When you state it as objective fact, i.e. "Maybe about the (audible) fact that different interconnect cords have different sound signature," then yes, you do indeed have a burden of proof.
se
So if there is nothing really to show as an ear shattering difference, is it I am measuring the wrong things or is there nothing there?
The THD measurements are fairly classic, generally well understood and some are quite skilled in reading them. The 19 and 20k IM plots are to me quite handy in understanding more complex interactions.
Live music and to a lesser extent recorded music have some interesting characteristics that have not been allowed for. As a general rule of thumb the energy in music rolls off at 3db per octave. In live music a 6db frequency response bump is almost always added. Throw in the Fletcher Munson curves that show low frequencies need to be 15 db-ish louder to be perceived as the same volume as mid range ones. (I take this to mean the masking is also reduced!)
So I decided to try an IM test with a signal more resembling music. A 100 hertz tone that was 26db higher in level than a 2000 hertz tone. (13+15+6 would be worst case for a duo matching levels by ear and then eq'd.)
The results are shown. Now this gets interesting. There are sidebands to the 2k tone. These are not caused by FM modulation. They are classic IM. In this case they are thermally induced by the feedback resistor.
The critical band method of rating masking puts it around 200 hertz at 2k so these sidebands are close to or may be a problem.
The open question is are there recordings that have enough low frequency energy that the thermal IM will yield enough energy to show up. Or are there gifted individuals who can pick it up at the levels shown.
So my conclusion is that under certain conditions it is in the realm of possibility that the difference of a single resistor may be perceived!
The THD measurements are fairly classic, generally well understood and some are quite skilled in reading them. The 19 and 20k IM plots are to me quite handy in understanding more complex interactions.
Live music and to a lesser extent recorded music have some interesting characteristics that have not been allowed for. As a general rule of thumb the energy in music rolls off at 3db per octave. In live music a 6db frequency response bump is almost always added. Throw in the Fletcher Munson curves that show low frequencies need to be 15 db-ish louder to be perceived as the same volume as mid range ones. (I take this to mean the masking is also reduced!)
So I decided to try an IM test with a signal more resembling music. A 100 hertz tone that was 26db higher in level than a 2000 hertz tone. (13+15+6 would be worst case for a duo matching levels by ear and then eq'd.)
The results are shown. Now this gets interesting. There are sidebands to the 2k tone. These are not caused by FM modulation. They are classic IM. In this case they are thermally induced by the feedback resistor.
The critical band method of rating masking puts it around 200 hertz at 2k so these sidebands are close to or may be a problem.
The open question is are there recordings that have enough low frequency energy that the thermal IM will yield enough energy to show up. Or are there gifted individuals who can pick it up at the levels shown.
So my conclusion is that under certain conditions it is in the realm of possibility that the difference of a single resistor may be perceived!
Attachments
What power level? How about nice big fat 5W resistors. A properly sized feedback network will have little or no thermal modulation. This has been dealt with before.
Last edited:
Indeed.
It is proved to myself.
I don't hear, therefore I'm deaf.
Actually the big fat 5 watters did not measure very well. This was a test between two 1/4 watt rated resistors.
Power level certainly plays into the amount of distortion, so I tried to keep levels in the range that would equate to a 70 to 90db reproduced level. Away from clipping and louder than headphones.
The question under test was is it possible to hear the difference between a CC and MF resistor. No more. Others have certainly made claims beyond that.
Oh by the way I tried reducing the bias back to the starting point and the THD v level curve only change a little bit!
The interesting result was that the distortion increase in the amplifier was greater than my measurement method was showing. As I use a bridge method some of the actual distortion is masked by the references.
So when my measurements show -120db in a circuit that is not a bridge that may come out as -80. Don't yet have exact numbers, but the math shouldn't be too hard.
John,
No matter how closely I hold power cords to my ears I have yet to hear anything
If the power cord affects the sound of an audio thingy, I would start looking at the induced AC onto the chassis, as that is really easy. Next I would look at the power supply.
ES
I love it. A subjective fact!!! You heard it. A vast power distribution grid of thousands of miles of wire, hundreds of generators, transformers, miles to your substation, hundreds of feet from the nearest transformer to your circuit breaker panel, dozens more to your receptacle, and the last three feet of power cord makes all that difference. My, my. Only one conclusion to draw from that. Your power supply stinks.
Since thermal distortion goes as V^2 you've got an 80db discrepancy there which one should be able to track down.
I find it hard to believe that my phono stage, CD player, pre-amp and power monoblocks (all are commercial ones), all have PSU malfunctions.
Listening to music, both live and reproduced, is subjective.
You know, on my stereo set I listen to music with my ears, not with measuring equipment. So, everything influencing the experience of listening to reproduced music have, necessarily, a subjective aspect to it.
Your mileage may differ.
You know, on my stereo set I listen to music with my ears, not with measuring equipment. So, everything influencing the experience of listening to reproduced music have, necessarily, a subjective aspect to it.
Your mileage may differ.
Yes we agree on this It is both level and that my reference should only be 6db lower than the source is resulting in low resultant readings.
Have you ever tried inverting the AC supply lines? Try disconnecting all of the interconnect cables and measuring the voltages between all the chassis. Also check to ground. Yes there always is some capacitive coupling. Now swap AC power lines for the lowest reading. You will be pleased with the results.
Why? Because the manufacturers have no standard for that and do what works best for them, so unless all your gear comes from one place expect some differences.
I haven't done it, however, none of it explains audible differences between different power cords and power distribution boxes.
Oh really? Are there no IEEE standards for shielding, line and load regulation, and AC ripple? No regulations for impedence. No standard for rating power supplies capabilities according to established specifications and accepted test procedures? What about FTC or can you just make any claim for any power supply?
- Status
- Not open for further replies.