After the last burst of funny cables stuff (is that something like Gorillagrams ...?) I'll go back to sensible ... perhaps. My aim, as a consumer of music with absolutely no pretensions to having professional knowledge of the ins and outs of recording studios, or recording techniques, is to enjoy whatever music is served up to me, by being able to connect, emotionally, to the musicians doing their thing. And I don't achieve that if the sound is harsh and hyper outlined, or if maple syrup has been poured over it.
Part of the "problem" of refining playback is that the better it gets the more you're aware of issues unresolved. Like driving a noisy car: you fix the muffler and you realise the the suspension is squeaking. Lubricate that, and you become aware of a rattle in the dashboard. Each step of quietening the vehicle allows you to hear more into,
, the sounds of the car and the last little scratchy sound will annoy the bejeezus out of you.The cable "problem" is one of those, so, hey, if one doesn't like to worry about such things then just wind down the windows, the wind noise will make it go away ...
Frank
Part of the "problem" of refining playback is that the better it gets the more you're aware of issues unresolved. Like driving a noisy car: you fix the muffler and you realise the the suspension is squeaking. Lubricate that, and you become aware of a rattle in the dashboard. Each step of quietening the vehicle allows you to hear more into,
Yes, I've noticed this effect too - I call it 'the onion' effect. Peel off one layer and lo and behold, another layer beneath appears. I think it has much to do with how perception works - our brains are creating the perceived sound, quite literally making sense of what the vibrations at our ear drums are indicating. I rather like the analogy from the book 'This is your brain on music' where Dan Levitin talks about ping-pong balls. The better the system is the less our unconscious perception has to work to create sensation out of those myriad bouncing pinig-pong balls. A better system means less mental interpolation but we're not quite aware just where the brain is interpolating and where it has the real data.
Nicely put. What I've found is that you have get to the point where that interpolation is totally unconscious, where no apparent effort is required. A quick way to "test" this is to switch your attention entirely to something else, like engaging in intense conversation with the person next to you. If the sound of the music continues to wash over you, with zero mental discomfort, then you're in pretty good shape ...
Ultimately, it doesn't matter whether the data is real or interpolation is happening, I believe; if your "hearing" side is convinced then that's all that matters. The extra step I perhaps take, compared to others, is that that level of "convincingness" has to occur for the "worst" recordings ...
Frank
Last edited:
Another point I would make: people talk of the volume of a system overloading a room. Perhaps my hearing is different from others but I don't believe that happens when a system is good enough. If a full, real drumkit is placed in to a reasonably lively, small room and you got a good drummer to go for it, the intensity of sound and the reflections would make the experience overwhelming. Yet my brain would soak it up, make sense of it all, as a powerful, captivating moment. Of course, my ear drums would rapidly start to protest, and I would need to withdraw before permanent damage occurred; but I wouldn't have complained that the room "overloaded".
So, as far as I'm concerned, a good system should be able to similarly perform. Not saying that that is what I have now, but I see absolutely no reason why that level of performance can't be achieved ...
Frank
So, as far as I'm concerned, a good system should be able to similarly perform. Not saying that that is what I have now, but I see absolutely no reason why that level of performance can't be achieved ...
Frank
I agree on this. But genuine experience is not transferable, especially not through web forum. You would hardly explain this to someone who has troubles with aggressive and disintegrating sound at high levels. And it is not a question of recordings.
If a static charge on the outside of cable insulation can affect the operating point of some part of the circuitry then you have big problems with your cables and your circuitry.fas42 said:
Except where there is significant movement, static tends to reduce not build up. Cars and people build up static charges by moving from here to there. Cables just sit there so a build up of static is unlikely.
However, let's assume for the moment that by some strange mechanism a build up of static has occured. It is on the other side of an insulator from the conductors so it can't affect circuit operating points. If any current flows then the static gets discharged. The only likely effect is to bias the insulation. If the insulation is a nonlinear dielectric then it could increase distortion if the source impedance is way too high. (I note that some people seem to want to deliberately introduce this problem by voltage biassing their cables.) The solution is simple: make the insulation slightly conducting so no charge builds up. In most cases a simple build-up of normal household grime will be sufficient, especially in the damper parts of the world. Avoid very high quality cable insulation.
So the solution to this problem, if it even exists, is cheap cables with poor insulation and cheap sources with a very low output impedance (such as a bog-standard opamp).
Yes, a cheap opamp with a series resistor. A resistor is better than an inductor in this position because it is much cheaper, and can't be tuned out by capacitance so it simply damps things down. An inductor would need resistors too anyway. People would still use a resistor even if it were more expensive than an inductor.
Agree that the solutions are relatively straightforward. But, unless the user of the system is aware of this as a possible problem he won't contemplate trying one of them.
What precisely occurs I can't say, I haven't the knowledge. But through trial and error, and constant experimenting, I know that for at least some systems it makes a very significant difference, if I do everything I can to eliminate this factor.
Edit: I'll just repeat, I can hear this problem on the majority of systems; most people are just used to systems sounding a certain way -- "That's a hifi ... all hifis have that certain quality to the sound ... that's just the way it is ... etc ..."
Frank
Last edited:
You are of course, incorrect.
There are 4 very important things you must remember.
1. There are no lumped element models presented which consider the speed of propagation along the cable. At the microsecond level where humans ability to discern interchannel temporal differences, this can not be ignored.
2. A load at the far end of a cable which has a V/I characteristic which is not identical to that cable, will require a different current than that which the cable will propagate at it's characteristic velocity. The RLCG model assumes instantaneous communication.
3. Measurement of these delays within the audio band is extremely difficult given the overall lack of ability of most to understand and implement a wideband test setup with the necessary capabilities, as well as a lack of adequate hardware.
4. Discussion with you regarding our different thinking on this topic has been nothing but pleasure for me. I do not care if you are wrong, or if I am wrong. This discussion is the way it should be. I thank you for this.
jn
I was speaking about more esoteric pseudophysics explanations.
Always. It's a perk of the job.
Hmm..please go back and read what I wrote.. I repeat....
Again, has anybody ever measured a 10 uSec delay on a 4 ohm load between a 20, 50, 1000, 10,000 hz signal at power? Not simulation. Measurement.
How long did it take in the dark room to see the traces?
What I like about the mercury wetted reed is it's by nature a bipolar device. I can easily change polarity just by changing the supply voltage presneted to the charge line. And the resulting output impedance is exactly that of the charging line impedance.
jn
I am not proposing a lumped model, but the low frequency transmission line model. This is the appropriate model for audio. This (the 'RLCG model') does not assume instantaneous transmission. On the contrary, it gives the behaviour of the line in terms of (complex and frequency-dependent) characteristic impedance and (complex) propagation parameters (speed and attenuation). The RF model you are using is a high frequency approximation; the LF model is exact. It can be used to design long distance audio transmission lines, so a few metres to a speaker is trivial.
Given the (frequency-dependent) load and source impedances and the (frequency-dependent) line parameters everything we are interested in can be calculated by known transmission line theory. Note that many textbooks just give the RF approximation, not the full theory.
Having said that, I believe the lumped model is good enough for most cases. I only bring up the LF line model as a counter to your impulse model.
I have not proposed measuring these delays. I am not interested in these delays. They will, broadly, be the same for each channel and insignificant when compared with what happens in the room.
You assert there are delays - I don't dispute this. You calculate these delays by using an RF impulse method when the full LF transmission line model would, in my opinion, be better. You assert that these delays matter - I and others doubt this.
We have been round this circle before so in future, unless one of us comes up with a better argument, could we just note that we differ? The only reason I butted in this time is that I didn't want other people to assume that your (otherwise unchallenged) claims are generally accepted.
Given the (frequency-dependent) load and source impedances and the (frequency-dependent) line parameters everything we are interested in can be calculated by known transmission line theory. Note that many textbooks just give the RF approximation, not the full theory.
Having said that, I believe the lumped model is good enough for most cases. I only bring up the LF line model as a counter to your impulse model.
I have not proposed measuring these delays. I am not interested in these delays. They will, broadly, be the same for each channel and insignificant when compared with what happens in the room.
You assert there are delays - I don't dispute this. You calculate these delays by using an RF impulse method when the full LF transmission line model would, in my opinion, be better. You assert that these delays matter - I and others doubt this.
We have been round this circle before so in future, unless one of us comes up with a better argument, could we just note that we differ? The only reason I butted in this time is that I didn't want other people to assume that your (otherwise unchallenged) claims are generally accepted.
Help! I found an old notebook with some of my stuff that has been stored for at least 25 years. It goes back to 1959 for the first entry, so it is interesting (to me).
However, this page has an entry that must be between 45 and 50 years old, but I can't exactly place its origin. Can any of you math wiz's help?
However, this page has an entry that must be between 45 and 50 years old, but I can't exactly place its origin. Can any of you math wiz's help?
Attachments
Trivial when the load at the other end is exactly compatible with the line. Not so when we require the line support a current/voltage characteristic which is not what the line is designed to support.
That is where the scientist in you gets in the way. You have adopted the assumption that interchannel delays are unimportant.
That is the wide chasm between you and the high end audio guys you are constantly at odds with. When they hear a system difference as a result of some error in ITD, you by definition dismiss it based on a lack of knowledge on what humans discern.
I do not casually dismiss what humans can discern. I recommed that approach for you as well.
Perhap you should do some reading. I recommend as a start, Nordmark. It is important you get an idea what humans can discern and what they cannot.
I have shown that the delays rise into the level of human discernment so cannot scientifically dismiss them. Had they not risen to a level where humans have demonstrated repeatable testable duplicated sensitivity, we would not be having this discussion.
I would not post on a forum where my statements are accepted without challenge.
I heartily recommend you learn what humans are capable of discerning.
jn
Richard,
The power amps sound like they were not released as " Marsh" commercial products. Basic topology jfet input, bipolar VAS and output?
You realize of course answering this with any information acknowledging their existence will result in you being badgered incessantly for more information, parts list, board layout, hand holding, baby sitting, and general DIY whining.......
The power amps sound like they were not released as " Marsh" commercial products. Basic topology jfet input, bipolar VAS and output?
You realize of course answering this with any information acknowledging their existence will result in you being badgered incessantly for more information, parts list, board layout, hand holding, baby sitting, and general DIY whining.......
Last edited:
- Status
- Not open for further replies.