the actual distance doesn't change that much near zero degrees, there will be no real amplitude variation.


    So the perception is straight ahead. right? :)

    Yes. I chose to use center stage as a start. It simplifies some of the equations..Unfortunately, to test lateralization at home, I only have the software ability to shift left to right by one sample increments, so the image shifts are going to be much larger than the inches I desired..perhaps my 24/192 card will have the ability to truncate much smaller time increments..

    This graph is the Nordmark results...it shows what humans are sensitive to w/r to l-r delays..

    Quote:

    Originally Posted by mtrycraft
    Of a single sound sources arrival times at each ear and the consequently different transfer function created in the canal and eventually interpreted by the brain, right?

    By keeping the image closer to center stage, I was hoping to keep the ear to ear transfer functions equivalent, leaving only the time delay as a measured entity..I recall radar o'riley had sent me some info that discussed how the ear frequency response varies with angle, so I wanted to avoid that confounding influence.


    I'm investigating whether or not the cable inductance can affect the sound timing at the level we can hear.So far, a cursory examination of the numbers via vector analysis has: 8 ohm, #12 zip, 2 uH, 12Khz... 1.07 degree shift, which is 230 nanoseconds..not anywhere near audibility findings..I still seek..

    Quote:

    Originally Posted by mtrycraft
    The wire is delaying the frequency to the single source of sound as a whole. In a mono speaker it would mean nothing, right? In a two speaker setup, you would be interested in the time delay, the delta, between the two runs of wire?

    The wire is delaying all the stuff within a channel differently, depending on the slew rate of the current..(the inductance of the wire will not allow the current to change as rapidly as the amp would like.)

    I am most certainly interested in the delta between the two runs..to image something off axis requires the two signals be timing different, and if the wire parameters cause a general shift in both channels, the different signals may shift differently..

    My example of 2uH/230 nSec shift is still unsettling to me...2 uH is a 63Khz breakpoint first order lowpass filter. the 230 nS was by inspection of a very simple vector analysis. I need to do a current/vector analysis to include the transfer function of a voice coil, the cone mass, all the inductive components, and I fear I will also have to include b-dot reaction within the speaker cabinet, which I haven't modelled yet for high slew rates...

    Wish I had my scanner setup here. I'd put a hand drawn pic up to snow tony...

    Hmmmm..images disappeared within the text body, so here they are again.

    Quote:

    Originally Posted by mtrycraft
    The world of magic :)

    It's probably that new site admin guy...hackin away..

    Seriously, this site has been really crazy w/r to images lately..first, it provides links at the bottom of the post, which I absolutely hate, cause it blows away any continuity of discussion..so I used "img src"....to get the pics within the text body...which worked for a while...then that plopped back to little x boxes..so I had to reload the images to that darn bottom of page stuff..not to mention, all the little pics that should be on the page are going south at various times..

    If eric is listening....I absolutely need to be able to use html "img src" code here...

    Cheers, John

    Quote:

    Originally Posted by woodman
    John (the Neutron bomber)

    ...your killin me...;)
  • 08-13-2004, 05:38 AM
    skeptic
    "I'm investigating whether or not the cable inductance can affect the sound timing at the level we can hear.So far, a cursory examination of the numbers via vector analysis has: 8 ohm, #12 zip, 2 uH, 12Khz... 1.07 degree shift, which is 230 nanoseconds..not anywhere near audibility findings..I still seek.."

    What is the typical unit to unit tolerance for loudspeaker system inductance when you take into account the crossover network and the voice coils themselves? How does this variation compare order of magnitude wise with the difference resulting from differences in cable lenghts?
  • 08-13-2004, 05:57 AM
    jneutron
    Quote:

    Originally Posted by skeptic
    "I'm investigating whether or not the cable inductance can affect the sound timing at the level we can hear.So far, a cursory examination of the numbers via vector analysis has: 8 ohm, #12 zip, 2 uH, 12Khz... 1.07 degree shift, which is 230 nanoseconds..not anywhere near audibility findings..I still seek.."

    What is the typical unit to unit tolerance for loudspeaker system inductance when you take into account the crossover network and the voice coils themselves? How does this variation compare order of magnitude wise with the difference resulting from differences in cable lenghts?

    Good questions...no direct answers yet..

    However, once the speakers are in place, and you sit down and listen, you accomodate those differences within your brain's interpretation of the virtual image...the grossest example would be sweet spot...if one cable caused a 1 dB amplitude drop, the sweet spot just shifts a little, so you move your chair..or head, whatever.. if it is half a dB, or 1/4....same thing..but if there is too much variance between speakers, say crossover tolerances or driver tolerances, the speakers probably won't sell..

    If however, the sweet spot for various frequencies is in different spacial locations as a result of some small inductance diff, say 1uH vs 2uH, then how does one accomodate??

    I'm even starting to re-think my use of pure resistance loads for cable testing, as an electrodynamic driver draws current as a result of velocity errors, whereas a resistor draws current directly proportional to voltage...it may be that I have to use drivers for real tests, or at least figure out a load which has mass equivalence..so the voltage/current vector analysis is consistent..Guess I'm gonna have to read up on that BEI site...

    Cheers, John
  • 08-13-2004, 06:05 AM
    skeptic
    "it may be that I have to use drivers for real tests, or at least design a load which has mass equivalence..so the voltage/current vector analysis is consistent.."

    It would be nice if people started testing amplifiers that way. Perhaps we could finally learn something from measurements about their differences instead of just their similarities. I'd also like to see them tested at levels other than one watt for all of the parameters that are now tested at this low level only.

    It seems to me that with modern equipment, it is entirely feasible to test amplifiers using very complex waveforms similar to music and analyze the differences between input and output with real world loads using computerized analysis to find out much more difinitively what types of distortions are really introduced and where meaningful performance differences exit.

    The problem with much current testing of equipment is that it uses 1930s concepts to measure the technology of 2004. By these antique standards, much equipment tests textbook perfect but does not always perform the same way. Push the envelope.
  • 08-13-2004, 06:20 AM
    jneutron
    Quote:

    Originally Posted by skeptic
    "it may be that I have to use drivers for real tests, or at least design a load which has mass equivalence..so the voltage/current vector analysis is consistent.."

    It would be nice if people started testing amplifiers that way. Perhaps we could finally learn something from measurements about their differences instead of just their similarities. I'd also like to see them tested at levels other than one watt for all of the parameters that are now tested at this low level only.

    It seems to me that with modern equipment, it is entirely feasible to test amplifiers using very complex waveforms similar to music and analyze the differences between input and output with real world loads using computerized analysis to find out much more difinitively what types of distortions are really introduced and where meaningful performance differences exit.

    The problem with much current testing of equipment is that it uses 1930s concepts to measure the technology of 2004. By these antique standards, much equipment tests textbook perfect but does not always perform the same way. Push the envelope.

    Agree.. Doesn't MLS testing do that?

    Push the envelope???nah...vaporize it..that's what they do here at work...

    I had to laugh at myself the other day (I spend a lot of time doing that).
    Discussing electron velocity a while back...curl asked what electron velocity was in supers..I gave a very good explanation as a first pass, to the best of my knowledge..
    Then went to a co-worker (who is one of the best superconductor physicists in the world), asked him....he said...and I quote.."what electrons?"

    Geeze, what we don't know...

    Cheers, John
  • 08-13-2004, 07:03 AM
    skeptic
    As I said in my posting about visiting the Bizzarro World of the Asylum, I recall Mr. Curl's postings on the subject of Fermi Velocity very distinctly. It seemed to me that for him, this was just one more technobabble buzz word used to snow tyros and win his points through intellectual intimidation rather than through a well reasoned arguement. I had to laugh my sides off when he cited the same sophmore undergraduate physics book that I used for his reference. In a way I felt sorry for him. He seems to be one of those people who are not only sometimes way out beyond their depth but don't have a clue about it. He may be a fine amplifier circuit design engineer but I got the strong feeling that as a physicist, he's not even competitive at the undergraduate level.

    Getting back to the topic at hand, you reported a difference of two microhenries for your run of 12 gage wire but won't most loudspeaker systems have an inductance of dozens or hundreds of millihenries considering just the choke in the woofer circuit alone? Aren't many tweeter circuits shunted by 1/2 millihenry (500 microhenries) or more directly across the load? This would put the cable inductance at a small fraction of one percent of the load inductance the amplifier sees. Don't you think the unit to unit variation far exceeds that amount given normal manufacturing tolerences?

    Of course while to most people 230 nanoseconds (less than a quarter of a millionth of a second) difference at 12 khz is inaudible, to a man with the hearing abilities of Jon Risch who can hear digital jitter in cables, even after the signals are reclocked, this is totally unacceptable. BTW, that is less time by almost a full order of magnitude than it takes for a nuclear explosion at the core of an atom bomb.
  • 08-13-2004, 08:01 AM
    jneutron
    Quote:

    Originally Posted by skeptic
    Getting back to the topic at hand, you reported a difference of two microhenries for your run of 12 gage wire but won't most loudspeaker systems have an inductance of dozens or hundreds of millihenries considering just the choke in the woofer circuit alone? Aren't many tweeter circuits shunted by 1/2 millihenry (500 microhenries) or more directly across the load? This would put the cable inductance at a small fraction of one percent of the load inductance the amplifier sees. Don't you think the unit to unit variation far exceeds that amount given normal manufacturing tolerences?

    The 2 uH is a ten foot length of #12 zip.

    Since the lateralization stuff I've come across goes from about 1Khz to 12Khz, I'll stick to that band..leaving the woof out of the pic w/r to imaging..although it may have an overall impact when it forces the amp into quadrants 2 and 4.

    Yes, the ratio's you state are right on, so that makes it odd that we (well, me anyway) are having a discussion of what the darn wire could possibly contribute. But it's that darn 1.5 uSec lateralization thing I can't get around...even 5 uS....the virtual image we look for is so damn...delicate, it's such a preposterous thing, sound where no object resides..(what are we thinking?)

    Theoretical example:
    Current slew rate into a pure resistor... 40 volts square wave..infinite rise speed, period 1 Khz.. two zips...one ten feet, other 20..for a ten footer, 2uH...other, 4 uH..both trivially small compared to the speakers..

    2 uH means it will slew initially at 20 volts per microsecond..meaning, about 3 to 4 uSec to nearly top... 4 uH means 10 volts per microsecond..about 6 to 8 uSec to nearly top..

    First pass...two ten footers...symmetry, totally..the image is dead center..

    Then, sub a 20 footer for the right one..all that was added was 2 uH..but look at the slews...the right speaker now takes about 6 to 8 uSec to achieve the same current that the left does in 3 to 4 uSec...

    Nordmark says that difference falls within the bounds of audibility..my thought is..the leading edge in this example is too fast, but if we wait until the slew rate falls to 12 Khz sine rates, maybe we key to that..

    Base L-R differences for a point source of 3-4 uSec is very small shift if you look at my simplistic graph....about 1-2 inches..I don't think I can hear that..but that is a dead center calculation, I will re-visit that graph, in two dimensions, for angles up to about 45 degrees off axis.

    So, this is a possible explanation for some kind of soundstage shift with a long and a short zip...not a concrete one, by any means..but an interesting start..


    Quote:

    Originally Posted by skeptic
    Of course while to most people 230 nanoseconds (less than a quarter of a millionth of a second) difference at 12 khz is inaudible, to a man with the hearing abilities of Jon Risch who can hear digital jitter in cables, even after the signals are reclocked, this is totally unacceptable. BTW, that is less time by almost a full order of magnitude than it takes for a nuclear explosion at the core of an atom bomb.

    I really enjoyed Clancy's description of that process in (I think) sum of all fears.

    Cheers, John
  • 08-13-2004, 08:29 AM
    skeptic
    You can look at it another way. There is a direct relationship between the phase shift and the fequency response falloff for the equivalent lumped sum parameter circuit. 230 nanoseconds only amounts to a tiny fraction of a tenth of a decibel at 12 khz. This is also within the variance of the manufacturing tolerences of even the best speakers. Even the grill cloth will introduce that much variance in falloff. In real world terms, don't you think this borders on the rediculous? Or maybe it has crossed the border and moved well into lunacy land.