Chris says he prefers one over the other and that is enough for anyone. Preference is a whole mind thing. Some of us might have been more comfortable had he not but no additional evidence is required.
It is easy to say how better sounding the Modulus-86 was, knowing who the designer was or how good the measurement was. They call it expectation bias, don't they?
How about recording them and ask even the designer for his preference in blind?
Chris is stating a preference, not claiming one is better than the other. Nothing to require me to go on Amazon for another order of Troll food.
In a blind test, "preference" is much harder than "difference". You can't prefer A or B if you can't differentiate between A and B. That's the point.
0.19% of what?Jay said:
I have no idea. My aim was hi-fi; I arrived at a sound which satisifies me.
That would depend on how far out they lie. I said "similar", not identical.
I have no idea why you consider that strange.
You listen with a microscope?
I have never conducted any formal tests. I assume your comment contains an implicit claim about the superiority of your hearing? Or merely the assumed inferiority of my hearing?
You believe you can hear something different about CFP amps. The most likely explanation is that they have been set with far too much quiescent current. This will result in gm-doubling in the central region, giving high order distortion.
His point was that while we can define distortion as X% on sine waves of a particular amplitude, we cannot say that music is being reproduced with X% distortion.
Sigh... Preference in the context of this thread is a whole body thing with all expectations and biases in place. Even though Toole seems to find no issue with preference blind testing we are not debating that.
Some people prefer having their scrotum nailed to a bit of 4x2 over a film and a cuddle. You cannot DBT that and I certainly don't want to try it for myself.
'Audiophoolery' includes the self-proclaimed defenders of scientific rigor in forums like this unless you're completely avoiding the social point of the post and playing light with Sean Olive's recent work. Perfectly flat was scientific perfection for decades, the requirement of the messy human element at the end of the listening chain at best deceived about what is 'right'.
OK here is my evidence Warren TenBrook's Summary of Head Measurements at Harman | InnerFidelity you will note the 1974 B&K curve that matches the Harman curve fairly closely. Where is your evidence?
I forget the detail of how the distortion was added in the klippel distortion test. I think it was similar to what Gedde did as detailed in his website, which I also forget. But I remember from listening that the distortion was not only one type.
I assumed that you must at least as experienced as I am in building amplifiers. I found it strange that you build valves but not solid state. One of my prejudice was that you don't want to mention an amp to avoid quantification of your subjective and qualitative statements...
You talk about all hifi amps should sound similar if competently design, but I don't know what is hifi to you or what is a competent amp to you. At least I expected you to mention amplifiers from Bob Cordell or Douglas Self or any other amps that I have built.
No, I listen with ears. Why I said "microscopic"? Because my experience told me that when people can hear something, I can too, easily. So when I reach my limit in listening, I call that "microscopic" and I don't expect people to understand how it is like... (Remember, we were talking about linearity. I thought that I know how linearity would be perceived by ears). So I avoided discussing about this because I thought it would be useless.
I wanted to know if you have conducted such blind test. Some people make a statement based on experienced, literature study, etc. I wanted to know which one is yours...
Because IF you had conducted the test, your "opinion" was interesting to me because I have a feeling that you should have found that 7th order would be harder to distinguish than 2nd order, so more or less the "opposite" of your statement.
I also have no experience in comparing 7th and 2nd. But I assumed that when the fundamental is at 1kHz, 7th order distortion is far too high in frequency. My way of perceiving such distortion is by "feeling" the fatigue, which is not easy at all. But perceiving 2nd order is easier. We just need to be "trained" or be familiar with the characteristics of sound when 2nd order is added.
As I have no idea how experienced you are in building amplifiers I cannot say. I can say that experience in building amplifiers can be unrelated to experience in designing amplifiers, and experience in designing amplifiers is not always well correlated with ability to understand amplifiers. Hence you could deduce nothing about the truth or falsity of my statements from knowledge of my experience in building.Jay said:
Hifi to me, and to everyone else who knows what the word means, is a reasonably successful attempt at sound reproduction. Anything else is just audio, including systems which cost a lot more than mere hifi.
Two errors there:
1. 7th harmonic will not necessarily be too high to hear, unless the fundamental is highish.
2. 7th includes IM, which can be at lower frequencies (using multitone or music as thr source).
The general consensus is that higher order distortion is more unpleasant, and therefore I would expect it to be more audible.
Broadly speaking, there seems to be two groups of people on here:
1. those who understand and largely believe the accepted knowledge in electronics and psychoacoustics which support hi-fi design - they may or may not have conducted suitable experiments themselves, as one of the advantages of writing and communication is that we can learn from what others have done; (so, for example, I can consider some aspects of how a CFP output behaves without having conducted listening tests because I can assume that CFP obeys circuit theory)
2. those who misunderstand or reject (with or without understanding) the accepted knowledge - they cannot have any evidence of their own version of electronics (as they don't have their own universe) and they usually have nothing other than anecedote to support their own version of psychoacoustics.
I realise that there may be a few who don't fit into either category, and newbies will be seeking to discover which category best suits them.
Mentioning an amp, even if I could, would not quantify anything.Jay said:
Good idea, but how? Weather you include speaker/amp/cable and use a microphone or simply record the line level output IMO you have changed too many variables hardly scientific (in the narrow sense of isolating the effect of a single change).
As for audiofoolery I still suggest starting with something simple like a disinterested third party cryoing/not cryoing and serializing several pairs of identical cables. Everyone can use all their senses and nobody knows.
Gee, I'm not even sure now if I can speak to what exact point I was trying to make in the context of this thread. Oh wait, I think it was something to the effect that after several decades of attempting to reconcile outstandingly low distortion / noise specs on numerous brands of commercial products ( Naim, Linn, Bryston, Classe, PS Audio, Sim Audio, Audio Research, Carver) to what I was(n't) hearing - e.g , as the electromagnetic transducers on either end of the chain generally have several orders of magnitude worse "specs" than the electronics, just how vanishingly low do those numbers need to be to deliver my own metric of "enjoyable hi fi" ? - and am I even permitted to have my own standards of "acceptable performance"?
Yes, by the time I had the opportunity to listen to the two amps mentioned in first my own system, then a couple of others at a DIY get together , I'd got to know the designer as a very thorough engineer and all round gentleman. Whatever label you want to put on the psychological effect with which that familiarity influenced my personal subjective assessment of what I'd heard is quite fine with me.
And yes, of the two amps I heard using the LM3886 chip, the Mod86 delivered more of what I'd be willing to pay for. I've owned enough audio equipment over the past 50yrs to realize that the satisfaction any piece / system can deliver is always subject to change.
your turn, Jay
Yes, by the time I had the opportunity to listen to the two amps mentioned in first my own system, then a couple of others at a DIY get together , I'd got to know the designer as a very thorough engineer and all round gentleman. Whatever label you want to put on the psychological effect with which that familiarity influenced my personal subjective assessment of what I'd heard is quite fine with me.
And yes, of the two amps I heard using the LM3886 chip, the Mod86 delivered more of what I'd be willing to pay for. I've owned enough audio equipment over the past 50yrs to realize that the satisfaction any piece / system can deliver is always subject to change.
your turn, Jay
Last edited:
If you have experience building solid state amps, you would have been able to answer my question regarding which (solid state) amp do you think is a competent design. But you cannot give a definitive answer...
If you have experience building CFP amp like Rod Elliot's P3A, I may be able to relate to your explanation about how CFP amps are usually mistakenly biased with too high current. P3A is biased with 75mA. Lower bias will have lower performance. Rod mentioned that 20mA is a minimum...
So, I have to conclude that, to you, P3A is mistakenly designed? (Okay, I think I know what you mean. It is just easier if the circuit is concrete don't you think?). And if only you can mention which CFP amp is the correctly designed, instead of qualitative statements, I will be able to build it...
Too simple intention on my part, imho.
That's what I referred to as subjective qualitative statement...
I have mentioned the assumption regarding the fundamental frequency of 1kHz.
From experience I can tell that, yes, in general the higher order distortion is more unpleasant. But "unpleasantness" is very hard to perceive (We don't "hear it" anymore but "feel it") and therefore less audible.
Okay, thanks...
Two comments:
(1) I forget the schematic of the Modulus-86. May be there is extra circuitry in front of the 3886 chip. Sometimes, it is like having a preamp or buffer in front of one of the amps, making the listening comparison not fair...
(2) We have heard: "Take every claim regarding amp superiority with a grain of salt". This is very true because most of the time it is an expectation bias, then people have so big limitation in listening and experience that they cannot tell the difference between "better" and "just different"... I used to do a "background check" to see if I should trust the man and build the amp. I don't know you so I will just note that you think this is much better than a lot of good amps you have heard 😉
Wwhattt??!!? 😀
75mA is far too high - unless you believe in the 'first watt' idea and are happy to have significant crossover glitches at moderate signal levels in preference to much smaller crossover glitches for small signals. See this thread - especially my post 8. I calculate that with resistors of 220 and 0.33 he should have a quiescent current of 5.6mA in the output stage. Perhaps he never explored this region, because he 'knew' it was unreasonable?Jay said:
Until Self wrote his articles, and I explained his CFP results, it is likely that everyone set CFP current far too high. The correct current is too counterintuitive for most people, so I guess people continue to get it wrong.
From the designer's website
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
one of these things is not like the other,
one of these things is not the same
If I failed to avoid using the expression "much better than" in any of my comments in this thread regarding Tom's amps in particular, or the hundreds of others I've heard / owned during the length of my addiction to this hobby, I apologize. I'd thought I'd used the words preferred / would pay for, etc.
I don't really know your personal well enough - so in case that's not rhetorical - the point was that not every piece of equipment I've owned has "kept me happy" permanently.
Is there only a binary choice here between total objectivity and warm and fuzzy feel good subjectivity?
We used as low as 250uA for a 400mA max output stage. But I wanted to ask if either of you addressed the questions of ft or quasi-sat in power transistors at very low quiescent current density? At high enough frequency the traditional crossover "glitches" end up having little to do with the net distortion.
- Status
- Not open for further replies.