HiFi vs. Computer

[Guest]

Archived from groups: rec.audio.tech (More info?)

In article <11a4o9593r5eq73@corp.supernews.com>,
Richard Crowley <rcrowley7@xprt.net> wrote:
>>
>> Suppose thereare two birds silent, and both start singing.
>>
>> What do you think is the minimum detectable time difference between the
>> start of the birds' songs? (That is, if the difference in time is greater
>> than this interval, we can determine which bird started singing first,
>> otherwise we are unable to determine which started singing first.)
>
>Picoseconds or even femtoseconds. But likely not reliable at that
>resolution in a real forrest with noise and air movement.

Picosecond resolution of time difference would require that the birds'
song have significant frequency content up in the GHz range. I doubt
that it does.

--
Dave Platt <dplatt@radagast.org> AE6EO
Hosting the Jade Warrior home page: http/www.radagast.org/jade-warrior
I do _not_ wish to receive unsolicited commercial email, and I will
boycott any company which has the gall to send me such ads!

 

[Guest]

Archived from groups: rec.audio.tech (More info?)

"Dave Platt" <dplatt@radagast.org> wrote in message
news:11a4rcqbf3tcs60@corp.supernews.com...
> In article <11a4o9593r5eq73@corp.supernews.com>,
> Richard Crowley <rcrowley7@xprt.net> wrote:
>>>
>>> Suppose thereare two birds silent, and both start singing.
>>>
>>> What do you think is the minimum detectable time difference between
>>> the
>>> start of the birds' songs? (That is, if the difference in time is
>>> greater
>>> than this interval, we can determine which bird started singing
>>> first,
>>> otherwise we are unable to determine which started singing first.)
>>
>>Picoseconds or even femtoseconds. But likely not reliable at that
>>resolution in a real forrest with noise and air movement.
>
> Picosecond resolution of time difference would require that the birds'
> song have significant frequency content up in the GHz range. I doubt
> that it does.

Are you assuming we are using only one microphone for both
birds?

 

[Guest]

Archived from groups: rec.audio.tech (More info?)

On Sat, 04 Jun 2005 19:43:05 GMT, donald@pearce.uk.com (Don Pearce)
wrote:

>On Sat, 4 Jun 2005 11:56:25 +0000 (UTC), Stewart Pinkerton
><patent3@dircon.co.uk> wrote:
>
>>On Sat, 04 Jun 2005 12:04:25 +0100, Laurence Payne
>><lp@laurenceNOSPAMpayne.freeserve.co.uk> wrote:
>>
>>>On Wed, 01 Jun 2005 20:59:25 -0400, Joe Kesselman
>>><keshlam-nospam@comcast.net> wrote:
>>>
>>>>Well, yes, analog is _theoretically_ infinite precision.
>>>
>>>Only if your theory allows zero noise and infinite signal amplitude.
>>>That's VERY theoretical
>>
>>And could only happen at absolute zero, which wouldn't be good for
>>your vinyl................
>
>Not even at absolute zero. No escaping the old zero point energy I'm
>afraid - Heisenberg will not be argued with.

Or perhaps he will, I'm uncertain.........
--

Stewart Pinkerton | Music is Art - Audio is Engineering

 

[Guest]

Archived from groups: rec.audio.tech (More info?)

On Sat, 04 Jun 2005 22:45:34 GMT, "Tim Martin"
<tim2718281@ntlworld.com> wrote:

>
>"Stewart Pinkerton" <patent3@dircon.co.uk> wrote in message
>news:7m53a1hqvjcmqec6g4ad5cpmafe3vjnti2@4ax.com...
>
> >Only if your theory allows zero noise and infinite signal amplitude.
>> >That's VERY theoretical
>>
>> And could only happen at absolute zero, which wouldn't be good for
>> your vinyl................
>
>Suppose thereare two birds silent, and both start singing.
>
>What do you think is the minimum detectable time difference between the
>start of the birds' songs? (That is, if the difference in time is greater
>than this interval, we can determine which bird started singing first,
>otherwise we are unable to determine which started singing first.)

That would depend on the bandwidth of the detector, and is in any
event a pointless question, since the birds are not coincident. This
being so, the orientation of the detection point vis a vis the two
sound sorces will determine which birdsong is heard first. The error
caused by this is significantly higher than that caused by any actual
transient delay. Sound travels at roughly 1135 m/sec at sea level and
20 degrees Celsius, so a 1/2 inch difference in position of the birds
will cause a +/1 10 microsecond error in the timing. This is pretty
close to the sampling interval of CD................
--

Stewart Pinkerton | Music is Art - Audio is Engineering

 

[Guest]

Archived from groups: rec.audio.tech (More info?)

On Sun, 5 Jun 2005 07:55:11 +0000 (UTC), Stewart Pinkerton
<patent3@dircon.co.uk> wrote:

>On Sat, 04 Jun 2005 19:43:05 GMT, donald@pearce.uk.com (Don Pearce)
>wrote:
>
>>On Sat, 4 Jun 2005 11:56:25 +0000 (UTC), Stewart Pinkerton
>><patent3@dircon.co.uk> wrote:
>>
>>>On Sat, 04 Jun 2005 12:04:25 +0100, Laurence Payne
>>><lp@laurenceNOSPAMpayne.freeserve.co.uk> wrote:
>>>
>>>>On Wed, 01 Jun 2005 20:59:25 -0400, Joe Kesselman
>>>><keshlam-nospam@comcast.net> wrote:
>>>>
>>>>>Well, yes, analog is _theoretically_ infinite precision.
>>>>
>>>>Only if your theory allows zero noise and infinite signal amplitude.
>>>>That's VERY theoretical
>>>
>>>And could only happen at absolute zero, which wouldn't be good for
>>>your vinyl................
>>
>>Not even at absolute zero. No escaping the old zero point energy I'm
>>afraid - Heisenberg will not be argued with.
>
>Or perhaps he will, I'm uncertain.........

In principle, maybe.

d

Pearce Consulting
http/www.pearce.uk.com

 

[Guest]

Archived from groups: rec.audio.tech (More info?)

"Stewart Pinkerton" <patent3@dircon.co.uk> wrote in message
news:81c5a15kmk2hu98vtpkaa3jhmhkuajmlks@4ax.com...

> Sound travels at roughly 1135 m/sec at sea level and
> 20 degrees Celsius, so a 1/2 inch difference in position of the birds
> will cause a +/1 10 microsecond error in the timing. This is pretty
> close to the sampling interval of CD................

Typo there, you mean feet a second. So if you're detecting singing by an
acoustic instrument placed some distance from each bird, an error of half an
inch in the measurement of the distance of each instrument from its bird
will cause a 50 microsecond error in the measurement of the start of the
birds' singing.

Is that the greatest error in the measurement?

Tim

 

[Guest]

Archived from groups: rec.audio.tech (More info?)

On Sun, 05 Jun 2005 09:38:40 GMT, "Tim Martin"
<tim2718281@ntlworld.com> wrote:

>
>"Stewart Pinkerton" <patent3@dircon.co.uk> wrote in message
>news:81c5a15kmk2hu98vtpkaa3jhmhkuajmlks@4ax.com...
>
>> Sound travels at roughly 1135 m/sec at sea level and
>> 20 degrees Celsius, so a 1/2 inch difference in position of the birds
>> will cause a +/1 10 microsecond error in the timing. This is pretty
>> close to the sampling interval of CD................
>
>Typo there, you mean feet a second.

Ooops, yes, 345 m/sec, much easier to remember!

> So if you're detecting singing by an
>acoustic instrument placed some distance from each bird, an error of half an
>inch in the measurement of the distance of each instrument from its bird
>will cause a 50 microsecond error in the measurement of the start of the
>birds' singing.

Pretty much.

>Is that the greatest error in the measurement?

Yes, since everything else can be measured to very high precision.
--

Stewart Pinkerton | Music is Art - Audio is Engineering

 

[Guest]

Archived from groups: rec.audio.tech (More info?)

"Stewart Pinkerton" <patent3@dircon.co.uk> wrote in message
news:rud6a15ll777b19bqufft1mpito560onm2@4ax.com...

Tim
> >Is that the greatest error in the measurement?
>
Stewart
> Yes, since everything else can be measured to very high precision.
> --

Mmm. Suppose you have six measuring devices arranged in three pairs; the
devices in each pair are on opposite sides of the bird, and the three lines
joining the pairs are orthogonal. You calculate the times and average them.

What effect will that have on the distance measurement error?

Tim

 

[Guest]

Archived from groups: rec.audio.tech (More info?)

On Sun, 05 Jun 2005 22:37:25 GMT, "Tim Martin"
<tim2718281@ntlworld.com> wrote:

>"Stewart Pinkerton" <patent3@dircon.co.uk> wrote in message
>news:rud6a15ll777b19bqufft1mpito560onm2@4ax.com...
>
>Tim
>> >Is that the greatest error in the measurement?
>>
>Stewart
>> Yes, since everything else can be measured to very high precision.
>> --
>Mmm. Suppose you have six measuring devices arranged in three pairs; the
>devices in each pair are on opposite sides of the bird, and the three lines
>joining the pairs are orthogonal. You calculate the times and average them.
>
>What effect will that have on the distance measurement error?

Who cares? Despite all your sidestepping, you can't have infinite
resolution or infinite bandwidth for *any* natural sonic event. And
certainly not in an electrical signal which is an analogue of that
event.
--

Stewart Pinkerton | Music is Art - Audio is Engineering

 

[Guest]

Archived from groups: rec.audio.tech (More info?)

Laurence Payne wrote:
> On Fri, 03 Jun 2005 02:19:12 GMT, "Tim Martin"
> <tim2718281@ntlworld.com> wrote:
>
>> An analog signal, such as a bird singing in the woods,
has infinite
>> bandwidth.
>
> No it isn't. It is constrained by all sorts of physical
limitations,
> both in the bird and in the environment.

Agreed. The spectral content of most natural sounds will
reveal the inevitable high end roll-off if measurered with a
mic capable of going out to about 50 KHz. In a few cases one
must go out to a few 100 KHz.

 

[Guest]

Archived from groups: rec.audio.tech (More info?)

"Stewart Pinkerton" <patent3@dircon.co.uk> wrote in message
news:k2o7a19j58medqs2ifv208b2talep9uk2u@4ax.com...

> Who cares?

Well, you did, You said the distance measurement was the limiting factor;
and I've given a way to reduce the error in distance measurement. So
presumably you now accept that distant measurement errors would be lower
than the 50 microseconds or so resulting from a 0.5-inch error in distance
measurement.

Tim

 

[Guest]

Archived from groups: rec.audio.tech (More info?)

Eiron wrote:

> Ben Bradley wrote:
>
>>>> From what I've seen on your posts, I think you need to forget about the
>>>> computer for a second. Start with getting an understanding of how a
>>>> square wave is a gazzilion sine waves, all slightly out of phase from
>>>> each other, summed together. Then you'll be off to a good start.
>>>
>>>
>>> They're not out of phase with each other, they are in phase..
>>
>>
>>
>> The phase doesn't even matter, except to that Gibbs guy...
>
>
> And it's not a gazillion sine waves. You would be surprised at how few
> harmonics are required to make it indistinguishable (to the ear) from a
> real square wave.
>
My assumption was that the sine waves were all of one frequency, say
1000 Hz, hence the need for all of them to be out of phase, and that you
would see the resultant square wave in an oscilloscope, therefore the
need for a whole lot of them needed to make the final square wave that
is seen.

CD

 

[Guest]

Archived from groups: rec.audio.tech (More info?)

On Mon, 06 Jun 2005 12:34:06 GMT, "Tim Martin"
<tim2718281@ntlworld.com> wrote:

>
>"Stewart Pinkerton" <patent3@dircon.co.uk> wrote in message
>news:k2o7a19j58medqs2ifv208b2talep9uk2u@4ax.com...
>
>> Who cares?
>
>Well, you did, You said the distance measurement was the limiting factor;

That doesn't mean I *cared* about it!

>and I've given a way to reduce the error in distance measurement. So
>presumably you now accept that distant measurement errors would be lower
>than the 50 microseconds or so resulting from a 0.5-inch error in distance
>measurement.

So what? This is mere obfuscation. What matters is that your previous
pronouncements concerning dynamic range and bandwidth of natural
sounds, are utter rubbish.
--

Stewart Pinkerton | Music is Art - Audio is Engineering

 

[Guest]

Archived from groups: rec.audio.tech (More info?)

"Stewart Pinkerton" <patent3@dircon.co.uk> wrote in message
news:m5s8a15kac1dg66r1sjls8hlq2sppvssqp@4ax.com...

> So what? This is mere obfuscation. What matters is that your previous
> pronouncements concerning dynamic range and bandwidth of natural
> sounds, are utter rubbish.

I said nothing about dynamic range.

And I suspect it has at last dawned on you that the time resolution of an
analog signal is not limited in the same way as the time resolution of a
sampled representation of that analog signal. So for non-repeating signals,
the frequencies being represented are not the only factor in determining the
sampling rate required to represent them precisely.

Tim

 

[Guest]

Archived from groups: rec.audio.tech (More info?)

On Mon, 06 Jun 2005 17:13:27 GMT, "Tim Martin"
<tim2718281@ntlworld.com> wrote:

>"Stewart Pinkerton" <patent3@dircon.co.uk> wrote in message
>news:m5s8a15kac1dg66r1sjls8hlq2sppvssqp@4ax.com...
>
>> So what? This is mere obfuscation. What matters is that your previous
>> pronouncements concerning dynamic range and bandwidth of natural
>> sounds, are utter rubbish.
>
>I said nothing about dynamic range.

You don't even understand your own arguments, do you? You talked about
signal, and about noise. Dynamic range is the difference between peak
signal and noise floor.

>And I suspect it has at last dawned on you that the time resolution of an
>analog signal is not limited in the same way as the time resolution of a
>sampled representation of that analog signal.

How many times do you have to be told that the original sound is *not*
an analogue - it is an *event*? You also seem to be unable to
understand that the original event most definitely *is* limited in
bandwidth, hence in 'time resolution', and in *exactly* the same way
as any sampling system is limited - by physical reality.

> So for non-repeating signals,
>the frequencies being represented are not the only factor in determining the
>sampling rate required to represent them precisely.

Yes, they are. Go *learn* about the subject, and stop spouting this
schoolboy rubbish.
--

Stewart Pinkerton | Music is Art - Audio is Engineering

 

[Guest]

Archived from groups: rec.audio.tech (More info?)

On Mon, 06 Jun 2005 15:26:58 -0400, Codifus <codifus@optonline.net>
wrote:

>Eiron wrote:
>
>> Ben Bradley wrote:
>>
>>>>> From what I've seen on your posts, I think you need to forget about the
>>>>> computer for a second. Start with getting an understanding of how a
>>>>> square wave is a gazzilion sine waves, all slightly out of phase from
>>>>> each other, summed together. Then you'll be off to a good start.
>>>>
>>>>
>>>> They're not out of phase with each other, they are in phase..
>>>
>>>
>>>
>>> The phase doesn't even matter, except to that Gibbs guy...
>>
>>
>> And it's not a gazillion sine waves. You would be surprised at how few
>> harmonics are required to make it indistinguishable (to the ear) from a
>> real square wave.
>>
>My assumption was that the sine waves were all of one frequency, say
>1000 Hz, hence the need for all of them to be out of phase, and that you
>would see the resultant square wave in an oscilloscope, therefore the
>need for a whole lot of them needed to make the final square wave that
>is seen.

That's not how it works. A square wave may be represented by a
fundamental sine wave at the prf of the square wave, plus a
theoretically infinite number of odd harmonics arranged in a
mathematically precise descending order of amplitude. All those sine
waves are in phase.
--

Stewart Pinkerton | Music is Art - Audio is Engineering

 

[Guest]

Archived from groups: rec.audio.tech (More info?)

Tim Martin wrote:
> "Stewart Pinkerton" <patent3@dircon.co.uk> wrote in
message
> news:m5s8a15kac1dg66r1sjls8hlq2sppvssqp@4ax.com...
>
>> So what? This is mere obfuscation. What matters is that
your previous
>> pronouncements concerning dynamic range and bandwidth of
natural
>> sounds, are utter rubbish.
>
> I said nothing about dynamic range.
>
> And I suspect it has at last dawned on you that the time
resolution
> of an analog signal is not limited in the same way as the
time
> resolution of a sampled representation of that analog
signal.

That's your big error Tim. The resolution of a sample is
limited by the size of the sample, whether analog or
digital.

> So for non-repeating signals, the frequencies being
represented are not the
> only factor in determining the sampling rate required to
represent them precisely.

What you don't seem to realize Tim is the fact that all of
the limiations that you've been obsessing over relate to
both analog and digital signals.

 

[Guest]

Archived from groups: rec.audio.tech (More info?)

"Stewart Pinkerton" <patent3@dircon.co.uk> wrote in message
news:itbaa1tt16evbb2mi5ue5so8pvnv9ubsbo@4ax.com...

> You don't even understand your own arguments, do you? You talked about
> signal, and about noise.

I did say that what Arny referred to as the "background noise" of the woods
was part of the signal. And I did say that I avoided assuming the use of a
microphones as the measurement device, because microphones are connected to
electrical circuits and so introduce noise.

I don't see how either of those comments can be interpreted as a remark
about dynamic range.

Tim

 

[Guest]

Archived from groups: rec.audio.tech (More info?)

"Arny Krueger" <arnyk@hotpop.com> wrote in message
news:B5-dnaJC2MZQ4jjfRVn-rg@comcast.com...

> What you don't seem to realize Tim is the fact that all of
> the limiations that you've been obsessing over relate to
> both analog and digital signals.

That's incorrect: analog signals are continuous. Between any two points in
time where the analog signal is changing, one can find another point in time
where the value of the analog signal differs from its values at either of
the first two points in time.

This is not true for digital representations of analog signals.

Of course, there are limits in how accurately one can represent analog
signals, whether the representation is digital or analog. But the limits
are in communication channels and storage schemes, not in the signals
themselves.

Tim

 

[Guest]

Archived from groups: rec.audio.tech (More info?)

"Tim Martin" wrote ...
> That's incorrect: analog signals are continuous. Between any two
> points in
> time where the analog signal is changing, one can find another point
> in time
> where the value of the analog signal differs from its values at either
> of
> the first two points in time.

Doesn't make sense.

> This is not true for digital representations of analog signals.

Wouldn't be true even if the previous statement did make sense.