Lets start by quoting you on this, since it really is the key point
imho:
bear this in mind in all of the below.
That is an inevitable logical conclusion from the above I think.
Here's an analogy for you: it doesn't do any good for my FM receiver to
have a threshold sensitivity of -100dBm if the ambient noise is at -80
dBm.
IOW, if hi-fi amp distortion is below a threshold of audibility, then
it doesn't need to be lower or 0. Moreover, if two amps both have
distortion below a threshold of audibility, it doesn't matter if they
are characteristically different.
I would suggest that is an unrealistic definition. Hifi buffs have not
generally claimed that lowest paper distortion spec is what sounds
best.
I could not call them "hi-fi buffs if they aren't clear regarding what
hi-fi is. Yes, one needs to use caution in reading specifications. It
is likely that most of us in s.e.d are well aware of specsmanship games
played by vendors.
It is well known different types of distortion sound different,
therefore it is elementary to conclude that the amp with the lowest THD
is not guaranteed to be the one that sounds best.
The point of a hi-if element is to not "have it sound" at all. To have
it below an audible level is to say it has no sound. This is a separate
task from intentionally processing and coloring. There is nothing wrong
with coloring sound sources, but that is a task separate from hi-fi, and
one that should be explicit and independent.
I am not aware of any scientific study that conclusively shows that
humans can reliably detect different characteristic distortions under
0.5 to 1.0%, for as good or poor as the gross THD spec is (a "paper
spec," as you rightly call it). I could never make the claim that
"different types of distortion sound different" because I am aware of no
evidence that such a statement is true *in general*. By "in general," I
mean allow for specific exclusions (guitar amps operating with high
distortion are perhaps a good example).
The same processing was used, a processing claimed to be superior by LP
fans.
sequence:
1. LP claimed to sound better than CD, for same master source
2. Copy LP to CD
3. listener can't tell the difference between LP and LP copied to CD
4. If #1 and #3 are both true, then it is a logical conclusion
no idea if there even is one. Trying it and seeing soon answers the
question in a practical way.
But perhaps you're saying the technical definition of warm is your 50 Hz
mod system.
This is s.e.d. Since engineers deal in applied science in designing
electronics, a scientific goal definition must be made. Otherwise,
there is nothing to work with.
missing the point. In the real world many cut price systems are used,
and it costs money to add bass boost, add speaker bass response, and
add amp power capacity to be able to run it. Given that cut price
systems are used, the logical thing to do with them is maximise their
abilities. This can be done at apx no cost by injecting mains junk into
the amp.
Even $50 boom boxes have bass boost and tone controls. What is the
price range you're looking for? Incidentally, I think most tube
audiophile gear is on the other end of the spectrum -- they should be
able to afford tone controls.
???
Its really not. The 50Hz modulation method gets fairly consistent
responses from listeners.
Evidence?
One must question why, if a positive and significant positive listening
response is shown across large populations, why this apparently
inexpensive enhancement is not more commonplace. Why isn't there a
"warm" knob on cheap, mid, and high priced gear? It's absence is very
puzzling for such a clear claimed inexpensive enhancement.
The contrast between 2nd and 3rd harmonic
distortion might also be, though I cant be sure. I really see no reason
to think its inconsistent.
I would have to see a good study, since there would seem to be no reason
to assume uniformity or even consistancy.
Whether that goal is achieved is another question. Many will argue that
its not.
To my knowledge, there is no double-blind evidence that people can
reliably detect the low distortions commonly available in today's hi-fi
power amps. It is one matter to buy what one likes, for whatever
reason. It is another to make general scientific claims about human
hearing.
Also, the ancedotal evidence, for myself, and the mass of people I know,
is they (and I) can't make quality distinctions between amps with low
distortion content. Thus the lack of scientific evidence by those who
claim ability to detect differences at these low levels, combined with
the anecdotal evidence, leads me to remain skeptical of the claims. I
have no reason to believe it. It *may* be true, but the evidence is
entirely missing.
If its not, we are indeed comparing one set of distortions
with another, in a world of not quite ideal amplifiers, and the one
with the most euphonic set of distortions wins.
Well in my world that wouldn't be so. I would pick the cheapest with
distortion low enough such that I couldn't detect the difference between
the cheapest one and another one that cost $1 more (certeris paribus).
That is, I don't care what the distortion is in either one, if I can't
detect it, and like we learned in ECON101, all decisions are made at the
margin.
Right, but thats a different concept to the one I was talking about. I
was talking about things like eg adding 50Hz modulation to partially
make up for lack of bass, and suggesting the same concept of
semi-compensation occurs with distortions present in much lower
percentages, ie in hifi amps.
Basically you are talking about masking. It is similar in a sense to
the lossy compression of mp3. Where some of the sound content
sufficiently masks other content, the content that is masked can be
dispensed with.
This is a processing that you believe sounds nice. That's fine, but it
has little to do with hi-fi.
I think its well demonstrated.
To my knowledge, it has not been demonstrated.
if that were true, one would simply look at the THD of amps, pick the
lowest, and that would be it, youd get best sound. Real life doesnt
work like that.
I bring up the subject of lo-fi a lot in this because with larger
levels of distortion it is extremely easy to see the effects, and thus
the concepts in action. It is easy to take a basic 0.5w amp, with
inadequate bass, and add 50Hz modulation.
The problem isn't inadequate bass, per se, it is insufficient power.
It is easy to compare 3rd harmonic with 2nd
when youre running at 2% distortion, not 0.002%. etc.
This is exactly the crux of the matter. You, as best I can tell here,
believe that because distortion is audible and distinguishable at higher
levels, it is audible and distinguishable at lower levels, but just at
some scaled down version.
There is no scientific reason to presuppose this sort of linear level
transposition of distortion audibility in human hearing. Nature is
replete with examples of non-linear responses (including threshold
effects) by living creatures to environmental phenomena. Perhaps the
window response of humans to many phamaceuticals is a good and
well-known example of non-linear response.
In short, there is zero evidence that I am aware of that would
demonstrate that humans do *not* have thresholds in their ability to
hear subtle differences in sounds. The instrument of the ear may indeed
be "sensitive," but that does not equate to infinite sensitivity, nor a
linear response in sensitivity.
Lots of people are familiar with the 50Hz modulation trick, though most
dont realise thats how its done.
You're posting to s.e.d. Most know how modulation is performed. If
they don't, they are probably a bit out of place here.
In a lo-fi one would usally add it at the speaker, with underdamped
mechanical resonance(s). In principle one can add these things
anywhere.
This is a compensation nulling (and to some measure imperfect) of
another hi-fi shortcoming. It is not a masking.
Certainly thats another valid question. Im not sure its one we can
asnwer in this thread though.
ignorance and unscience permeate most fields. That only means its not
worth looking in the places where one finds no science. It doesnt make
a field invalid.
Well sure. But the fact of life is we only have so many resources.
Basically there is no low-hanging fruit left to pick when it comes to
amplifier distortion. Today's hi-fi fan who commands even a modest
income can get a low distortion and high power amp for fairly low cost
in real dollars. This is another way of saying _it isn't worth
pursuing_ in nearly all cases.
The fact that the "audibility question" is not "one we can answer in
this thread," and about a billion others sends us the message of severe
doubt when it comes to valuation of extreme efforts in simply proving or
disproving these marginal claims regarding amplifier distortion in hi-fi
amps. Information can be expensive at times. In this case, the
information is not worth the cost of procurement, as best I've been able
to tell.
This is why I keep mentioning lo-fi. With lo-fi you can turn the
distortion levels right up, and the results are immediate and obvious.
Again, this is apparently an assumption of transposition of audibility.
There is no reason to presuppose this that I am aware of.
If
there have been breakthroughs in the past 10 years, I am not aware of
them because I'm not really looking.
Are you saying you don't like the mix/production
of the source material and [adding]
distortion will make it better?
I daresay thats so in some cases
I think the difficulty with using an amplifier unit as a non-linear
distortion adder/processor is that it is not very controllable. That
is, a hi-fi amp is designed to *not* distort, not to distort. Different
distortion levels will be highly dependent on the volume level -- that
is, volume is not independent of distortion. I could not call this a
good audio processing design by any engineering standards I am familiar
with.
this is really missing the point that no amp is in reality distortion
free. Any good designer is handling distortion: no-one has yet found a
way to the 0% amp.
More importantly, there is no reason to assume such a thing is
necessary, for any practical purpose of hi-fi listeners.
I would suggest changing your definition. That may have been good when
the term was coined.
No, the thing to do would be to dispense with the term "hi-fi"
altogether rather than promote an oxymoron that obliterates our
language. But who can command people to lose interest in hi-fi? Why
would anyone do such a thing?
Amplifier distortion has not been brought down to 0% yet, the dragon
has not been slayed. Easiest to make good maybe, but to eliminate all
problems from, no.
If people can't hear it, and to my knowledge they can't, it has been
slayed. There is no reason to presume a "need" for a zero level of
distortion. That would be a solution in search of a problem.
me too. Also I think more amplifier classes will come along with better
efficiency, cost, or other advantages. I worked on a design that
shifted the heat diss away from the output trs, but unfortunately it
did not in the end enable any less silicon to be used.
But in reality they do not correct such defects. A look at any
commercial speaker's response curve shows it can not be corrected with
a bass or treble control.
Not completely, that is certainly true. But no claim of perfection was
made. A *partial* compensation was affordably provided. A less
affordable solution such as a graphic equalizer will do a better job.
Life, and hi-fi, are about tradeoffs.
These controls will lessen the infidelity in some areas of the
spectrum, while worsening it in others. IOW the control's frequency
response does not match the speaker's.
Obviously. The question for the user is the mix they deem optimal; that
would be balancing the defects as best they can. The designer simply
allows them a measure of choice in determining that mix for the local
situation.
not usually.
Im sorry but theyre inevitable. Its thus meaningless to say theyre a
no-no. Theyre an inevitable which must be handled.
Oh, it is definitely a no-no.
maybe one day that will be reality.
Regarding scientifically *demonstrable* human ability to detect
distortion, it is a reality and has been for some time. Sure, some
folks *claim* ability to hear minute levels of distortion, but they
haven't been able to demonstrate it.
Maybe they can detect it, but who else cares since the information to
really know is exceedingly expensive? It is probably cheaper for those
who make these claims to simply buy the stuff they believe is best than
to prove the matter scientifically. But this is *sci* electronics
design. Without science, it is irrelevent here.
IMO, the people who make these claims should stop making claims they
can't prove. They should simply say it is their belief. Then their
subjective wants, and real purchases are based on those beliefs, can be
made and no rational justification is needed. Doing so would remove the
rationality aspect; they would not need to attempt defending what has
been an intractable problem given the cost of good information. Just
treat it like religion: people can beleive what they want if it doesn't
hurt anybody else. What do I care if someone pays $3k for a 50 watt
tube amp? I could not care less.
It is part of every amplifier design on the planet, good and bad alike.
There is no 0% distortion stage.
There is no evidence that the fact of non-zero distortion is relevent to
human hearing. There is no necessary linkage.
It is normal in amp design, and is used to lower overall distortion,
but not cancel it.
There is a better way to phrase the distortion control techniques in
audio amps. The intent is certainly to *cancel* distortion, and in fact
this is what is done. It is only that it cannot be "done" in the
absolute (zeroed) -- it cannot *totally* cancel it. To say it is
lowered is indeed to say it is cancelled or controlled *to a measure*.
I think that describes graphic equalisers, not tone controls.
You are correct. A bass an treble control are cheaper and don't provide
the precision of a graphic equalizer. That is explicitly why I used the
term "partially compensating," among others. For tonal adjustment, no
one claims absolute perfection any more than one claims amps have 0%
distortion.