Can you explain that in reference to Amir’s graphs that everyone loves to copy and paste? What is he feeding it? Test tone I imagine. How does his measurements show this?
1 Like
@Elk Believe me, I know about the musical envelope created by varying instruments. My point was the frequency sweep isn’t the test to determine why a cable can sound different. You would have to test under load in the circuit under test with FFT, which would reveal it’s reactance to the circuit.
1 Like
I see. It appears @Shazb0t and I misunderstood your point in the same way.
My exact point in my question above. How does Amir’s simple tests equate to sound? Freq. sweep and noise. if they are the same they sound the same no matter what they are. I just don’t follow the reasoning. I have seem some really 3D graphs in reviews. I do not know how to read them. I have to go on the graphs creator to explain the results. In the case of Stereophile, and a few sites like Secrets who do produce these types of graphs, I have to go on trust of them interpreting them. In the case of ASR graphs, they are pretty simple to see and follow, but something is missing. IMO they are just the beginning, and dont tell the whole story. They are only the start, but they stop and draw their conclusions and move on.
To me the scary part is people are using them as the one and only… and spreading the word like a virus. Its simple and clean and easy to get sucked in and believe. I get it. But its not the whole story but they BELIEVE it is. Even @Shazb0t seems to know that, but is having a hard time to say Amir is misleading. That is probably the best way to describe ASR, misleading, and showing only 30% of the story. And the fact that they just chastise openly in public forum anyone that does not believe. That is why hobbyists, and not just audiophiles have issues with it. There are tons of people who dont have 50k worth of stuff that disagree with Amir’s methodology as well.
He has a very open public forum and if he started to produce more meaning measurements and not be such an egotist in his remarks, it could be a very useful site. But in its current form its more harm than good.
5 Likes
@Elk The problem with trying to deal with this issue is the complexity of the envelope created by music. It isn’t a single tone, and the subtleties we all know exists between equipment and cables cannot be easily dismissed by the simplistic testing methods many want to use to prove there is no differences.
1 Like
Moreover, the tone and colour of the same not can be varied numerous ways by different fingering, the note effectively being constructed from different harmonics.
You can hear these things, three different oboes playing the same notes in a piece by JS Bach, but you can’t detect identify them on an audio Precision plot.
This isn’t true. As pointed out above. These differences would 100% be measurable.
2 Likes
So, a question that has been asked here many times, by me and others is: can the sound quality differences of audio cables, for example, interconnects, all of the exact same design, but of different conductor material, for example, silver plated copper, vs OCC, be measured? (All other variables the same except the conductor material)
EDIT: Generally, as a group, we believe we can hear a difference in material and many of us have paid premiums to get that material.
1 Like
I just finished watching the video that started this thread and have a few questions.
It used the low rez built-in display on the power regeneration unit to show a regenerated 60hz waveform when much more accurate devices were at arms length away. Would it be possible to show the input/output wave forms on one of the more accurate o-scopes on the bench ? It would be nice to see the changes on a higher resolution display.
Has anyone connected one of these devices to an active amplifier running under load and analyzed the DC output coming from the amp’s power supply ? That seems like a more real world test.
3 Likes
Yes, the differences in the timbre of the three oboes is easily measurable.
In fact, they can be synthesized. The tricky part is understanding and replicating how the harmonics change over time, even on just a single held note.
4 Likes
boo.
Agreeing with you, let me add another perspective.
When a signal from an instrument is amplified (lets take the violin example @Elk posted), the amp treats each one of the harmonics as a “fundamental”. The amp does not know the difference between harmonics and fundamentals. So it will amplify each and everyone of said harmonics.
Should the amp add 0% distortion, the output would be the exact envelope that @Elk posted. But, if the amp adds, lets say 0.1% 2nd harmonic distortion, such distortion will be added to each of the harmonics of the envelope (assuming all distortion is correlated). So the 2nd harmonic will be a bit higher (in this case the equivalent to -60dB from the amplitude of the fundamental will be added to the harmonic), the third harmonic a bit higher (by the same -60dB of the 2nd harmonic of the instrument) and so on and so forth, so that you will have a 9th harmonic that will be purely the distortion product.
That is why measuring the amp by “static” tones is sufficient. This measurment will clearly show what the amp is adding to the signal it got from the source.
At least, this is the way I understand it. Should I be wrong, please do correct me.
1 Like
Hello Shazb0t,
Steady state at at any frequency won’t show how music is heard under dynamic conditions. We test amplifier’s into resistive loads for consistency but it doesn’t involve the reactance off the load. Does that matter.
Below is a plot showing IMPEDANCE and the reflection of a “100-ohm” vector load for a typical Ethernet cable that has varying reflections based on the size of the real (resistive) vector component. All the points on the Y-axis are @ “100-ohm” yet the reflection (X-axis) varies considerably. Why is that?
Just because we have a 100-ohm vector impedance does not mean we will remove the reflection coefficient. Audio has all sorts of reactive impedance data sets (real+imaginary) and different combinations of equipment will alter those issues. The Vp non linearity across frequency aggravates it, where the trace above is RF and not as impacted except by mechanical geometry variations. If we are almost a resistive vector, we see the minimum of reflections, -55 dB or so. But there are vectors with more and more reactive component @ -23 dB or so, not nearly as good. BOTH are 100-ohm vector magnitude!
If we also look at how cable really moves an EM field, we see the following graph. The Vp per physics has to go to “zero” @ DC and to the limiting speed of the dielectric’s Vp at RF. The data shows exactly that. Can we improve that “distortion”? Yes, we can. I have the engineering paper on this if you like. It is all accepted physics.
If we alter C and R we can determine a better optimization if the cable structure can yield the calculated variable values. We need about 60 pF/foot and a 28 AWG wire to make meaningful changes in the Vp slope. Higher or smaller values on either R or C impacts manufacturability and amplifier stability(*) into reactive loads.
By Soufiane Bendaoud [soufiane.bendaoud@analog.com]
Giampaolo Marino [giampaolo.marino@analog.com]
() Capacitive loads often give rise to problems, in part because*
they can reduce the output bandwidth and slew rate, but mainly
because the phase lag they produce in the op amp’s feedback
loop can cause instability. Although some capacitive loading is
inevitable, amplifiers are often subjected to sufficient capacitive
loading to cause overshoots, ringing, and even oscillation. The
problem is especially severe when large capacitive loads, such
as LCD panels or poorly terminated coaxial cables, must be
driven—but unpleasant surprises in precision low-frequency
and dc applications can result as well.
If we do make the calculated changes, we can graph the expected changes. Notice again it goes to zero at DC and to the dielectric’s Vp at RF, as theory says it has to.
Does the graph agree with the open-short impedance data? Yes, it does. It flattens at ~ 50% Vp AS PREDICTED IN CALCULATION. The higher capacitance of the series II pulls down the impedance at very low frequency, but neither design can be an eight-ohm cable. The drop in Vp raises the impedance as frequency goes down, and that’s the way it is unless we use excessive capacitance. Give me a cable and test it open-short and you will see this effect and right were the majority of the energy is being moved, and reflections are the worst (not attenuated out as much). We can lower impedance, yes, but to be an eight-ohm cable is not going to happen.
To hold inductance low in a speaker cable is important, and significant design changes need to be made to hold “L” low while changing the R and C to better improve Vp coherence. The ICONOCLAST does hold “L” to the same uH/foot through value while adjusting the Vp group delay properties with R and C changes.
Now the proper context. EM waves travel a 3E8 meters per second. That’s fast even at lower percentages. Our cables are shorter. Is there enough time to make a “difference” in the sound? We loop back to the “can we hear that” question. As long as the attributes are REAL, I have no issues with logical comparisons on the time base, and then actual in-use applications. We can provide either design extreme on speaker and IC both to compare. What we show is cable’s are indeed able to be better optimized for analog.
Use a 1313A zip cord style and use a series II and see. Both are available for trial. Both are different electrical for sure and to proper calculation and measurement. If we want better analog cables this is how I decided to design them and why. The design ethos also applies to IC cables as well, but the R and C values need to be changed for a voltage verses current signal transfer (high impedance verses low impedance load).
Best,
Galen
6 Likes
Thanks for the reasoned response. I agree that this is an interesting study. I do wonder, however, if this study concerning differences in the sound of cables simply verified expected results, why did it prompt such vicious ad hominem attacks against the author on ASR?
A lot of the comments I have read about differences between cables attribute those differences to marked variations in frequency response, resistance, inductance, or capacitance. However, those factors did not appear to have major effects in Kunchur"s AES study.
For those who haven’t read the study, I’ll just quote part of the conclusion below:
“This work did not conduct an exhaustive determination of all possible physical causes of sonic differences in interconnects. For example, time-domain effects such as reflections were not studied because a balanced cable requires a differential amplifier and extra cable (both adding their own noise and distortions) before an oscilloscope. However, the electrical measurements conducted here indicate that noise levels may be one determining factor of sonic performance. The measurements also show that characteristics such as resistance and frequency response, that naive consumers may focus on, are irrelevant for distinguishing HEA [High End Audio] interconnect cables.”
1 Like
No, it isn’t silly. If we have increased amplitude where did it come from? The cable is too short to have classic attenuation, so the superposition of the underlying variables are shifted in PHASE and this changes the final single voltage we measure. The change in the voltage is from a different spectral energy superposition and will impact how the cable sounds too, not simple attenuation. Move two like amplitude sine waves, and the energy will shift the amplitude as well as what you hear. Composite audio is no different. Part of the change is the perceived change in amplitude and sound quality. What causes that? Group Delay distortion in ALL cable’s ability to move data at the same speed across frequency. We can improve, but not eliminate that as the physics says we can’t. DO NOT CHANGE THE SETTINGS, what you hear is indeed the cable’s properties.
Best,
Galen
5 Likes
@jvvita Amp A and Amp B both sweep flat over the audio spectrum, and show similar THD properties, yet Amp A can sound totally different than Amp B under load. A simple single tone sweep does not tell us much other than it is capable of reproducing the audio spectrum.
1 Like
Paul touched on it pretty early on in the thread. It shouldn’t be an issue in current P12s. Older (and all for that matter) P12s with the current FW shouldn’t have any issue with the HC relay.
1 Like
WOW, my head is spinning. Much of this i sort of follow. But i will say I see more “Science” in this one thread, then the entire ASR site.
Also, I keep seeing people say “why didn’t Paul use the super sensitive equipment” instead of the cheaper Flukes? I have a question to the much smarter people here, Is that even nesessary for Paul to show what he was showing? Is there that big of a difference? I get it, in engineering you want best reading, but he is just doing something to prove a point. So what if its .05% off. Which I believe is probably the only difference we are talking about, right? People are so stuck on these numbers that in the end what is the diff. between this and that is so so small, its meaningless. These people seem to think they have to buy the one with the best numbers, thats it, thats all and there is nothing else.
Finally, I think we have proved that the few tests ASR does show, are just the tip of the iceberg of measurements that SHOULD be done before he declares his recommendation. Galen has schooled us on that.
2 Likes
I feel like we’re starting to talk past each other slightly. I am not contesting the validity of these electrical properties existence, but I am suggesting to you that they are so small in properly chosen audio cables as to be audibly negligible in output measurements from an audio system. I have provided the evidence of instrument grade measurements of said outputs showing negligible changes in properly designed and chosen cables for audio application. You have acknowledged these measurements, indicated system output measurements would be a valid way to make comparisons, and suggested that there are additional measurement setups involving playing music that will show the audible differences in the end system output that you claim will be due to the electrical properties that you’re discussing. I have suggested that if this is the case that you demonstrate those measurements. Do you have an example of any such audio system output measurements that compare cables in this way? If not, what is the reason that you believe this can not be done?
2 Likes
Huh?