Wavebourn (et al),
OK. So, assuming that we agree that steady-state THD-type specs are way over-emphasized, and that good-enough ss distortion specs are relatively easy to achieve, anyway, then where is a list of more-important (or other important) types of measurements (or phenomena to watch out for)? (if for nothing else then just so that we part-time amateurs can try to more-efficiently get "up to speed")
My primary goal is simple: the most-accurate reproduction of the source material (whether I like it better that way or not).
Books? Websites? A short list that could be used for searches?
(Just scratching the surface is fine, at this point. But please be as specific as possible.)
Regards,
Tom
Last edited:
Go the SoundStage website. Under speaker reviews you will find a list of NRC tests of various speakers. Look at those especially in the response and distortion graphs. This is a good starting place. Also waterfall plots are good if you can be sure they used the same time interval. Regards
Hi Tom;
you can find lots of details here on the forum, people in polemics often reveal what is needed, but it is hard to understand if not to have first - hand experience, because all arguments sound valid. 🙂
Yes, my goal is the same, and it is simple: the most-accurate reproduction of sounds as we hear them, And I go to this goal the whole my life step by step, but can't tell that I achieved the goal.
I can't give you a recipe, because the recipe depends on ingredients. I am still learning and searching for best ingredients, and best recipes.
What is most impotent to me? Most important is when subconscious reaction on sounds happen before I realize that sounds are not real, they are reproduced.
Anatoliy
Last edited:
The exceptions that prove the rule..?
If there's one thing I've learned from this forum, it's that EE101 is not all that useful in the world of hi fi. You can plug numbers into formulae all you want, but all you will get out are numbers, and the interpretation of numbers is entirely subjective in these parts. If I calculated a noise & distortion figure of -110dB from using a certain configuration, in the past I would have thought it satisfactory. However, I have since learned that such a level of noise causes oracular haemorrhaging if it's the wrong kind. 1% THD can apparently be quite acceptable if it's a different type.
A few posts back you mentioned that by lowering earth impedance with a special mains cable you could suppress noise (hum?) by a further 20dB than it would be otherwise. (It's simple EE101.) But 20dB of nothing is still... nothing. What would be the special feature of this mains cable anyway? How thick would it be? If I moved the amplifier closer to the wall socket and shortened my standard cable to 5cm would that achieve the same effect? (I could also move the wall closer to the amplifier by extending the unseen twin-and-earth cable behind it...)
Hi,
Perhaps, perhaps not. It is worthwhile having a look around the insides of real top-notch studio's that turn out good recordings and mastering. You would surprised how many "exceptions" there are (more than rules)...
Not quite.
The problem is that you must apply your formula thoroughly, first order approximations will not do.
How much ground fault current is allowed for a piece of equipment that is "safe"?
What are the impedances in the loops in which the currents circulate?
What is the spectrum of this current?
What is the expected noise spectrum that will appear with the signal?
What is the resultant contamination of the output of the system in terms of direct noise, intermodulation?
All that needs solid EE101.
Of course, it is much easier to just say "I cannot see how the last 1m interconnect/mains cable [delete as appropriate] can make a difference." than to rigorously apply correct analysis...
Certainly these effects are clearly measurable using suitable methodology. Nothing esotheric involved either. But if I want to know the impact on a given cable in a given system, then to measure the cable independently and without reference to what happens in the systems does not seem like a good way to me.
Now we are debating audibility.
That is we now need to place our measured performance data in a context to make it information. Sadly this context is incomplete and replete with bad science and faith based systems of thought.
Well, but do you have "nothing"? have you actually ever tested a complete interconnected system, for arguments sake from the pickup leads (which may plugged into an AP2 instead of a cartridge) all the way to speakers (from where we may run a set of cables back to the AP2 - headphones recommended during measurements) or even better by measuring the acoustic output of the speakers?
Have you verified that under THOSE conditions you have "nothing", actually measured?
Appropriate electrical parameters, what else?
As thick as needed to fulfil the electrical parameters needed and to comply with electrical safety.
No, as it is not an issue with distance to the wall as such. Please see the diagrams posted earlier. It should be quite clear what is going on.
Ciao T
Perhaps, perhaps not. It is worthwhile having a look around the insides of real top-notch studio's that turn out good recordings and mastering. You would surprised how many "exceptions" there are (more than rules)...
Not quite.
The problem is that you must apply your formula thoroughly, first order approximations will not do.
How much ground fault current is allowed for a piece of equipment that is "safe"?
What are the impedances in the loops in which the currents circulate?
What is the spectrum of this current?
What is the expected noise spectrum that will appear with the signal?
What is the resultant contamination of the output of the system in terms of direct noise, intermodulation?
All that needs solid EE101.
Of course, it is much easier to just say "I cannot see how the last 1m interconnect/mains cable [delete as appropriate] can make a difference." than to rigorously apply correct analysis...
Certainly these effects are clearly measurable using suitable methodology. Nothing esotheric involved either. But if I want to know the impact on a given cable in a given system, then to measure the cable independently and without reference to what happens in the systems does not seem like a good way to me.
Now we are debating audibility.
That is we now need to place our measured performance data in a context to make it information. Sadly this context is incomplete and replete with bad science and faith based systems of thought.
Well, but do you have "nothing"? have you actually ever tested a complete interconnected system, for arguments sake from the pickup leads (which may plugged into an AP2 instead of a cartridge) all the way to speakers (from where we may run a set of cables back to the AP2 - headphones recommended during measurements) or even better by measuring the acoustic output of the speakers?
Have you verified that under THOSE conditions you have "nothing", actually measured?
Appropriate electrical parameters, what else?
As thick as needed to fulfil the electrical parameters needed and to comply with electrical safety.
No, as it is not an issue with distance to the wall as such. Please see the diagrams posted earlier. It should be quite clear what is going on.
Ciao T
Well, as you say above, audibility is debatable, so the appropriate electrical parameters are also debatable.
But just to humour me, what are the vital differences in electrical parameters between this:
UK KETTLE LEAD C13 (2098/2098) - www.misco.co.uk
(seems a bit pricey at £3.45)
and this? (or any other mains cable you might care to mention)
Audioquest NRG WEL Signature Series Mains Power Cable
(very reasonable at £3149)
I guess it's this section that will help with our calculations:
Just because some cable manufacturers choose not to use such 'imaginative' marketing doesn't mean they're any different.
What if the 'appropriate' cable is combined with 15m of (solid core!) twin-and-earth prior to it? (Plus the unknown cable back to the substation? What interference is there on the same phase? Is there a car body repair shop nearby? A taxi office? What equipment do they use? What is its "spectrum"? In other words where do you get your figures from to plug into your EE101 calculations?)
What if it turns out not to be 15m, but 20m of twin-and-earth? Does this upset our EE101 calculations?
Edit: the more I think about it, the more absurd it seems. If I had a problem with breakthrough from taxis or a nearby welding shop, the last thing I would put money on to cure it would be a replacement mains cable. It's just silly, isn't it? EE101 or not.
Last edited:
How do they know? Have they measured this distortion? Silly me, of course this is immeasurably small distortion yet which is audibly blindingly obvious to those with fine ears and the appropriate bank balance.
Shields itself, from what? A spiral which is wound the opposite way to what? Hyperlitz: the only meaning which could be attached to this is that it is like Litz only more so, so very fine separately insulated conductors - yet they say they use solid conductors. Did they mean to say Hypolitz?
Even if this were true, which it is not, the mains cable is not carrying multi-octave audio but 50/60Hz plus harmonics (varying at syllabic rate) and possible some LF.
Hi,
First, I said that audibility is a different debate. We had more than enough of that.
Let me put it this way, contacts have usually 50mOhm or more contact resistance, loop resistance for the cable may be several 100 milliOhm and several 100nH.
Leakage currents up to 10mA are allowed. If these flow halve in the mains cable and halve in the audio cable, with around 500milliOhm loop impedance for each loop, what level of noise do we have in the loop, compared to a 2V "0dB" level?
I would not know, nor do I particularly care.
If you want to know, you'll have to buy them and test them, won't you?
Then you have created a very badly set up system that almost invariably will have massive hum necessitating a "cheater adaptor" to stop the earth loop from making the system literally unlistenable.
What has that got to do with anything? Do you actually understand the electronics involved and referenced or are just mindlessly parrotting counter-factual BS others have equally brainlessly advanced to support their own agenda?
Then you have yet another problem, distinct from those you have due to the ones discussed by me. Cable construction may also have an impact here, of course, be it due to shielding, inductance etc.
Why, I either measure them or work from worst case based on regulatory compliance (UL/CE/FCC etc.). How else would I do it?
No, not in a system where all units are plugged into a local power distribution strip, one with a modicum of filtering that ensures that live, neutral and PEN are at the same potential for most of the audio range, which helps if you have a mix of Class I and Class II gear (and only then).
Ciao T
First, I said that audibility is a different debate. We had more than enough of that.
Let me put it this way, contacts have usually 50mOhm or more contact resistance, loop resistance for the cable may be several 100 milliOhm and several 100nH.
Leakage currents up to 10mA are allowed. If these flow halve in the mains cable and halve in the audio cable, with around 500milliOhm loop impedance for each loop, what level of noise do we have in the loop, compared to a 2V "0dB" level?
I would not know, nor do I particularly care.
If you want to know, you'll have to buy them and test them, won't you?
Then you have created a very badly set up system that almost invariably will have massive hum necessitating a "cheater adaptor" to stop the earth loop from making the system literally unlistenable.
What has that got to do with anything? Do you actually understand the electronics involved and referenced or are just mindlessly parrotting counter-factual BS others have equally brainlessly advanced to support their own agenda?
Then you have yet another problem, distinct from those you have due to the ones discussed by me. Cable construction may also have an impact here, of course, be it due to shielding, inductance etc.
Why, I either measure them or work from worst case based on regulatory compliance (UL/CE/FCC etc.). How else would I do it?
No, not in a system where all units are plugged into a local power distribution strip, one with a modicum of filtering that ensures that live, neutral and PEN are at the same potential for most of the audio range, which helps if you have a mix of Class I and Class II gear (and only then).
Ciao T
Silly of you to expect a straight answer.
It seems you know the correct answer. But that may be because you're not selling anything. 😀
Hi,
Do you know how an AC signal actually travels through a conductor (hint, it does not do that at all)?
Copper oxides have interesting electrical properties. For multi-stranded copper cable it's presence may be safely assumed, more-so if the insulation is PCV.
And the distortion of a sufficiently degraded multi-strand cable is hardly only theory, nothing more exotic than am AP2 will do.
Ciao T
Do you know how an AC signal actually travels through a conductor (hint, it does not do that at all)?
Copper oxides have interesting electrical properties. For multi-stranded copper cable it's presence may be safely assumed, more-so if the insulation is PCV.
And the distortion of a sufficiently degraded multi-strand cable is hardly only theory, nothing more exotic than am AP2 will do.
Ciao T
how did this digress from respectable physics and issues about mains filters to mains cables?
A mains cable must :
one, not catch fire
two,not fall out
three, not open
If you believe different, someone will take your money. Keeps the economy going.
Now can we can get back to what makes amps actually sound different?
Updated my sim with the correct Exicon MOSFETs. No luck at all getting stability by taking VAS output to the feedback point. If I can hold my head up long enough today, I will play with the margins and figure out the Simetrix pole analysis tool. The model says I have 100Khz bandwidth. How does one reduce that without slew limiting the stages or the excessive loading the original VAS pole imposed? Maybe the output newtork is not so evil after all.
A mains cable must :
one, not catch fire
two,not fall out
three, not open
If you believe different, someone will take your money. Keeps the economy going.
Now can we can get back to what makes amps actually sound different?
Updated my sim with the correct Exicon MOSFETs. No luck at all getting stability by taking VAS output to the feedback point. If I can hold my head up long enough today, I will play with the margins and figure out the Simetrix pole analysis tool. The model says I have 100Khz bandwidth. How does one reduce that without slew limiting the stages or the excessive loading the original VAS pole imposed? Maybe the output newtork is not so evil after all.
Are you saying that the electrons in the conductor don't move and so don't contribute to the signal? For almost all audio purposes the quasi-static (i.e. circuit) approximation of EM theory is valid. If not we would not be able to design amplifiers using lumped components like resistors and capacitors, and circuit theorems like Kirchoff's.ThorstenL said:
My point is that a mains cable should not be carrying audio signals, and even if it is the equipment should ignore it. Any audio distortion produced in the mains cable will be very small compared to the junk already present on the mains supply. If the junk doesn't hurt us too much then the distortion won't either. We really must stop straining at very expensive gnats. If people want to sell us special high-quality gnats then they should do us the courtesy of giving a plausible explanation based on accepted science, backed up by appropriate measurements. My view is that cable fussiness is a sign of poor equipment, not fine equipment.
Hi,
Belief does not come into it and the aggressively dismissive attitude exhibited by some here are probably the single biggest effort in support of the "voodoo cable" industry going, as they effectively blackball the subject for serious investigation.
The facts are that as long as gear is mains powered you will have issues.
If an AP2 was designed in terms of powersupplies, inputs and outputs the way
most HiFi is, it would not be able to measure 16 Bit audio with any accuracy!
While I am much in favour of a fair deal for Ostriches (my party is lobbying for 1/3 of all legistlative bodies to be made up by Ostriches, for the awesome dignity they will thus lend to the proceedings), behaving the way that is falsely alleged for ostriches (sticking ones head in the sand) will not do.
In fact, there is another act of double-think - applauding the design of the AP2 and almost worshiping it's output graphs and completly ignoring why it had to be like that.
As Sy would point out, an amp has no sound... 😉
You may have to remove all the other compensation and still have a minimal miller compensation. See how it works out.
Ciao T
Belief does not come into it and the aggressively dismissive attitude exhibited by some here are probably the single biggest effort in support of the "voodoo cable" industry going, as they effectively blackball the subject for serious investigation.
The facts are that as long as gear is mains powered you will have issues.
If an AP2 was designed in terms of powersupplies, inputs and outputs the way
most HiFi is, it would not be able to measure 16 Bit audio with any accuracy!
While I am much in favour of a fair deal for Ostriches (my party is lobbying for 1/3 of all legistlative bodies to be made up by Ostriches, for the awesome dignity they will thus lend to the proceedings), behaving the way that is falsely alleged for ostriches (sticking ones head in the sand) will not do.
In fact, there is another act of double-think - applauding the design of the AP2 and almost worshiping it's output graphs and completly ignoring why it had to be like that.
As Sy would point out, an amp has no sound... 😉
You may have to remove all the other compensation and still have a minimal miller compensation. See how it works out.
Ciao T
I honestly, REALLY don't want to get into another pointess cable debate, but even as a sceptic, I have to admit that truly not all cables sound the same. On the other hand, since I have lost faith in Mother Goose, Santa Claus and myth and magic, I do believe that if one tries hard enough, one CAN measure the differences. I seem to remember a rather educational text on this subject in UK's Hi Fi News & Record Review around the late 90ies by one Ben Duncan.
Ultimately, different materials conduct differently.
What I'd like to point out is that perhaps even nominally same cables COULD behave differently in a circuit, including the mains power cable. I was told by a friend who sells electrical hardware that Chinese cables had appeared and some of them were tested and found to be substandard regarding metal purity to our usual Belden type cable, for example. He said their R and C values were quite a bit above what we take as standard, but that it was much cheaper than the same standard cable.
An old story, no doubt, a fair share of Chinese price advantage comes from complete disregard for technical standards to which we are used to and take for granted. Can anyone say with absolute certainty that such a cable cannot make any difference under any circumstances? And without any knowledge of what that power cable is attached to on the device end?
Ultimately, different materials conduct differently.
What I'd like to point out is that perhaps even nominally same cables COULD behave differently in a circuit, including the mains power cable. I was told by a friend who sells electrical hardware that Chinese cables had appeared and some of them were tested and found to be substandard regarding metal purity to our usual Belden type cable, for example. He said their R and C values were quite a bit above what we take as standard, but that it was much cheaper than the same standard cable.
An old story, no doubt, a fair share of Chinese price advantage comes from complete disregard for technical standards to which we are used to and take for granted. Can anyone say with absolute certainty that such a cable cannot make any difference under any circumstances? And without any knowledge of what that power cable is attached to on the device end?
Last edited:
I sure wish that I found amps to sound all the same, even my own designs. Life would be perfect, and I could live off my royalties long into the future. '-)
A fair and interesting response to the original article showing the scientific method occasionally surfaces.
However my stand is pretty much that with today's easy access to test equipment and the quality of it, that you can measure difference in just about anything. The issue is at what level of difference is it significant? I just saw a documentary where they claim tasting of some molecules can detect 1 part per billion.
I can certainly measure differences in cables, that doesn't mean you can hear them. It gets a bit more interesting when the measured differences get to where folks say they can hear them. (A level well above the measurement threshold.)
Now power cords and line filters are a different issue. The typical sound system component that puts the most noise back into the AC power line is the CD player. (Duh!) ((As I measure it.))
There are reasonably well documented methods for power line noise to get into the signal path. If that noise can be heard by most folks then we would consider the design to be not very good. A valid issue is at what level the noise or the modulation or other effects of the noise cannot be heard by the outriders of the hearing or perception curves.
- Status
- Not open for further replies.