In article , Andy Evans
wrote:
This is the snag with making obervations without reliable measurements.
You end up having to try things 'at random' in the hope some might seem
better, but not then be sure if or why... (JL)

This is true, but it doesn't rule out trial and error as a legitimate
way of getting results.

Not sure how you are defining "legitimate". However I'd agree that we need
not exclude trial and error. But my points we

1) That we have to *learn* from the 'error' part, and be able to understand
the results so they guide what we do next.

2) Using 'trial and error' does not rule out *other* methods which may be
more systematic and better based on a large amount of prior knowledge which
people have collected, analysed, and systematically understood.

Thus as well as 'trial and error' we can then go on to add, 'use the error
as data for a diagnosis to tell you what tests to perform next'. Then
repeat this in a systematic and logical manner.

'trial and error' can work. But alas as soon as there is more than one
variable it can take infinitely long, and not even converge... :-)

Understanding tends in the end to save a lot of time and puzzlement, and
improves your chance of a good end-result. Trial and error, by itself, does
not.

When you think how many times Beethoven re-wrote
the 5th symphony (a process typical to composition), you can see that
the final result was in fact a process of trial and error.

Not really. He spent a lot of time and effort first learning about music.
I'm also pretty sure that he tried things and then considered them
carefully, not just wrote down the first thing that came to him and stopped
there. Hence his 'trial and error' wasn't the functional equivalent of
spraying some music paper with spots of ink, then turing the dots into
quavers! For that reason it is misleading to describe such a process as
'trial and error' unless you include th bits about understanding and
learning and then exploring/using the new understanding as well. :-)

I think it's quite fair to think of creating a sound system as an act of
creativity, and as we know from the initial stages of the creative
process multiple choices are combined with feedback of the effects of
these choices to gradually narrow down the practical possibilities. This
process of "divergent thinking" is the crucial part of creativity,

So is analytical and methodical thinking and being able to be
self-critical, etc. Hence you can't pick one plum out of this cake and
expect it to be the whole cake.

[big snip]

As above.

Far simpler IMHO to buy a Meridian and listen to the music. :-)


Saying this to a creative is like asking a claustrophobic person to go
potholing.

Sorry, I must not be 'creative' then, despite whatever I may have done.
:-)

Must go and get myself a purple hankie to wave about, and perhaps some
frills for my shirt cuffs... ;-

Note that I said 'simpler'. If you just want to listen to music, then
Meridian makes sense. OTOH of you want to fiddle about, then trying things
at random is fine.

However if you want to *learn* from 'fiddling about' or 'trial and error'
and hence both produce improved results, and have the satisfaction of
having learned, then you need more than just 'try things at random and
don't bother with any measurements or actual understanding', I'm afraid.

FWIW I regularly used to get hold of other people's amp and pull them apart
and find out how they worked. Good way to learn in my experience.

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