Hi:

Hypothetical situation: a PCM audio signal [24-bit and monoaural] is
transmitted through an analog 3 Mhz AM carrier, an AM receiver on the
other end [tuned to 3 Mhz] picks up the signal, and the reciever is
attached to a device that can recieve, process, and decode the PCM
audio back to analog and then send it to a loudspeaker. However -- in
this theoretical situation -- the environment is filled with EMI, RFI,
and heterodynes that affect all AM stations.

My question: Will the received PCM audio signal remain noticeably
"clean" to the listener or will he/she notice the EMI, RFI, and
heterodynes affecting the audio?

I ask because I think -- but definitely don't know -- that because the
received signal is digital, it is less likely that the EMI, RFI, and
heterodynes would cause noticeable auditory disruptions when compared
to analog. Do I guess correct?


Thanks,

Radium

Radium wrote:
My question: Will the received PCM audio signal remain noticeably
"clean" to the listener or will he/she notice the EMI, RFI, and
heterodynes affecting the audio?

I ask because I think -- but definitely don't know -- that because the
received signal is digital, it is less likely that the EMI, RFI, and
heterodynes would cause noticeable auditory disruptions when compared
to analog. Do I guess correct?

You guess correct, assuming that FEC is applied to the digital signal
before it is used to modulate the transmitter.

But you wouldn't do it that way anyway. Raw PCM is too bandwidth
inefficient. You'd use MPEG layer 2, or apt-X, or something like that
to reduce the bandwidth without noticeably degrading the audio quality.
You'd probably also multiplex several different channels (programmes)
together onto one RF carrier as well, to make better statistical use of
the RF bandwidth.

Cheers
Mike

Mike Gathergood (G4KFK) wrote:
Radium wrote:
My question: Will the received PCM audio signal remain noticeably
"clean" to the listener or will he/she notice the EMI, RFI, and
heterodynes affecting the audio?

I ask because I think -- but definitely don't know -- that because the
received signal is digital, it is less likely that the EMI, RFI, and
heterodynes would cause noticeable auditory disruptions when compared
to analog. Do I guess correct?

You guess correct,
assuming that FEC is applied to the digital signal
before it is used to modulate the transmitter.

What is FEC?

But you wouldn't do it that way anyway. Raw PCM is too bandwidth
inefficient.

Isn't 3 Mhz enough to transmit a data rate of 1.06 mbps?

44,100 X 24 = 1,058,400

You'd use MPEG layer 2, or apt-X, or something like that
to reduce the bandwidth without noticeably degrading the audio quality.
You'd probably also multiplex several different channels (programmes)
together onto one RF carrier as well, to make better statistical use of
the RF bandwidth.


Cheers
Mike

Radium wrote:
What is FEC?

Forward Error Correction. Google it :-)

Isn't 3 Mhz enough to transmit a data rate of 1.06 mbps?

Yes - but the chipsets to compress the digitised audio are much cheaper
than the notional value of the bandwidth you would be trashing with
your 24bit PCM.

Cheers
Mike

Mike Gathergood (G4KFK) wrote:
Radium wrote:
My question: Will the received PCM audio signal remain noticeably
"clean" to the listener or will he/she notice the EMI, RFI, and
heterodynes affecting the audio?

I ask because I think -- but definitely don't know -- that because the
received signal is digital, it is less likely that the EMI, RFI, and
heterodynes would cause noticeable auditory disruptions when compared
to analog. Do I guess correct?

You guess correct, assuming that FEC is applied to the digital signal
before it is used to modulate the transmitter.

What if FEC is not used?


But you wouldn't do it that way anyway. Raw PCM is too bandwidth
inefficient. You'd use MPEG layer 2, or apt-X, or something like that
to reduce the bandwidth without noticeably degrading the audio quality.
You'd probably also multiplex several different channels (programmes)
together onto one RF carrier as well, to make better statistical use of
the RF bandwidth.

Cheers
Mike

"Mike Gathergood (G4KFK)" wrote in message
oups.com...

Radium wrote:
What is FEC?

Forward Error Correction. Google it :-)


Just to add to Mike's comment, FEC works by send the same message several
times- in simple terms, in the hope that one with get through correctly. It
is simple to implement, your recieving system just needs to be able to
identify a correct message and use it, not spot a bad message and initiate
either a request for resend or applly some sort of correction method
(assuming there is error correcting code in the message).


--
73
Brian
www.g8osn.org.uk

Brian Reay wrote:
"Mike Gathergood (G4KFK)" wrote in message
oups.com...

Radium wrote:
What is FEC?

Forward Error Correction. Google it :-)


Just to add to Mike's comment, FEC works by send the same message several
times- in simple terms, in the hope that one with get through correctly. It
is simple to implement, your recieving system just needs to be able to
identify a correct message and use it, not spot a bad message and initiate
either a request for resend or applly some sort of correction method
(assuming there is error correcting code in the message).


--
73
Brian
www.g8osn.org.uk

These comms are in one direction, so you don't resend a packet as there
is no way to make such a request. In practice, the codes have both
error detection and correction capabilities, so to the degree the
coding allows, the signal can be corrected with the bits that were
received. Note nobody mentioned a modulation scheme for sending this
data.

As a bit of trivia, Reed Solomon encoding was invented without a way to
decode it. That's what you get when you let mathematicians run wild.
For absolutely nothing of any value other than bragging rights, name
the guy who invented the decoding scheme for Reed Solomon. [Hopefully
this isn't wikied someplace. I did one class in grad school on error
detection and correction, and it was a pain in the ass if you get into
the theory. Implementation is quite simple.]

Radium wrote:
Mike Gathergood (G4KFK) wrote:
Radium wrote:
My question: Will the received PCM audio signal remain noticeably
"clean" to the listener or will he/she notice the EMI, RFI, and
heterodynes affecting the audio?

I ask because I think -- but definitely don't know -- that because the
received signal is digital, it is less likely that the EMI, RFI, and
heterodynes would cause noticeable auditory disruptions when compared
to analog. Do I guess correct?

You guess correct, assuming that FEC is applied to the digital signal
before it is used to modulate the transmitter.

What if FEC is not used?

With no FEC, your receiver would be more prone to those errors that you
were worried about in the first place.



But you wouldn't do it that way anyway. Raw PCM is too bandwidth
inefficient. You'd use MPEG layer 2, or apt-X, or something like that
to reduce the bandwidth without noticeably degrading the audio quality.
You'd probably also multiplex several different channels (programmes)
together onto one RF carrier as well, to make better statistical use of
the RF bandwidth.

Just to amplify on this, anything that you can do to reduce the
bandwidth of the digital signal before it's used to modulate the
transmitter will help. One very big advantage is that you can wind down
the bandwidth of the receiver, thus reducing the level of background
noise in the system.

Why did you choose 3MHz? What's the application?

Cheers
Mike

Brian Reay wrote:
Just to add to Mike's comment, FEC works by send the same message several
times- in simple terms, in the hope that one with get through correctly. It
is simple to implement, your recieving system just needs to be able to
identify a correct message and use it, not spot a bad message and initiate
either a request for resend or applly some sort of correction method
(assuming there is error correcting code in the message).

FEC is generally used in applications where you don't have time to
request a retransmission of an errored packet, and/or where the
transmission path is simplex, and/or where there are multiplex
receivers for a single transmitter.

If you have the luxuries of time and a full-duplex point-to-point
environment, ARQ is better.

Cheers
Mike

Mike Gathergood (G4KFK) wrote:
Radium wrote:
Mike Gathergood (G4KFK) wrote:
Radium wrote:
My question: Will the received PCM audio signal remain noticeably
"clean" to the listener or will he/she notice the EMI, RFI, and
heterodynes affecting the audio?

I ask because I think -- but definitely don't know -- that because the
received signal is digital, it is less likely that the EMI, RFI, and
heterodynes would cause noticeable auditory disruptions when compared
to analog. Do I guess correct?

You guess correct, assuming that FEC is applied to the digital signal
before it is used to modulate the transmitter.

What if FEC is not used?


With no FEC, your receiver would be more prone to those errors that you
were worried about in the first place.

What would these errors sound like?




But you wouldn't do it that way anyway. Raw PCM is too bandwidth
inefficient. You'd use MPEG layer 2, or apt-X, or something like that
to reduce the bandwidth without noticeably degrading the audio quality.
You'd probably also multiplex several different channels (programmes)
together onto one RF carrier as well, to make better statistical use of
the RF bandwidth.

Just to amplify on this, anything that you can do to reduce the
bandwidth of the digital signal before it's used to modulate the
transmitter will help. One very big advantage is that you can wind down
the bandwidth of the receiver, thus reducing the level of background
noise in the system.


Why did you choose 3MHz?

44,100 X 24 = 1,058,400

1,058,400 bps requires that the frequency of the carrier be at least
2,646,000 Hz. To make it safe, use 3 MHz.

What's the application?

Well, my application was more to do with reception than transmission.

I'd like to know what I would hear on a 3MHz AM carrier whose receiver
[both the AM and the linear PCM part] is at its maximum bandwidth. The
3 Mhz AM receiver is attached to a linear-PCM receiver [once again,
both receivers have the maximum bandwidth possible for them]. The
linear-PCM receiver is attached to a DAC which converts the linear-PCM
signal to analog. This analog signal [which was PCM] is then sent to a
loudspeaker. Just to make things more interesting, the antennae and
receivers are so sensitive that they can pick signals as low as
..00000001 dB. Most likely, what would I hear?


Cheers
Mike