Thanks Wahab.
You say that the less THD the better, for 'true hifi reproduction', which is probably a purely technical notion without any psychoacoustic implications.
I wasn't actually trying to be offensive. No problem, we agree to disagree. But I agree emphatically about stability; many amp faults are instability and other transient misbehaviour.
I have heard the Aphex, it doesn't really impress, or the ES8 for that matter.
Thanks for replying,
Hugh
You say that the less THD the better, for 'true hifi reproduction', which is probably a purely technical notion without any psychoacoustic implications.
I wasn't actually trying to be offensive. No problem, we agree to disagree. But I agree emphatically about stability; many amp faults are instability and other transient misbehaviour.
I have heard the Aphex, it doesn't really impress, or the ES8 for that matter.
Thanks for replying,
Hugh
what is HiFi?
Does it only sound good?
Does it try to reproduce the sound of the original (High Fidelity)?
Can it both reproduce accurately and sound good?
Define true hifi.
What it certainly is not are the systems that most of us refer as "hifi".
The signal measured close to the speakers must be as much
as possible identical to the one entering the amp.
Speakers distort a lot, but using a Set amp, you will measure
more THD than when using a SS amp, even when using your
ears as sole measurement instrument...
my definition must exclude intermediate measurements.
One can conditionally bring in intermediate measurements, but overall they must not determine Lower Fidelity.
High Fidelity, to me, means the reproduced sound at my ears must be as close as possible the same as that sound arriving at a well positioned microphone/s.
You essentially say the same as in my post, set apart that
i gave the precision : "close to the speakers." , since it remove
the room influence..
Of course that measurement close to the speakers implicitly
say that micros are used for the said measurement..
I took this to mean measuring the electrical signal leaving the amplifier and arriving at the speakers must be
Since i m not a supporter of the heavy cable influence dogma,
it could only be the acoustic signal stage i was refering to.
I can't say that I neccessarily agree with everything Self has written but at least he is very clear about his design philosophy that's seems to be based on things we can actually measure. And he can present plenty of research and measurements to back up his arguments.
So, in most cases I rather agree with him than blindly trust people who design "by ear" and simply disregard all the measurements because "they don't tell everything". Which, is a philosophy that basically gives nothing concrete in the end.
Bob Cordell's remarks about some stuff he disagrees with are pretty much on point as well as the comment that Self has also ignored a lot of alternative ways and circuits that also make up an amplifier. That latter reason may also explain why Self may have had different results with different designs and components than some others. Self's books / work are not Gospel but they are extremely good references.
So, in most cases I rather agree with him than blindly trust people who design "by ear" and simply disregard all the measurements because "they don't tell everything". Which, is a philosophy that basically gives nothing concrete in the end.
Bob Cordell's remarks about some stuff he disagrees with are pretty much on point as well as the comment that Self has also ignored a lot of alternative ways and circuits that also make up an amplifier. That latter reason may also explain why Self may have had different results with different designs and components than some others. Self's books / work are not Gospel but they are extremely good references.
Hi,
very much so.
He shows how and why an amplifier can be made to perform well in terms of distortion.
Without using his techniques at reducing voltage amplifier distortions, do you think he could measure the capacitor distortion of a slightly too small NFB cap with any certainty?
Similarly, could supply line induced distortions have been tracked down and a scheme to eliminate them proposed with some certainty?
Make the amplifier "Blameless" and you get the opportunity to investigate other inaccuracies.
We have a lot to thank D.Self for.
So we agree that "hi-fi" means nothing added, nothing taken away. This must mean a straight wire with gain, except possibly for some compensation for transducer problems.
Some audiophiles find this boring and clinical, so they redefine "hi-fi" to mean whatever they happen to prefer yet still pretend that they are using the original meaning. So 'distorted' becomes the new 'linear', 'bent' becomes the new 'straight', 'lop-sided' becomes the new 'balanced'. When pressed, some of them even launch into an ignorant post-modern criticism of science itself. Self may overstate his case, but strong evidence for the objectivist view can be found on these very forums when you see the number of people who pop up and claim to have improved the sound of their SS system by adding a valve cathode follower running from a 12V HT supply. The rich but stupid can even buy such nonsense ready-made, complete with arty blurb.
By all means listen to the sound you prefer, just don't try to pretend that it is better than accuracy or, even worse, deny the concepts of 'better' or 'accurate'. I am too old for all that post-modern claptrap!
Some audiophiles find this boring and clinical, so they redefine "hi-fi" to mean whatever they happen to prefer yet still pretend that they are using the original meaning. So 'distorted' becomes the new 'linear', 'bent' becomes the new 'straight', 'lop-sided' becomes the new 'balanced'. When pressed, some of them even launch into an ignorant post-modern criticism of science itself. Self may overstate his case, but strong evidence for the objectivist view can be found on these very forums when you see the number of people who pop up and claim to have improved the sound of their SS system by adding a valve cathode follower running from a 12V HT supply. The rich but stupid can even buy such nonsense ready-made, complete with arty blurb.
By all means listen to the sound you prefer, just don't try to pretend that it is better than accuracy or, even worse, deny the concepts of 'better' or 'accurate'. I am too old for all that post-modern claptrap!
It should be clear in most older guys minds at least, That Hi-Fi has been used consistently since the 1950s to describe audio systems complete or in part that are intended to reproduce audio with high fidelity. e.g. Hi-Fi cabinet, Hi-Fi speakers.
In recent years, the term has certainly shifted, as the priority in typical affluent cultures has shifted, I believe, to increase traditional stereophonic to multichannel home theatre installations.
Here in Oz, I regularly hear the term "Hi-Fi" to now mean simply " Stereophonic Audio". The newer term "Hi-end" to imply expensive and quirky though not necessarily good electronic entertainment equipment. Also the term is applied in replacement for the now degraded "Hi-Fi".
I'll venture that simililar abused terminology exists elsewhere (asbestos suit on!)
Hence, I see no point in arguing over where in the audio system "High Fidelity" ends or begins. It is really a generic package of devices, each classed, somewhat by industry consensus and now by Forum consensus as "Hi-Fi" or not in accordance with our arbitrary standards for individual devices. (asbestos suit on!)
In recent years, the term has certainly shifted, as the priority in typical affluent cultures has shifted, I believe, to increase traditional stereophonic to multichannel home theatre installations.
Here in Oz, I regularly hear the term "Hi-Fi" to now mean simply " Stereophonic Audio". The newer term "Hi-end" to imply expensive and quirky though not necessarily good electronic entertainment equipment. Also the term is applied in replacement for the now degraded "Hi-Fi".
I'll venture that simililar abused terminology exists elsewhere (asbestos suit on!)
Hence, I see no point in arguing over where in the audio system "High Fidelity" ends or begins. It is really a generic package of devices, each classed, somewhat by industry consensus and now by Forum consensus as "Hi-Fi" or not in accordance with our arbitrary standards for individual devices. (asbestos suit on!)
Even Doug himself in the original Wireless World articles mentioned tongue in cheek the idea of a "niceness control" in the form of signal processing before the (blameless) amp to add distortion and "colour".
If that approach could be realised and made to work... it is an intriguing thought.
If that approach could be realised and made to work... it is an intriguing thought.
I have been waiting 10 years for Doug to design one and still no dice, Mooly.
Wanna Try? - I'll buy a PCB at least!
Wanna Try? - I'll buy a PCB at least!
I always hate it when nice terms/words get hijacked. Political correctness and UK equality laws prevent me from mentioning one such example.
It is unfortunate that almost anything which plays CDs through two speakers is now called a hi-fi. What should we now call something which has the same aims as what we used to call hi-fi? We can't call it hi-end, because much of that is just expensive pretty effects boxes. My preference is to stick with hi-fi, and say of some cheaper CD playing equipment and some hi-end "That is not hi-fi!".
If you want a niceness control, see the Harmonizer thread in the Tube section.
It is unfortunate that almost anything which plays CDs through two speakers is now called a hi-fi. What should we now call something which has the same aims as what we used to call hi-fi? We can't call it hi-end, because much of that is just expensive pretty effects boxes. My preference is to stick with hi-fi, and say of some cheaper CD playing equipment and some hi-end "That is not hi-fi!".
If you want a niceness control, see the Harmonizer thread in the Tube section.
Hey, if we can just decide what "niceness" is, building a circuit to do it and attaching a knob shouldn't be that hard. My computer sound card has all sorts of effects like "concert hall" and "bathroom" that are all bad. Still, a bit of response shaping plus a subtle amount of reverb can be a wonderful thing. Maybe the box needs to preview the music and choose settings based on characteristics of what it measures. Think about a modern dSLR camera. Some have extensive lookup tables of scenes (thousands) and they base their exposure on a match between the current scene and one in the database, rather than a simple meter measurement.
Maybe it is time to recognize the cognitive amplification factor, even if remains unexplained, even if there is no decent litterature about it. The cognitive amplification is the crazy effect you perceive when listening to a high quality audio system, with your mind generating supplementary information like the face expression of the singer or the way the two hands do walk on a piano keyboard. For enjoying cognitive amplification at the max, you need to enjoy the reproduced music without consciously asking yourself "hold on a second, how if this possible ?". As soon as you start realizing the madness, the cognitive amplification effect vanishes. It happened to you on some occasions, isn't ?
Your mind is far more sophisticated than the best equalizer, dynamics processor and reverb-cancelling device.
In this context, I would say that all what is needed is to allow your mind to work, allow it to boot using clues buried into the recorded material.
I don't know what those clues are. From my experience, you only need a few seconds for getting your mind deciding between garbage (audio coming from an average audio system) or gold (the perception that there was something happening on stage, with skins, textures and colours).
Our memory is resolute and powerful, even if the access remains limited for species survival purposes. If our memory had unlimited access, we would prefear playing and replaying our best moments without concerns for preserving life and reproducing ourselves.
In the context of an inner memory more capable and more resolute than the access method would suggest, a possibility is that every sound coming from an audio system gets internally convolved in our brain. Convolved with a high resolution prototype elaborated by our Gestalt.
Such internal high-resolution prototype may thus embed very special clues, subjective clues. Such internal high-resolution prototype is thus a moving target, depending on the culture and the experience of each individual.
And still, the big question remains : is it true some golden ears can perceive subjective differences between 0.01% THD amplifiers ?
The brain may be doing something completely different than just a convolution. For life preservation purpose, a mature brain always ranks each auditive perception using three scales : is it a predator coming, is it my baby crying, or is it a sexual partner entering communication. Knowing that when the same brain was in infancy, in the mother's womb, the sounds looked radically different.
I guess the brain gets confused when it gets 33% on all scales, not being able to command the body in a clear way.
The brain needs thus to process burst of audio signals containing clues. The detection process needs to be sensitive. A sensitivity of one part per million (1ppm) doesn't shock me at this stage. Think in terms of biochemistry, instead of dB in a spectrum analyzer.
Consider the latest audio IFF systems used on submarines. The clues over there are a few ppm down the gross signal. And they seem to pass intact through a blameless 0.1% THD underwater sensor and a blameless 0.1% THD operator headset.
The most sensitive detection methods are model-based.
A possibility is that our brain, why listening an instrument on a high-quality audio system, is entering a model-driven behaviour. This way, - big surprise -, what we perceive is not the sound, but the output of our internal model.
With a satisfaction coefficient being generated, based on the differences between the sound pressure and the model output.
Your mind is far more sophisticated than the best equalizer, dynamics processor and reverb-cancelling device.
In this context, I would say that all what is needed is to allow your mind to work, allow it to boot using clues buried into the recorded material.
I don't know what those clues are. From my experience, you only need a few seconds for getting your mind deciding between garbage (audio coming from an average audio system) or gold (the perception that there was something happening on stage, with skins, textures and colours).
Our memory is resolute and powerful, even if the access remains limited for species survival purposes. If our memory had unlimited access, we would prefear playing and replaying our best moments without concerns for preserving life and reproducing ourselves.
In the context of an inner memory more capable and more resolute than the access method would suggest, a possibility is that every sound coming from an audio system gets internally convolved in our brain. Convolved with a high resolution prototype elaborated by our Gestalt.
Such internal high-resolution prototype may thus embed very special clues, subjective clues. Such internal high-resolution prototype is thus a moving target, depending on the culture and the experience of each individual.
And still, the big question remains : is it true some golden ears can perceive subjective differences between 0.01% THD amplifiers ?
The brain may be doing something completely different than just a convolution. For life preservation purpose, a mature brain always ranks each auditive perception using three scales : is it a predator coming, is it my baby crying, or is it a sexual partner entering communication. Knowing that when the same brain was in infancy, in the mother's womb, the sounds looked radically different.
I guess the brain gets confused when it gets 33% on all scales, not being able to command the body in a clear way.
The brain needs thus to process burst of audio signals containing clues. The detection process needs to be sensitive. A sensitivity of one part per million (1ppm) doesn't shock me at this stage. Think in terms of biochemistry, instead of dB in a spectrum analyzer.
Consider the latest audio IFF systems used on submarines. The clues over there are a few ppm down the gross signal. And they seem to pass intact through a blameless 0.1% THD underwater sensor and a blameless 0.1% THD operator headset.
The most sensitive detection methods are model-based.
A possibility is that our brain, why listening an instrument on a high-quality audio system, is entering a model-driven behaviour. This way, - big surprise -, what we perceive is not the sound, but the output of our internal model.
With a satisfaction coefficient being generated, based on the differences between the sound pressure and the model output.
Last edited:
Aware that being a "subjectivist" doesn't turn someone into an idiot. There's no objection that DS is very clever. Just imagine someone as clever as DS, knows electronics equally well, but was born with better hearing, would you say that he is a subjectivist who disregards all the measurements? I believe many designers, if not all, can measure what DS can.
And I don't really understand what is the problem with trusting or not trusting people. I myself don't have a need to trust anyone or anything. I have built the blameless before I knew who DS is. I just don't want to live with the amp (thus it will be difficult to persuade me to build his other amps if any). If you like the blameless, stay with it, it is not too expensive.
I guess the fundamental problem with such statement is the inability to notice that it is perfectly impossible to throw an amp in the signal chain and pretend as if nothing is added to the sound. There's no such thing as straight wire with gain...
I had a trought letaly.
During recording audio signal is passinag a lot stages(opa/discret/valve) and still when reproduced it sounds different on differene amplifier-loudspeaker combination, sometimes good and sometimes not so.
Maybe there is different interaction between loadspeaker and output stage, NFB included. EMS from loudspeaker could enter input stage trough NFB in differen way for different AMPs. There could be a reason way someone prefers non NFB(global) AMPs.
To prevent EMS from the loudspeaker to enter back input could be a way to beter reproducted sound. Low output impendance maybe is not good enough solution for that.
Dado
During recording audio signal is passinag a lot stages(opa/discret/valve) and still when reproduced it sounds different on differene amplifier-loudspeaker combination, sometimes good and sometimes not so.
Maybe there is different interaction between loadspeaker and output stage, NFB included. EMS from loudspeaker could enter input stage trough NFB in differen way for different AMPs. There could be a reason way someone prefers non NFB(global) AMPs.
To prevent EMS from the loudspeaker to enter back input could be a way to beter reproducted sound. Low output impendance maybe is not good enough solution for that.
Dado
Niceness?....A Selfish term for nicely coloured sound. As we are all different in our tastes of nice, the control will need 2 Functions IMO. Degree and type as perhaps in spread of harmonics 2nd, 3rd, 4th. Now, will my fuzz-box handle that, or do we need dspmicros?
Actually I'm not sure whether his was a tongue-in-cheek or clenched teeth remark. The carefully clipped prose of his earlier editions probably hides a bit of frustration with press, readership and clients. It would be enlightening to hear from someone closer.
Actually I'm not sure whether his was a tongue-in-cheek or clenched teeth remark. The carefully clipped prose of his earlier editions probably hides a bit of frustration with press, readership and clients. It would be enlightening to hear from someone closer.
- Status
- Not open for further replies.