I guess that is the difference between reproduction and instruments, and perhaps between hi-fi and audiophilia.andyjevans said:
There may be a difference here between how scientists and engineers (and laymen) use the word 'theory'. To a scientist a theory is a useful model of how reality works; he expects it to be 'true' in some sense. To an engineer a theory may be a way iof calculating something, given certain (often unspecified or even unknown) approximations. When theory and practice differ the scientist looks for a better theory; the engineer may just shrug his shoulders and say that he never trusted theory anyway.
It is something more than this. Unfortunately physics cannot describe the musical sound (i.e. what you really perceive and associate with that sound). But even assuming absurdly this were possible I really wouldn't see the point in measuring the musical sound other than using ears and brain as this has always been the way used for creating music, instruments and places where you play music. Back to real world, physics can only describe the physical sound under invariant conditions. When you have a listener in the system you are no more in invariant conditions both in the short and long term. You change something, even a little detail, and you cannot tell the consequences until you listen regardless of the possible measurements....
For me measurements are only a gross way to check that things work fine ( they work like I wish) and I have not made trivial mistakes.
Last edited:
The time constant and -3db point are directly related. For example, if you halve the capacitor value, the time constant is halved and the -3dB frequency is doubled.
As a rule of thumb, you want the -3dB frequency as low as possible, to minimize the effects in the audio band. Bear in mind it's a gradual rolloff. For example, if you set the -3dB point to 10Hz then there will be 3dB attenuation at 10Hz. However there will also be about 1dB attenuation at 20Hz, 0.3dB at 40Hz etc.
The effect is also cumulative, so if your source, preamp and power amp all have low frequency rolloffs with the -3dB points at 10Hz, then the system as a whole will be -3dB at 20Hz and -1dB at 40Hz
P.S. Interesting to see the results of your blindish listening tests. I'm surprised the polyester one fared so badly. Did you pursue that result any further, as you did with the good-sounding electrolytic?
Last edited:
This is in reply to all the previous replies that said no or I don't think so or some similar variation.
At some point you have to connect to the next stage. OK the load may be high (470K say) but it is a lot less than a CCS provides.
The effect is real, especially if you build a regular CC circuit. Measure the distortion at a given output level with and without a cathode bypass resistor and the distortion invariably increases with the bypass removed. I have tested this many times and it is always so. The increase is not large but it is always there.
The precise cause I agree is open to debate but of all the triode parameters the one most largely determined by the physical construction of the tube is mu and varies the least with operating point. Minimising distortion by minimising the load on ra or by minimising gm variations are effectively the same thing and a CCS or mu follower do both to a greater or lesser extent.
Wire up a few triodes in CC configuration and check it out for yourselves.
Cheers
ian
“If the correct capacitor is in there, then it should be impossible to hear the difference when another correct capacitor is swapped in.”
So what? This hasn’t answered my question – a correct capacitor is “inaudible” compared to WHAT? You still have a capacitor in the circuit. If you take it out you have just a resistor which is a different set of parameters. You still have to compare both to different ways of biasing. All forms of biasing will have one effect or another. If you are actually saying "unmeasurable" that's not the same thing as "inaudible".
All this makes me think of the old Whiskas ad: "eight out of ten owners said their cat prefers it". Prefers it to what? Chewing a lump of concrete? What are we comparing here?
"When theory and practice differ the scientist looks for a better theory; the engineer may just shrug his shoulders and say that he never trusted theory anyway".
I’m a psychologist, and it’s not only scientists who look for better theories – so do police detectives, stand-up comedians, classical composers or just human beings in general. So I think this is unnecessarily reductionist.
“Bald assertion, anecdote, claims that something has to be true because so many people believe it”
So “people fall in love”, “children fear dogs” and such is just anecdotal because really there’s no evidence so it’s just a bald assertion? You’re welcome to go down that road but others simply won’t ignore empirical findings, and will regard words like “opinion” and “anecdote” as patronising when they would use the word “observation”.
I’m not happy with any of the above because I believe there’s a different explanation for this phenomena. I also believe that since nobody is really happy with extreme views on this matter it keeps coming up in arguments, so we may be missing another dimension.
Here’s what I think happens. The human brain is pretty good at perception. The retina communicates to brain at 10 million bits per second (Tufte et al). Our auditive system is rather slower, but the real bottleneck is the actual processing power of the brain. The processing power of the brain has been estimated as 126 bits per second (Miller, 1956). It’s easy to see that what the brain does in practice is to ignore an enormous amount of incoming information and select what it wants to attend to. This explains eyewitness divergence and the whole phenomenon that “people see and hear different things”. Actually they see and hear the same things but their SELECTION of what to attend to and process is different. So what they report as their “experience” of perception is very different.
Think of a photograph. That’s accurate – all the details are faithfully reproduced. The analogy would be a sound system with low distortion. But that actually doesn’t satisfy the narrow window of what people WANT to see. We might call this the “essence” of what is perceived – the reduction into the most important features the brain wants to pick up. So photos routinely get processed to enhance the “essence” – the greens are greener, the sky is bluer.
“Distortion” you will say. But our brains are evolutionarily programmed to select and enhance certain cues in our environment. Some of these are important emotive cues like faces, voices and sounds. Look at how many audiophiles use SETs for the human voice “because it sounds more real”. They are picking up more of this “essence”. Consider how a caricature of a face may pick up more of the “essence” of a personality than a photo – people say “that’s just like him” while the photo “doesn’t really look like him”.
So I think it’s really unfair to criticise our human brain for wanting to select the “essence” of what we hear. This human brain may not be interested in a full frequency response to within 1 db. It may want to hear the woodiness of a clarinet, a bass that makes the body move and dance and so on.
So rather than wasting a lot of time and energy in theorising about what the brain “ought” to be perceiving we should give more consideration to the way our human brain actually works, and how our evolution has programmed us to pick up, select and enhance perceptions that are important to us in our lives.
So what? This hasn’t answered my question – a correct capacitor is “inaudible” compared to WHAT? You still have a capacitor in the circuit. If you take it out you have just a resistor which is a different set of parameters. You still have to compare both to different ways of biasing. All forms of biasing will have one effect or another. If you are actually saying "unmeasurable" that's not the same thing as "inaudible".
All this makes me think of the old Whiskas ad: "eight out of ten owners said their cat prefers it". Prefers it to what? Chewing a lump of concrete? What are we comparing here?
"When theory and practice differ the scientist looks for a better theory; the engineer may just shrug his shoulders and say that he never trusted theory anyway".
I’m a psychologist, and it’s not only scientists who look for better theories – so do police detectives, stand-up comedians, classical composers or just human beings in general. So I think this is unnecessarily reductionist.
“Bald assertion, anecdote, claims that something has to be true because so many people believe it”
So “people fall in love”, “children fear dogs” and such is just anecdotal because really there’s no evidence so it’s just a bald assertion? You’re welcome to go down that road but others simply won’t ignore empirical findings, and will regard words like “opinion” and “anecdote” as patronising when they would use the word “observation”.
I’m not happy with any of the above because I believe there’s a different explanation for this phenomena. I also believe that since nobody is really happy with extreme views on this matter it keeps coming up in arguments, so we may be missing another dimension.
Here’s what I think happens. The human brain is pretty good at perception. The retina communicates to brain at 10 million bits per second (Tufte et al). Our auditive system is rather slower, but the real bottleneck is the actual processing power of the brain. The processing power of the brain has been estimated as 126 bits per second (Miller, 1956). It’s easy to see that what the brain does in practice is to ignore an enormous amount of incoming information and select what it wants to attend to. This explains eyewitness divergence and the whole phenomenon that “people see and hear different things”. Actually they see and hear the same things but their SELECTION of what to attend to and process is different. So what they report as their “experience” of perception is very different.
Think of a photograph. That’s accurate – all the details are faithfully reproduced. The analogy would be a sound system with low distortion. But that actually doesn’t satisfy the narrow window of what people WANT to see. We might call this the “essence” of what is perceived – the reduction into the most important features the brain wants to pick up. So photos routinely get processed to enhance the “essence” – the greens are greener, the sky is bluer.
“Distortion” you will say. But our brains are evolutionarily programmed to select and enhance certain cues in our environment. Some of these are important emotive cues like faces, voices and sounds. Look at how many audiophiles use SETs for the human voice “because it sounds more real”. They are picking up more of this “essence”. Consider how a caricature of a face may pick up more of the “essence” of a personality than a photo – people say “that’s just like him” while the photo “doesn’t really look like him”.
So I think it’s really unfair to criticise our human brain for wanting to select the “essence” of what we hear. This human brain may not be interested in a full frequency response to within 1 db. It may want to hear the woodiness of a clarinet, a bass that makes the body move and dance and so on.
So rather than wasting a lot of time and energy in theorising about what the brain “ought” to be perceiving we should give more consideration to the way our human brain actually works, and how our evolution has programmed us to pick up, select and enhance perceptions that are important to us in our lives.
Last edited:
But you still have to set up a reproduction system for people to listen to, and that will have inherent colorations. And you have to select material to listen to and you still have to select listeners and so on and so on. What do you mean by "ears-only conditions" - blind A-B testing?
Rather than regurgitating all this old stuff, what do you think of the proposition that we should respect the way our brain actually works in evolutionary terms? Could you comment on that?
The topic here is electrical circuits. You can either hear the effect of the circuit under test (which is not hifi) or you can't (which is). If you engineer a circuit properly with hifi in mind, you won't be able to hear its effect. If you're after an effects box, that's a different issue, and the sorts of sensory testing that one needs to do to judge the success is a bit more complicated.
I don't like processed photos. On my TV the colour control is turned down lower than most people have it, so the colours look a bit more realistic. I don't like digital artifacts on TV or photos; I guess I find them easy to see as I know what to look for. A little background noise would be preferable to artifacts.
I like carrots to taste like carrots, and potatoes to taste like potatoes etc. I like women to look like women (not men or Barbie dolls).
I take the same view for sound. Others may not. That is up to them. The task of an amplifier, in my view, is to amplify a voltage signal and add some power to drive a speaker. If I want effects I will add an FX box.
So the only issue for me is whether a cathode bypass capacitor aids or disrupts the process of faithful amplification. I believe circuit theory is the main thing I should be thinking about in this context; add to that component quality. Sorry if that seems boring or reductionist.
I like carrots to taste like carrots, and potatoes to taste like potatoes etc. I like women to look like women (not men or Barbie dolls).
I take the same view for sound. Others may not. That is up to them. The task of an amplifier, in my view, is to amplify a voltage signal and add some power to drive a speaker. If I want effects I will add an FX box.
So the only issue for me is whether a cathode bypass capacitor aids or disrupts the process of faithful amplification. I believe circuit theory is the main thing I should be thinking about in this context; add to that component quality. Sorry if that seems boring or reductionist.
I said no because I have measured it as well. That's why it depends on the application and on the actual device. In particular at low output level irreducible distortion makes the difference and this depends mostly on the quality of the actual valve you are using. However THD alone doesn't tell all the story. Its impact on the sound is actually overrated in many cases. I am not saying that it doesn't matter but just that when it is low enough other things become more important.
Recently I made a quasi-replica of a Mullard circuit with PCL82's and after trying all possible things and tweaks I and all the others listening to it had to accept the fact that 1970's NOS Orion-branded PCL82's (printed "made in USA" on the glass) sounded far better than more recently made Eastern European Tungsram's. Difference in distortion was really a matter of hair splitting up to the clipping. It was really low even without feedback and nearly identical. Guaranteed.
Hello Sy,
I don't mean to be confrontational here, because I have huge respect for your work and posts, but I don't think electrical circuits ARE "the" topic here. They are part of the totality, and they are the topic you have chosen to address. But unless you just measure stuff in labs, you have to take into account the listening experience and how the human brain works.
Good heavens no - your posts are never boring! And actually when you say "I like carrots to taste like carrots - I take the same view for sound" I COMPLETELY agree with you. But I agree with you for different reasons. I'm an ex-professional musician, and the overwhelmingly important thing for me, and how I design and build everything, is that I want flutes to sound like flutes, clarinets like clarinets, crash cymbols like crash cymbols, Steinway pianos to sound like Steinways and not Bechsteins and so on. I grew up playing a Bechstein grand and I played right alongside a variety of musical instruments for half my life, and these sounds are indelibly imprinted on my brain. My whole pleasure in listening is to feel "wow - that rim shot on the snare drum is spookily like it is in my head".
But I keep coming back to the point of human perception. If in terms of human perception a duck "quacks like a duck" then it's a duck. It may not measure like a duck but if its quack is more duck-like it's more of a duck. I've referred in my previous post to perception and processing in the human brain. This isn't woolly anecdotal stuff, it's contemporary neuroscience.
Last edited:
Of course, and I've done so again and again; I get the feeling that many human brain functions are not appreciated by those who prefer faith-based approaches to engineering. Perhaps it's my fault that I haven't been clear enough about validation through ears-only listening tests when an effect not explained by simple measurements is believed to be heard?
As an engineer (in another field) its pretty obvious that we hobbyist aren't designing circuits for studio monitoring, good engineers know their design goal, to reproduce the live art. Calling this an effects box is an oversimplification IMHO. A typical commercial recording goes thru so many "effects boxes" before massed produced as a CD it is laughable to think they get it right all the time. Unless you limit yourself to a few "perfect" recordings, the DAC and amp at home require finesse.
I really want to see the electrical engineers measurements of a live performance, we hear a lot about measuring voltage across a load but that is about as far as the EE's go, sure we have FFT to make things visually organized, but how about mikes in a manikin head dancing to a performance? That has only been has barely been touched on.
Perception tough for engineers, "effects" are an engineering requirement if we want our ear to come close to hearing live music from a pair of home transducers. Blindly relying on a 1000 different studio geeks to "get it" right is an exercise in futility much worse than amplifier creativity. Good amplification is still more art than engineering, that's what makes it interesting.
If by "effects box" perhaps you mean we should use an at home DAW to remaster our recordings, I would be all for it if we had recordings with dynamics left and all 16-24 tacks, but we don't.
We need things like tube finesse, and just one skill in that is eliminating electrolytic capacitors as much as possible. I built a spud amp with no electrolytics in the signal current loop, it is amazing, even measures well.
The only situation I can see the need for an "instrumentation" home playback is if one is mixing, mastering his own recordings, beyond that we need skillful creative playback designs.
Neuroscience has to challenge the most fundamental philosophical foundations of science because it deals with consciousness, this is why engineers have such a hard time with it. Its very foreign for an engineer to think of perception as a hard science rather than psychobabble. There are so many variables much easier to measure voltage across a load, make FFTs, plots and call it a day.
So all these various studio FX can be magically removed by using tubes and avoiding electrolytics at home? I didn't know it was that easy; I needn't have spent 45 years learning electronics.
No and no. Exceptions are 'high-end' people who need a good story to extract money from the gullible, and DIYers who seem to imagine that success in a pursuit can be separated from appropriate knowledge and understanding.regal said:
I suppose it's fortunate, then, that I'm not an engineer. 20-odd years doing sensory science as part of my job was a real eye-opener, so to speak. In any case, this has little-to-nothing to do with cathode bypass capacitors, so let's try to stick to that.
We're not talking here about con artists who drive Maseratis, but we are talking about DIY. Success in an amplification pursuit is partly or even largely dependent on "knowledge and understanding", by which I assume you mean knowledge and understanding of engineering. But it also involves other human factors, such as the quality of choices made, the ability to judge the results aesthetically, the nature of perception and its processing and the quality and extent of creativity applied to any task. It doesn't matter whether you refer to these as art or sociology or green cheese as long as you take them into consideration. If you are eliminating all these factors we're seriously not on the same page.
Well, it has everything to do with whether you can hear them or not and whether you want to use them or eliminate them. But if you start from the point that you do want to use a bypass capacitor, then no it's not relevant because you just want to know what value and type to use. But even then somebody is going to say "polypropylenes sound better than electrolytics"..... It's a cat that doesn't want to go back in the bag without a good sniff around.
Being creative is not an alternative to engineering; creativity is an essential part of engineering. I deplore the way that some arty people (often the least talented ones, too!) have tried to claim the word exclusively for themselves. My TV set lives in mortal danger of assault every time a news report talks about 'the creative industries' as shorthand for pop music and fashion.andyjevans said:
Taken in isolation that statement, which may be true, says nothing about sound reproduction. It could be that PPs sound better because they distort less; it could be that they sound better because they distort more and some people prefer that.andyjevans said:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.