"What HiFi" - can it be trusted?
 

In message , chris
writes
This should be true ! BUT. Unfortunatly due to the design spec this
is not true.
There was some deep discussion on this whole issue a while back on
RAHE,
I too had thought like you. Then The Man from Belden explaind it
rather well: you could do a google on it his expanation should be a
lot better than mine.
But basically the Fibre call for in the spec is many times larger in
diameter that the wavelenght of the light used so instead on the light
bouncing down the fibre in a controlled fashon it bounces about in a
lot more random fashon and after a metre or three the uncontrolled
light bounces interfere with the main light signal generating extra
noise. This can cause misreading of the recieved signals, hence
errors, noise, distortion.
So by using a better fibre (which is not as cheap) that has a smaller
diameter, will improve the quality (by reducing the noise) at the
recieving end, resulting in less errors etc.


All audio fibre-optic links use multimode fibre. Single mode fibre
(with no bouncing about) is only used on telecomms links at hundreds and
more Megabits/sec.

At the VERY low bit rate used for SPDIF it really doesn't matter a damn.

Likewise, as many others have pointed out, for cable runs of a metre or
so, phono-plugs are quite OK for copper connections. Use a proper
RG-spec cable and BNC connectors for long lengths by all means.

I'm now playing with multi-channel 24bit, 48kHz sample-rate pro-audio
over Cobranet at work. Have a look at the Cirrus web site, some
seriously good work being done on the distribution of digital audio
feeds there.
--
Chris Morriss

"What HiFi" - can it be trusted?
 

On Sat, 3 Jan 2004 11:38:30 -0000
"chris" wrote:

So by using a better fibre (which is not as cheap) that has a smaller
diameter, will improve the quality (by reducing the noise) at the
recieving end, resulting in less errors etc.

Assuming you were getting errors in the first place, that is...

--
Spyros lair: http://www.mnementh.co.uk/ |||| Maintainer: arm26 linux

Do not meddle in the affairs of Dragons, for you are tasty and good with ketchup.

"What HiFi" - can it be trusted?
 

On Sat, 3 Jan 2004 11:38:30 -0000
"chris" wrote:

So by using a better fibre (which is not as cheap) that has a smaller
diameter, will improve the quality (by reducing the noise) at the
recieving end, resulting in less errors etc.

Assuming you were getting errors in the first place, that is...

--
Spyros lair: http://www.mnementh.co.uk/ |||| Maintainer: arm26 linux

Do not meddle in the affairs of Dragons, for you are tasty and good with ketchup.

"What HiFi" - can it be trusted?
 

"chris" wrote in message
...

"Oliver Keating" wrote in
message ...
I was flicking through "What HiFi" magazine and I came across
something very
odd - a series of reviews on *digital* audio cables, for connecting
a CD
player to an amp.

There are comments such as "this cable brings across a crisp sound a
cut
above the rest"

Now, at this point I have to shake my head in disbelief. Surely a
digital
cable about 1 metre long can easily carry a 1mbit data stream with
no
errors. Bear in mind ethernet has to carry 100mbits.


Yes a "DIGITAL cable most certainly can, but a lot of so called
"digital" interconnects arn't made with DIGTAL grade cables, and the
plugs also are not true 75 ohm, so you will start to get line
reflections, ringing on the recieved pulses (and if they are of
sufecent signal level cause pulse signal coruption and drop-outs) this
could start adding to the error rate, resulting in a less "good a
sound".

The thing is though, most of these phenominan only occur when you have
discreet pulses which travel down the line (and indeed get reflected), i.e.
you have a transmission line.

However, if you consider a data rate of 1mbit, or one pulse every
1microsecond, then say the pulse duration is 0.5micros, and say the
transmission line has a transmission speed of one tenth the speed of light
(in reality probably much faster, especially for good cables.) This means
the physical length of the pulse is 0.5e-6*3e7 = 15 metres.

So even with a lousy cable, the wavelength of the pulse is still 15 metres,
which is much longer than the typical length of an audio interconnect, so
you don't really have to treat the line as a transmission line - any
reflections would have the opportunity to bounce 15 times before the pulse
even ended, by which time it has decayed to zero and there can be sufficient
interval between that and the next pulse for any other artifacts to die
away.

Now, with a 100Mbit Ethernet cable, over a length of 100 metres, errors can
be a problem. But 1mbit over 1 metre? I would put money on it that the error
rate is zero, or negligably small.

about optical cables. There is absolutely
*no* way an optical pulse can be distorted sufficiently over 1 metre
that it
could result in an error.

This should be true ! BUT. Unfortunatly due to the design spec this
is not true.
There was some deep discussion on this whole issue a while back on
RAHE,
I too had thought like you. Then The Man from Belden explaind it
rather well: you could do a google on it his expanation should be a
lot better than mine.
But basically the Fibre call for in the spec is many times larger in
diameter that the wavelenght of the light used so instead on the light
bouncing down the fibre in a controlled fashon it bounces about in a
lot more random fashon and after a metre or three the uncontrolled
light bounces interfere with the main light signal generating extra
noise. This can cause misreading of the recieved signals, hence
errors, noise, distortion.
So by using a better fibre (which is not as cheap) that has a smaller
diameter, will improve the quality (by reducing the noise) at the
recieving end, resulting in less errors etc.

But the same arguement as above comes into play, especially with light,
where the typical pulse length is 150 metres. There is no way more than one
pulse can exist in the cable at once, so how is there any error possible?

I can only think that the testers suffered some sort of placebo
effect, or
they are in cahoots with the industry in order to provide a supply
of
customers who will shell out £350 for the "best" optical cable.

No, what they heard in the test is valid.
Sorry, its just good engineering to fix, bad engineering in the first
place.
As to the question is a cable worth £350 ? Well only your ears, brain
and wallet can make that value judgment.
I personaly would look for a much less expensive option (good dacs are
less than that).

In the end though, should you connect your CD player to your Amp using an
analogue or digital connection (I always assumed digital).

And if it is the latter, is there any point buying an expensive CD player?

snip

Who ever said life was easy :¬)

Indeedily :)

Happy New Year
Chris

"What HiFi" - can it be trusted?
 

"chris" wrote in message
...

"Oliver Keating" wrote in
message ...
I was flicking through "What HiFi" magazine and I came across
something very
odd - a series of reviews on *digital* audio cables, for connecting
a CD
player to an amp.

There are comments such as "this cable brings across a crisp sound a
cut
above the rest"

Now, at this point I have to shake my head in disbelief. Surely a
digital
cable about 1 metre long can easily carry a 1mbit data stream with
no
errors. Bear in mind ethernet has to carry 100mbits.


Yes a "DIGITAL cable most certainly can, but a lot of so called
"digital" interconnects arn't made with DIGTAL grade cables, and the
plugs also are not true 75 ohm, so you will start to get line
reflections, ringing on the recieved pulses (and if they are of
sufecent signal level cause pulse signal coruption and drop-outs) this
could start adding to the error rate, resulting in a less "good a
sound".

The thing is though, most of these phenominan only occur when you have
discreet pulses which travel down the line (and indeed get reflected), i.e.
you have a transmission line.

However, if you consider a data rate of 1mbit, or one pulse every
1microsecond, then say the pulse duration is 0.5micros, and say the
transmission line has a transmission speed of one tenth the speed of light
(in reality probably much faster, especially for good cables.) This means
the physical length of the pulse is 0.5e-6*3e7 = 15 metres.

So even with a lousy cable, the wavelength of the pulse is still 15 metres,
which is much longer than the typical length of an audio interconnect, so
you don't really have to treat the line as a transmission line - any
reflections would have the opportunity to bounce 15 times before the pulse
even ended, by which time it has decayed to zero and there can be sufficient
interval between that and the next pulse for any other artifacts to die
away.

Now, with a 100Mbit Ethernet cable, over a length of 100 metres, errors can
be a problem. But 1mbit over 1 metre? I would put money on it that the error
rate is zero, or negligably small.

about optical cables. There is absolutely
*no* way an optical pulse can be distorted sufficiently over 1 metre
that it
could result in an error.

This should be true ! BUT. Unfortunatly due to the design spec this
is not true.
There was some deep discussion on this whole issue a while back on
RAHE,
I too had thought like you. Then The Man from Belden explaind it
rather well: you could do a google on it his expanation should be a
lot better than mine.
But basically the Fibre call for in the spec is many times larger in
diameter that the wavelenght of the light used so instead on the light
bouncing down the fibre in a controlled fashon it bounces about in a
lot more random fashon and after a metre or three the uncontrolled
light bounces interfere with the main light signal generating extra
noise. This can cause misreading of the recieved signals, hence
errors, noise, distortion.
So by using a better fibre (which is not as cheap) that has a smaller
diameter, will improve the quality (by reducing the noise) at the
recieving end, resulting in less errors etc.

But the same arguement as above comes into play, especially with light,
where the typical pulse length is 150 metres. There is no way more than one
pulse can exist in the cable at once, so how is there any error possible?

I can only think that the testers suffered some sort of placebo
effect, or
they are in cahoots with the industry in order to provide a supply
of
customers who will shell out £350 for the "best" optical cable.

No, what they heard in the test is valid.
Sorry, its just good engineering to fix, bad engineering in the first
place.
As to the question is a cable worth £350 ? Well only your ears, brain
and wallet can make that value judgment.
I personaly would look for a much less expensive option (good dacs are
less than that).

In the end though, should you connect your CD player to your Amp using an
analogue or digital connection (I always assumed digital).

And if it is the latter, is there any point buying an expensive CD player?

snip

Who ever said life was easy :¬)

Indeedily :)

Happy New Year
Chris

"What HiFi" - can it be trusted?
 

......."basically the Fibre call for in the spec is many times larger in
diameter that the wavelenght of the light used so instead on the light
bouncing down the fibre in a controlled fashon it bounces about in a
lot more random fashon and after a metre or three the uncontrolled
light bounces interfere with the main light signal generating extra
noise. This can cause misreading of the recieved signals, hence
errors, noise, distortion.
So by using a better fibre (which is not as cheap) that has a smaller
diameter, will improve the quality (by reducing the noise) at the
recieving end, resulting in less errors etc"............

Mmmmm

Wavelength of light used is?
Diameter of cable used is?

Even the finest fibre is many many times greater in diameter than the
wavelenghth / amplitude of teh transmitted light but that's not relevent!
DIGITAL data is transmitted as a series of pulses.
Light on = 1
Light off = 0.
Interference is insignificant unless you think that some form of lasing is
taking place within the fiber to smear the time base.

I can almost believe that analogue interconnects make a difference (but then
didn't we all throw up our hands in horror at the thought of tone controls
"interfering" with the signal) but I have yet to hear any difference
between digital interconnects given that the plugs and sockets are kept
clean.

Most differences in analogue leads also disapear given a few hours use or a
clean up of the "old" plugs Any lead which sounds significantly different
to a "standard" interconnect is almost certainly "damaging" the sound in
some way - whether by phase shifts or by acting as a filter and I can't see
that any such "change" is a good thing!

Use your ears, get out more, listen to some live music then see whether your
hi-fi sounds real - don't compare it with other hi-fi systems. Most of the
big name set-ups at shows or in dealers demo rooms sound awfull - start from
there and try to improve things from there - and if they even suggest a sub
woofer set the dogs on them!

"chris" wrote in message
...

"Oliver Keating" wrote in
message ...
I was flicking through "What HiFi" magazine and I came across
something very
odd - a series of reviews on *digital* audio cables, for connecting
a CD
player to an amp.

There are comments such as "this cable brings across a crisp sound a
cut
above the rest"

Now, at this point I have to shake my head in disbelief. Surely a
digital
cable about 1 metre long can easily carry a 1mbit data stream with
no
errors. Bear in mind ethernet has to carry 100mbits.


Yes a "DIGITAL cable most certainly can, but a lot of so called
"digital" interconnects arn't made with DIGTAL grade cables, and the
plugs also are not true 75 ohm, so you will start to get line
reflections, ringing on the recieved pulses (and if they are of
sufecent signal level cause pulse signal coruption and drop-outs) this
could start adding to the error rate, resulting in a less "good a
sound".

about optical cables. There is absolutely
*no* way an optical pulse can be distorted sufficiently over 1 metre
that it
could result in an error.

This should be true ! BUT. Unfortunatly due to the design spec this
is not true.
There was some deep discussion on this whole issue a while back on
RAHE,
I too had thought like you. Then The Man from Belden explaind it
rather well: you could do a google on it his expanation should be a
lot better than mine.
But basically the Fibre call for in the spec is many times larger in
diameter that the wavelenght of the light used so instead on the light
bouncing down the fibre in a controlled fashon it bounces about in a
lot more random fashon and after a metre or three the uncontrolled
light bounces interfere with the main light signal generating extra
noise. This can cause misreading of the recieved signals, hence
errors, noise, distortion.
So by using a better fibre (which is not as cheap) that has a smaller
diameter, will improve the quality (by reducing the noise) at the
recieving end, resulting in less errors etc.

I can only think that the testers suffered some sort of placebo
effect, or
they are in cahoots with the industry in order to provide a supply
of
customers who will shell out £350 for the "best" optical cable.

No, what they heard in the test is valid.
Sorry, its just good engineering to fix, bad engineering in the first
place.
As to the question is a cable worth £350 ? Well only your ears, brain
and wallet can make that value judgment.
I personaly would look for a much less expensive option (good dacs are
less than that).

snip

Who ever said life was easy :¬)

Happy New Year
Chris

"What HiFi" - can it be trusted?
 

......."basically the Fibre call for in the spec is many times larger in
diameter that the wavelenght of the light used so instead on the light
bouncing down the fibre in a controlled fashon it bounces about in a
lot more random fashon and after a metre or three the uncontrolled
light bounces interfere with the main light signal generating extra
noise. This can cause misreading of the recieved signals, hence
errors, noise, distortion.
So by using a better fibre (which is not as cheap) that has a smaller
diameter, will improve the quality (by reducing the noise) at the
recieving end, resulting in less errors etc"............

Mmmmm

Wavelength of light used is?
Diameter of cable used is?

Even the finest fibre is many many times greater in diameter than the
wavelenghth / amplitude of teh transmitted light but that's not relevent!
DIGITAL data is transmitted as a series of pulses.
Light on = 1
Light off = 0.
Interference is insignificant unless you think that some form of lasing is
taking place within the fiber to smear the time base.

I can almost believe that analogue interconnects make a difference (but then
didn't we all throw up our hands in horror at the thought of tone controls
"interfering" with the signal) but I have yet to hear any difference
between digital interconnects given that the plugs and sockets are kept
clean.

Most differences in analogue leads also disapear given a few hours use or a
clean up of the "old" plugs Any lead which sounds significantly different
to a "standard" interconnect is almost certainly "damaging" the sound in
some way - whether by phase shifts or by acting as a filter and I can't see
that any such "change" is a good thing!

Use your ears, get out more, listen to some live music then see whether your
hi-fi sounds real - don't compare it with other hi-fi systems. Most of the
big name set-ups at shows or in dealers demo rooms sound awfull - start from
there and try to improve things from there - and if they even suggest a sub
woofer set the dogs on them!

"chris" wrote in message
...

"Oliver Keating" wrote in
message ...
I was flicking through "What HiFi" magazine and I came across
something very
odd - a series of reviews on *digital* audio cables, for connecting
a CD
player to an amp.

There are comments such as "this cable brings across a crisp sound a
cut
above the rest"

Now, at this point I have to shake my head in disbelief. Surely a
digital
cable about 1 metre long can easily carry a 1mbit data stream with
no
errors. Bear in mind ethernet has to carry 100mbits.


Yes a "DIGITAL cable most certainly can, but a lot of so called
"digital" interconnects arn't made with DIGTAL grade cables, and the
plugs also are not true 75 ohm, so you will start to get line
reflections, ringing on the recieved pulses (and if they are of
sufecent signal level cause pulse signal coruption and drop-outs) this
could start adding to the error rate, resulting in a less "good a
sound".

about optical cables. There is absolutely
*no* way an optical pulse can be distorted sufficiently over 1 metre
that it
could result in an error.

This should be true ! BUT. Unfortunatly due to the design spec this
is not true.
There was some deep discussion on this whole issue a while back on
RAHE,
I too had thought like you. Then The Man from Belden explaind it
rather well: you could do a google on it his expanation should be a
lot better than mine.
But basically the Fibre call for in the spec is many times larger in
diameter that the wavelenght of the light used so instead on the light
bouncing down the fibre in a controlled fashon it bounces about in a
lot more random fashon and after a metre or three the uncontrolled
light bounces interfere with the main light signal generating extra
noise. This can cause misreading of the recieved signals, hence
errors, noise, distortion.
So by using a better fibre (which is not as cheap) that has a smaller
diameter, will improve the quality (by reducing the noise) at the
recieving end, resulting in less errors etc.

I can only think that the testers suffered some sort of placebo
effect, or
they are in cahoots with the industry in order to provide a supply
of
customers who will shell out £350 for the "best" optical cable.

No, what they heard in the test is valid.
Sorry, its just good engineering to fix, bad engineering in the first
place.
As to the question is a cable worth £350 ? Well only your ears, brain
and wallet can make that value judgment.
I personaly would look for a much less expensive option (good dacs are
less than that).

snip

Who ever said life was easy :¬)

Happy New Year
Chris

"What HiFi" - can it be trusted?
 

Oliver Keating wrote:

I was flicking through "What HiFi" magazine and I came across something
very odd - a series of reviews on *digital* audio cables, for connecting a
CD player to an amp.

There are comments such as "this cable brings across a crisp sound a cut
above the rest"

Now, at this point I have to shake my head in disbelief. Surely a digital
cable about 1 metre long can easily carry a 1mbit data stream with no
errors. Bear in mind ethernet has to carry 100mbits.

It got even worse when they talked about optical cables. There is
absolutely *no* way an optical pulse can be distorted sufficiently over 1
metre that it could result in an error.

I can only think that the testers suffered some sort of placebo effect, or
they are in cahoots with the industry in order to provide a supply of
customers who will shell out £350 for the "best" optical cable.

Which brings me onto CD players. I always thought that amplifier and
speakers mattered the most, but What HiFi reckons CD players are
important, and worthing spending loads of money on. Now, if you have a CD
player in a half decent Hi-Fi setup then you use a digital interconnect,
so really, all the CD player is having to do is read the raw data off the
CD and feed it to the Amp, and the cleverness of its own DAC is neither
here nor there.

So in a £1,000 CD player are you paying for a great DAC (which you won't
use) or simply some very good error correction in the reading process?

I just wonder if this magazine is just designed so that the industry is
able to flog expensive kit.

Precisely.

Ian

"What HiFi" - can it be trusted?
 

Oliver Keating wrote:

I was flicking through "What HiFi" magazine and I came across something
very odd - a series of reviews on *digital* audio cables, for connecting a
CD player to an amp.

There are comments such as "this cable brings across a crisp sound a cut
above the rest"

Now, at this point I have to shake my head in disbelief. Surely a digital
cable about 1 metre long can easily carry a 1mbit data stream with no
errors. Bear in mind ethernet has to carry 100mbits.

It got even worse when they talked about optical cables. There is
absolutely *no* way an optical pulse can be distorted sufficiently over 1
metre that it could result in an error.

I can only think that the testers suffered some sort of placebo effect, or
they are in cahoots with the industry in order to provide a supply of
customers who will shell out £350 for the "best" optical cable.

Which brings me onto CD players. I always thought that amplifier and
speakers mattered the most, but What HiFi reckons CD players are
important, and worthing spending loads of money on. Now, if you have a CD
player in a half decent Hi-Fi setup then you use a digital interconnect,
so really, all the CD player is having to do is read the raw data off the
CD and feed it to the Amp, and the cleverness of its own DAC is neither
here nor there.

So in a £1,000 CD player are you paying for a great DAC (which you won't
use) or simply some very good error correction in the reading process?

I just wonder if this magazine is just designed so that the industry is
able to flog expensive kit.

Precisely.

Ian

"What HiFi" - can it be trusted?
 

MC does sometimes quote a sort of
'magic number' that he makes up

This is true, of course, but nevertheless he does rank componants which is
frankly the only way to do it. The ear is a very sensitive instrument, the
memory very fallible for intervals of days and weeks between listening tests.
The best we can do for this basic level of evaluation is say 'L sounds better
than M' however fallible the criteria. We can then say 'if G sounds better than
L then it also sounds better than M'. Over time you build up rankings. Audio
Amateur did this, and so did Stereophile I believe. Of course manufacturers
hate it so magazines have dropped the whole thing. Commercial pressure. There
you are. You end up with nonsense like What HiFi - seventeen different products
with five stars and no idea which to buy. At least MC tried, and I thought
some of his ideas were interesting enough and miles better than the rest of the
field.

=== Andy Evans ===
Visit our Website:- http://www.artsandmedia.com
Audio, music and health pages and interesting links.