Advice: Amp building
 

Jim Lesurf wrote:
In article , Rob
wrote:
Jim Lesurf wrote:

What is 'the' scientific method - intrigued!
Pretty much, it what I summarised. The key point is that you set out
to collect data in a way that allows you to assess the results for
relevance, reliability, etc. In particular, you test in ways where the
outcome can 'falsify' the ideas you before you started.

That's fine, but I'm concerned about where your original
ideas/hypotheses come/came from.

That would depend on the case/circumstances.

In general, the process of the scientific method starts from some
observations about the real world. It is then open to anyone to propose a
'hypothesis' which tries to explain these observations in terms of a
proposed underlaying physical process. This may be based upon spotting
apparent similarities, or parallels, or links with other observations or
theories which have been successful in explaining observations in other
areas, or in the past.

However that is just the start. It is then vital that the hypothesis be
*testable*. i.e. that we can then carry out some new observations,
generally as part of an experimental process, to gather new results which
might either 'support' or 'contradict' the hypothesis.

The above point is important in the scientific method. We need to be able
to carry out tests *whose outcome might show our hypothesis was
ill-founded*. In the jargon of experimental method the hypothesis must be
'falsifiable'. i.e. we have some experiment or new observations we can
obtain which would have the ability to indicate that the idea (hypothesis)
wasn't a good one.

Thus experiments are designed with the specific aim of having the ability
to show the idea(s) they are testing may be wrong as descriptions of
reality. If we can't do any such experiments, or make any such
observations, then the idea would be regarded as 'not a scientific
description of reality'.

If the results of the experiment support the hypothesis, then we would
regard the hypothesis as being of some use. But we would then try again to
see if we could 'catch it out' with some fresh observations or experiments.
We'd also search for any possible contradictions between the idea and other
ideas (theories) that have stood up to repeated tests in the past.

The results of this process are a set of interlinked ideas which have
'survived' a number of 'challenges' in the form of experiments that tried
to find flaws in their description of reality. But these ideas are always
'provisional' in that we may eventually find flaws, and then have to come
up with a new idea. We leep on trying to find flaws, and then use these as
the 'seed' of some new ideas.

So, for example, Newton's Laws of motion 'survived' many such experimental
tests over many years, but were eventually superceeded by Einstein's
theories. The reason being that we found circumstances where Newton's ideas
did not accurately describe how things behaved, but Einstein's ideas did.
We still use Newton's ideas as an 'approximation', though, which gives
useful results in many cases without the complications of Einstein's.

But given what I have described, it should be no surprise that there are
people proposing new ideas which could superceed Relativity as the basis of
gravitation, etc. Here the main bone of contention are issues like galactic
rotation and having to use the idea of 'dark matter' to make the
observations fit Einstein's theories. At present we have no idea if
Einstein's ideas will 'survive' this - but even if they do, they may be
replaced at some point - when the evidence shows this is appropriate.

The purpose of collecting a fair amount of data/results is to enable
you to see if the results aren't likely to be due to 'chance'. It also
is to allow you to 'randomise' other factors which may be altering the
individual results as a result of an uncontrolled variable.


Are you saying that any one circumstance (a fast lap time, say) can be
explained by a finite number of variables? Is that the theory behind
your method? I don't think you're saying quite that (yours might be a
reasonable or approximate truth), but in this context I'm unclear
how/why/if you rank variables. To understand 'the' method I'd have to
know how, as a point of method, these things are dealt with.

This is a different, but related issue. This the topic of experimental
design and the methods/protocols used to actually collect observations.
This is quite a complex topic, and the details vary depending on the area
of study to some extent. However I will try to give an outline.

The basic problem is that when you make an observation, that - in itself -
does not generally tell you *why* the outcome was as observed.

There will tend to be 'uncontrolled' variables which may affect the
results. Some of these may be 'known' in the sense that you are aware of
them, and would expect them to be likely to have some effect. Others may be
'unknown' in that you either have no awareness that the variable exists, or
it has not occurred to anyone that it could affect the results you obtain.

I would not describe these as a 'finite number' of variabled. Better to say
an 'undefined' number or 'unknown' number.

We can try to deal with these in various ways, and I can give some
examples.

In some cases the variable may be controllable. For example for some
experiments we may expect the results to be affected by temperature, and
can then run the experiment and collect observations while keeping the
system being experimented on in a 'controlled environment' where we
maintain a steady, known, temperature. In this way some 'uncontrolled'
variables can be controlled, and we tackle the problem in that way.

In some cases we can't control the variable. So we may know that
temperature has some effect, but the nature of what we are studying makes
it impossible to actually control the temperature. We can then proceed in
various ways.

One way would be to only run the experiments when the temperature just
happens to be a pre-chosen value. Thus we aren't controlling the variable,
but selecting when it is at a given level. This means that we we repeatedly
take observations, any change should not be due to temperature, but may be
for some other reason, and so we have excluded temperature effects.

The above may be frustrating as we waste time waiting. So a common
alternative is to keep repeating the experiment and taking observations,
noting the temperature each time we do so. When doing this, we keep
repeating the same conditions of interest. For example, lets assume that we
are interested in the effects of, say, magnetic field, on something. We
make repeated measurements, but use the same set of field levels on many
occasions. This means we now have many observations for *each* chosen field
level, but for various temperatures.

We'd group the results into sets with the same field level, and in each
group see how the results show any temperature effect. By selection we are
controlling the field level to discover the temperature effects.

We'd then reverse this and group the results into sets with the same
temperature, and in each group see how the results show field effects.

This is seperating out different variables.

An alternative, and common, approach is 'randomisation'. Here we keep
running the experiment, applying different fields, but using the same set
of levels many many times. We go on doing this over a long enough period
that the temperature will have varied around, going up and down many times.
We then regard the temperature effects as a source of 'noise'. This
randomising works provided we can ensure the temperature isn't correlated
with the thing (field in this example) whose effects we are trying to
determine.

The snag is that we have to make many measurements, and the do an
appropriate statistical analysis to determine if the results show anything
at all with any level of reliability. We also have to accept that there is
always a non-zero chance that an outcome was produced by the 'noise' and
has fooled us. The purpose of many repeated 'randomised' tries is to reduce
the chance of such a 'noise-created' outcome to being so small that we
don't feel we have to worry about it.

The purpose of trying to exclude some factors (e.g. not telling the
'driver' what arrangement/chance is being tried) is to see if their
opinions on what is going on mean anything.

That's OK as it goes, and is a reference to DBT I suppose. The problem
in an audio context (and motoring) is that you cannot, I think, isolate
or reduce testing to fixed/certain variables.

DBT is one form of protocol, but I wasn't necessarily meaning DBT.

The experimental protocol, however, has to employ a mix of 'controlling
variables' and 'randomisation'. So, for example, ensuring a 'blind' test
means we can see if the judgements of the people who say they can tell A
from B are actually reliable. But to do this, the test has to be done
enough times for us to establish what level of 'confidence' we can put on
the result.

The randomisation for listening tests also helps us with uncontrolled
variables. e.g. lets imagine that the speakers used warm up with us, or
during the day, and this alters their behaviour to an audible extent.

if we simply allowed someone to listend to item 'A' via the speakers, and
then item 'B'. they might say, "A sounds different to B' and give a
description that others taking the same test would agree with
independently. However unless we take care, the conclusion may be that 'A
sounds different to B in this way...'

But the reality might be that 'A was always tried before B, and the
speakers were temporarily altered by trying A'.

Hence such an experimental protocol might simply lead to a mistaken belief.

To avoid this, we would have to repeat the comparison many times,
sometimes using A then B, sometimes B then A. This would 'randomise' the
effects of the speaker warming up upon the end results when they were
analysed.

I have tried to give some simplified examples, so I have left out a lot,
but you can see that the above is already quite a long description. In
practice, the design of experimental protocols and methods can be very
demanding.



The idea is also to see if you can 'catch out' the ideas people have
and find they have flaws, or are making an error. Not simply to find
'support for a belief'.

The actual experimental protocol would vary according to the aims of
the specific area being investigated. So for a specific case we'd have
to lay out a more detailed set of methods.


There is a simple difference in your view of 'the' scientific method and
mine. I believe that some data is not not 'seen', and some is not
measurable in a consistent fashion.

If it is not seen then it is not 'data'. However if you mean that there are
variables of which we may be unaware, then I agree. This is where the
correct experimental methods come to our aid but we cannot give an
absolute guarantee we have 'eliminated' any effects. All we can do is
reduce the chance that they have caused a misleading result, and perhaps do
more experiments in the future. Thus scientific understanding is always
provisional and a matter of finite levels of confidence rather than
absolute certainty.

Does the above make sense as an explanation? Afraid it is longer than I
might like, but as someone else once said, "I could have made it shorter if
I'd had more time". :-)


I think that's a pretty good explanation, thanks. I think one thing you
highlight is that the method you describe is capable of considerable
interpretation.

One thing that could be expanded upon is *who* is interpreting the 'real
world' - it's pretty obvious that most of what we know has arisen from
the work of, say, white men. Does that matter - even if they have
followed this method?

Anyhow, bla bla :-)

It would be very useful to me, and I suspect others, if you could post
your explanation on your web site.

Rob

Advice: Amp building
 

In article , Rob
wrote:
Jim Lesurf wrote:
[big snip]

Does the above make sense as an explanation? Afraid it is longer than
I might like, but as someone else once said, "I could have made it
shorter if I'd had more time". :-)


I think that's a pretty good explanation, thanks. I think one thing you
highlight is that the method you describe is capable of considerable
interpretation.

The method itself is fairly well-defined. The difficulties can tend to
arise in people developing and agreeing - and being able to carry out - an
experimental proceedure. This is becuase the specific nature of the
'problems' to overcome vary from case to case, and there may be debate or
disagreement regarding the 'best' experimental methods or protocols to
tackle the problems.

One thing that could be expanded upon is *who* is interpreting the 'real
world' - it's pretty obvious that most of what we know has arisen from
the work of, say, white men. Does that matter - even if they have
followed this method?

Can't comment on the second part of the above as I don't know.

re the first part: One of the advantages of the scientific method is that
it sets out that both the experimental process and the observed results
have to be available for others to examine. This means they can form their
own views regarding the appropriateness of the specific experimental
method, and the reliability of any conclusions. It also allows them to
either repeat the experiment, or having identified any 'flaws', to design
and carry out a different one of a 'better' design which tries to overcome
those 'flaws'. In turn, their experiment and results should then be open
for others to repeat this process

Anyhow, bla bla :-)

It would be very useful to me, and I suspect others, if you could post
your explanation on your web site.

It is on my 'to do' list. :-) Alas, so are many other things, but I hope
to get around to it soon as the topics of "do amp/cables/etc sound
different?", "are subjective reviews worth the paper they're printed on?",
etc, crop up repeatedly, and it would probably be useful if the relevant
arguments from the scientific method and experimental design were explained
so people can see the reasons why there are serious doubts about a lot of
what is claimed in consumer magazines, etc.

Slainte,

Jim

--
Electronics http://www.st-and.ac.uk/~www_pa/Scot...o/electron.htm
Audio Misc http://www.st-and.demon.co.uk/AudioMisc/index.html
Armstrong Audio http://www.st-and.demon.co.uk/Audio/armstrong.html
Barbirolli Soc. http://www.st-and.demon.co.uk/JBSoc/JBSoc.html

Advice: Amp building
 

"Dave Plowman (News)" wrote in message
...
In article ,
Keith G wrote:
Wrong Don. (At least in my case!) I have almost no originals to compare
- when I play music it is the original. Fidelity to me is being able to
tell a violin from a viola, or a clart from an oboe or tenor saxophone,

You have pretty low sights, then, given very basic equipment will allow
this.


First, ignoring for a moment the recording medium, it is interesting
to find out what people can and cannot hear sitting in an auditorium
with real instruments in front of them. There are a very large number
of interesting tests to determine this.

The violin/viola example quoted above is a very good one.
Likewise the alto and tenor saxophone playing the same melody in
the same (overlapping) register. Can you tell the difference between
them Dave? Most people cannot. It is the most frequently used test.

Now try to differentiate between a Guild and a Martin acoustic
guitar. They let's try a Bosendorfer and Steinway concert grand.
Try some of these experiments - you may be sadly disappointed
at your own ability:-)

It has been suggested that people who cannot differentiate between
an oboe and a cor anglais, have not chance of differentiating between
two amplifiers. This could be true.

I took part in organising an experiment which involved a small ensemble
recording - solo clarinet with a string quartet. The clarinet was recorded
on
its own track and the separation was good enough for us to be able to pitch
shift this track only, with regard to the ensemble. The pitch was
incremented
at the rate of 2% over sixteen bars. By the time we had reached 4%
(a demi-semitone) most of the musicians and studio personnel in the panel
were cringing. In contrast most of the others showed no reaction.

The second point is that differentiation between the timbre of instruments
is
clearer on some amp/speaker combinations than on others. This is one
of the reasons why I so much enjoy the Shostakovich Quartets on the
Reznekov SET amp, which reproduces them with stunning realism. For
me, that's what listening to music should be about.

Regards to all
Iain

Advice: Amp building
 

Jim Lesurf wrote:
In article , Rob
wrote:
Jim Lesurf wrote:
[big snip]
Does the above make sense as an explanation? Afraid it is longer than
I might like, but as someone else once said, "I could have made it
shorter if I'd had more time". :-)


I think that's a pretty good explanation, thanks. I think one thing you
highlight is that the method you describe is capable of considerable
interpretation.

The method itself is fairly well-defined. The difficulties can tend to
arise in people developing and agreeing - and being able to carry out - an
experimental proceedure. This is becuase the specific nature of the
'problems' to overcome vary from case to case, and there may be debate or
disagreement regarding the 'best' experimental methods or protocols to
tackle the problems.

One thing that could be expanded upon is *who* is interpreting the 'real
world' - it's pretty obvious that most of what we know has arisen from
the work of, say, white men. Does that matter - even if they have
followed this method?

Can't comment on the second part of the above as I don't know.

re the first part: One of the advantages of the scientific method is that
it sets out that both the experimental process and the observed results
have to be available for others to examine. This means they can form their
own views regarding the appropriateness of the specific experimental
method, and the reliability of any conclusions. It also allows them to
either repeat the experiment, or having identified any 'flaws', to design
and carry out a different one of a 'better' design which tries to overcome
those 'flaws'. In turn, their experiment and results should then be open
for others to repeat this process


I'm a little stuck on the notions of 'the' method and 'a' process. No
matter ...

Anyhow, bla bla :-)

It would be very useful to me, and I suspect others, if you could post
your explanation on your web site.

It is on my 'to do' list. :-) Alas, so are many other things, but I hope
to get around to it soon as the topics of "do amp/cables/etc sound
different?", "are subjective reviews worth the paper they're printed on?",
etc, crop up repeatedly, and it would probably be useful if the relevant
arguments from the scientific method and experimental design were explained
so people can see the reasons why there are serious doubts about a lot of
what is claimed in consumer magazines, etc.


I might have a go myself, just for no one to read :-)

A problem for people 'like me' is that I simply don't have the requisite
knowledge to take on technical argument. I was reading a semi-technical
book, to do with solid state and valves, and it was going fine until
about a third of the way through. I simply couldn't move from one page
to the next - the assumptions I'd been asked to accept were simply too
much. The same with papers on psychoacoustics - i just couldn't see the
link between theory and method and conclusions.

I'm increasingly drawn to Wikipedia as a catalyst for discussion:

http://en.wikipedia.org/wiki/High_fidelity

.... raises more points than it answers, but this (audio reproduction)
subject is incredibly complex, and 'put up or shut up' is not
satisfactory for reasons I've touched on. The quest for realism and DBT
sections are at least mentioned, with some attempt at balance.

Rob

Advice: Amp building
 

"Iain Churches" wrote


The second point is that differentiation between the timbre of instruments
is
clearer on some amp/speaker combinations than on others. This is one
of the reasons why I so much enjoy the Shostakovich Quartets on the
Reznekov SET amp, which reproduces them with stunning realism. For
me, that's what listening to music should be about.



You're going to have a hard time peddling *realism* over 'sonic accuracy'
here, Iain - I'm sure one or two of the most *opinionated* posters here
believe we choose and prefer to use the kit we do because it is *less*
realistic.....

Advice: Amp building
 

In article ,
Keith G wrote:
You're going to have a hard time peddling *realism* over 'sonic
accuracy' here, Iain - I'm sure one or two of the most *opinionated*
posters here believe we choose and prefer to use the kit we do because
it is *less* realistic.....

That's a reasonable conclusion.

--
*I love cats...they taste just like chicken.

Dave Plowman London SW
To e-mail, change noise into sound.

Advice: Amp building
 

"Keith G" wrote in message
...

"Iain Churches" wrote


The second point is that differentiation between the timbre of
instruments is
clearer on some amp/speaker combinations than on others. This is one
of the reasons why I so much enjoy the Shostakovich Quartets on the
Reznekov SET amp, which reproduces them with stunning realism. For
me, that's what listening to music should be about.



You're going to have a hard time peddling *realism* over 'sonic accuracy'
here, Iain - I'm sure one or two of the most *opinionated* posters here
believe we choose and prefer to use the kit we do because it is *less*
realistic.....


Dave. You should try to forget the test bench for a moment, foget
your biases, close your eyes and listen. What we are talking about
here is a listening experience.

Are you one of those people who would choose a transparent acrylic
cellos in preference to a baroque instrument by Marcello Armand-Pilon?
The state-of-the-art engineered cello (especially the one
with the internal flourescent lamp and the pickup on the bridge:-)
*must* be better. It has 400 years of technical deveopment, is held
together with superglue (not boiled fish entrails) and has all the
advantages of CAD design and CNC production:-)

But seriously, beg borrow build or steal a *good* SET and a pair
of ELS63 (which is the combination I have been using) and the
Shostakovich Quartets (Supraphon) Follow the score if you
choose to, but most importantly, listen. It's an incredbly emotional
experience.

Then get back to us.

Cheers
Iain

Advice: Amp building
 

"Rob" wrote in message
...

I don't agree.

Of course not, that's part of your current religion.

The technology is led by the tobacco and petrochemical industries from
what little I see and know.

That would be very little.

Now the art *and* science of motor racing (and indeed audio) is something
I'm interested in, but I can't see it getting many ng inches here at
least.

hmm.

Can you explain in scientific terms why Schumacher (say, erm Michael)
drives cars faster than anyone else?

Which week?

Which kind of racing?

I would wager that his car is set up in some very peculiar ways, as his
team techs see it.

I doubt that the setup is truly peculiar.

People who don't study and learn from history just keep making the same
mistakes over and over again.

Well, you know of course, it tends to go first as tragedy, second as
farce. But that's just how people interpret history - you have your lens,
I have mine.

My lens sees technology and its underpinnings exceptionally well.

Advice: Amp building
 

"Andy Evans" wrote in message
oups.com...

note that extensive corrections were required to makes Andy's quoting
legible and even semi-standard

We have low cost high volume production amps that exceed any reasonable
need. AK

OK - so the 'need' for a reference system, e.g. for professional
musicians, that is better than high volume production ss amps is
"unreasonable".

First prove the need.

Amps are only useful and interesting as components of more complex
systems,

they are almost completely disinteresting as ends in themselves. AK

So why do you and Jim (DHTs) insist on pontificating over componants
which you have no interest in?

Me, pontificate?

Why should people who are seriously
involved in building and improving such componants be REMOTELY
interested in your pompous views when you clearly are only interested
in breezing in and out of threads without contributing anything of
value to the actual R+D of such componants.

Pointing out redundacy and wasted effort can be valuable services.

This is EXTREMELY tiresome
when it has to be endured on a regular basis.

Then don't read our posts. Is there a man with a gun that makes you hurt
yourself this way?

On other newsgroups
people actually help each other build projects, go through different
iterations, contribute experience of their own builds, suggest
modifications.

Been there, done that.

Result - great equipment.

Not in every case. There are pompous people who pretend to be experts and
lead people in circles. Pointing those people out is a valuable public
service that some of us provide.

Think about it.

Talk about being pompous. Note Andy's suggestion that people need his
prompting to think.

The 'analogy' Andy gave was, actually, quite useful - it does show his
lack of grasp of the scientific method and how it can be applied in such
cases. (Jim)

"Useful" only in the sense of getting Jim his usual kicks from
appealing to the whole of the world who he assumes is on his side and
ready to tut tut in rhythm.

Whatever that means.

Jim's a good guy with lots of valuable information and wisdom to share.

We used to have people like that in school
- we called them "teachers pets".

Yeah, and they probably got far better grades than you did, Andy.

Still devoting your life to resolving your childhood hurts?

I'm sure you remember the kind of
thing "Please miss, Penny's stolen the blackboard cleaner and she's got
it under her desk".......

There's no relevance to that, and what Jim contributes. In fact Andy, your
post right here reads more like that.

Advice: Amp building
 

"Keith G" wrote in message
...


OK, quite by chance (I didn't take both these pix) I have what I think is
a good analogy for what I perceive to be the difference between analogue
and digital, valves and SS. See these very similar pix:

http://www.apah69.dsl.pipex.com/show...Screenshot.jpg

http://www.apah69.dsl.pipex.com/show...Screenshot.jpg

They are both handheld snaps of virtually the same frame from the same
film film - one from a DLP projector and one from an LCD projector.

Someone ought to tell you how to properly adjust the pickture on a video
projector.

Initially, the DLP image doesn't strike you as sharp as that from the LCD
and DLP projectors suffer from 'artifacts' that are not present with LCD.

The LCD projector pix has color shift up the ying-yang. Anybody with
non-trivial experience with video projectors should immediately notice that
some adjustment is in order.

Also the DLP projector is noisier, runs hotter and is much bigger than the
LCD. If you have not seen a DLP projector and only use an LCD one, it will
satisfy you completely and they are very easy and convenient to use.

In fact I have one of each in use in the same room. The two projectors
aren't really comparable and neither are the applications. However, there's
no overwhelming advantages of one over the other for the given application.

On paper the case for the LCD is strong, in reality the DLP ****es all
over it....

In reality, the differences in implementation and installation quality
outweigh most technical details related to the basic principles of
operation.