Feedback Cancellation

[Pat]

Recall that radio stations have call in programmes. During those phone i
programmes a caller/listener would call the radio station from his hous
or vehicle and hear himself on the radio station. This causes feedbac
which is a nuisance to the radio dj on air.

Now I am thinking of coming up with a system that would do away with tha
feedback, regardless of the distance from which the caller is callin
from. Does someone have an idea, please help.

Thank you

[Greg]

On Mon, 24 Jul 2006 08:50:24 -0500, "Pat" <[email protected]> wrote:

>Recall that radio stations have call in programmes. During those phone in
>programmes a caller/listener would call the radio station from his house
>or vehicle and hear himself on the radio station. This causes feedback
>which is a nuisance to the radio dj on air.
>
>Now I am thinking of coming up with a system that would do away with that
>feedback, regardless of the distance from which the caller is calling
>from. Does someone have an idea, please help.
>
>Thank you
>

The problem is not so much feedback as it is delay. The caller is waiting
to hear his voice from the radio, but due to the delay, is confused, or
unsure if he is on the air. When he does hear his voice it out of sync and
that causes further confusion. Consequently he speaks in fits and starts.

A human factors problem, not a DSP problem.

Greg Knox

[Bob]

"Greg" <[email protected]> wrote in message
news:[email protected]...
> On Mon, 24 Jul 2006 08:50:24 -0500, "Pat" <[email protected]> wrote:
>
> >Recall that radio stations have call in programmes. During those phone in
> >programmes a caller/listener would call the radio station from his house
> >or vehicle and hear himself on the radio station. This causes feedback
> >which is a nuisance to the radio dj on air.
> >
> >Now I am thinking of coming up with a system that would do away with that
> >feedback, regardless of the distance from which the caller is calling
> >from. Does someone have an idea, please help.
> >
> >Thank you
> >
>
> The problem is not so much feedback as it is delay. The caller is waiting
> to hear his voice from the radio, but due to the delay, is confused, or
> unsure if he is on the air. When he does hear his voice it out of sync and
> that causes further confusion. Consequently he speaks in fits and starts.
>
> A human factors problem, not a DSP problem.
>
> Greg Knox
>

True. Delay is the problem. However, if it was feedback, this little box
would be a COTS solution:
http://www.behringer.com/DSP1124P/index.cfm?lang=ENG
I picked one up a couple of months ago to kill a terrible resonance problem
in my sound system (in the bathroom <g>). Of course, first thing out of the
box, we put it on an audio FFT analyzer to check S/N, distortion, and filter
shapes. We started scratching our heads about why the output level was so
low: we were putting in a sine at 1Vp-p and only getting 8mV out. WTF?
Answer: it defaults to "feedback destroyer" mode. It saw our sine wave as a
feedback problem and did its best to destroy it. :-) Not bad for less than
$100 US.

Bob
the bathroom sounds GREAT!

[glen herrmannsfeldt]

Pat wrote:

> Recall that radio stations have call in programmes. During those phone in
> programmes a caller/listener would call the radio station from his house
> or vehicle and hear himself on the radio station. This causes feedback
> which is a nuisance to the radio dj on air.

I believe they will always have a delay long enough to keep bad words
off the air. Still, one must keep the echo off the air.

> Now I am thinking of coming up with a system that would do away with that
> feedback, regardless of the distance from which the caller is calling
> from. Does someone have an idea, please help.

I once knew someone working on a system using double balanced mixers
with something like a 1Hz mixer input. I never heard the results, but
I don't remember that it worked very well. If you arrange it to add 1Hz
to any signal coming in, that would affect the feedback, and maybe not
change a voice signal too much.

-- glen

[Mad Prof]

"glen herrmannsfeldt" <[email protected]> wrote in message
news:[email protected]. ..
> Pat wrote:
>
> > Recall that radio stations have call in programmes. During those phone
in
> > programmes a caller/listener would call the radio station from his house
> > or vehicle and hear himself on the radio station. This causes feedback
> > which is a nuisance to the radio dj on air.
>
> I believe they will always have a delay long enough to keep bad words
> off the air. Still, one must keep the echo off the air.
>
> > Now I am thinking of coming up with a system that would do away with
that
> > feedback, regardless of the distance from which the caller is calling
> > from. Does someone have an idea, please help.
>
> I once knew someone working on a system using double balanced mixers
> with something like a 1Hz mixer input. I never heard the results, but
> I don't remember that it worked very well. If you arrange it to add 1Hz
> to any signal coming in, that would affect the feedback, and maybe not
> change a voice signal too much.
>
> -- glen
>

All you need to do is phase -advance the signal.

M.P



--
Posted via a free Usenet account from http://www.teranews.com

[Jerry Avins]

glen herrmannsfeldt wrote:

> I once knew someone working on a system using double balanced mixers
> with something like a 1Hz mixer input. I never heard the results, but
> I don't remember that it worked very well. If you arrange it to add 1Hz
> to any signal coming in, that would affect the feedback, and maybe not
> change a voice signal too much.

Up to 5 Hz is OK for most voices. I built a tube version that worked
using filtering, not phasing. The biggest problem was keeping the up-
and down-converters from pulling one another. With the best power-supply
filtering and other isolation I could manage, 2 HZ was impossible. # Hz
worked, but I didn't trust it for long-term stability. 4 Hz was solid,
but 5 Hz gave better feedback-suppression and still sounded all right,
so I used it.

Home Depot and others use a different system altogether. They have a
digital recording chip in the PA connected to the phone, and nothing
comes out of the speaker until the phone is hung up.

Jerry
--
Engineering is the art of making what you want from things you can get.
¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯ ¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯

[Pat]

Thank you ladies and gentlemen for your responses.

Well I will try the adding of 1Hz to 5Hz and also will try the phasin
ideas. In this case all ideas are worth trying out. As for the proble
being delay, I have heard feedback coming out when a caller calls th
radio station while the volume of their radio receiver is high. That's m
concern.

Some callers can call from long distances while some would call fro
within the same town with the radio station, how do I manage the feedbac
control and callers are from different places, say the distances of up t
1600km. We know the delays of telephone lines and microwaves depend on th
distance. Theproject is manageable for the same distances like in
conference room but I am stuck in this case.

I thank you once again

[Jerry Avins]

Jerry Avins wrote:


> ... With the best power-supply
> filtering and other isolation I could manage, 2 HZ was impossible. # Hz
> worked, but I didn't trust it for long-term stability. 4 Hz was solid,
> but 5 Hz gave better feedback-suppression and still sounded all right,
> so I used it.

Note that # = <shift>3 :-)

Jerry
--
Engineering is the art of making what you want from things you can get.
¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯ ¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯

[Jerry Avins]

Pat wrote:
> Thank you ladies and gentlemen for your responses.
>
> Well I will try the adding of 1Hz to 5Hz and also will try the phasing
> ideas. In this case all ideas are worth trying out. As for the problem
> being delay, I have heard feedback coming out when a caller calls the
> radio station while the volume of their radio receiver is high. That's my
> concern.
>
> Some callers can call from long distances while some would call from
> within the same town with the radio station, how do I manage the feedback
> control and callers are from different places, say the distances of up to
> 1600km. We know the delays of telephone lines and microwaves depend on the
> distance. Theproject is manageable for the same distances like in a
> conference room but I am stuck in this case.
>
> I thank you once again

The feedback that disturbs you is not entirely the same as what limits
the gain in a PA system. Many advances have happened since I built the
frequency shifter more than 50 years ago, and now there are better ways.
Most talk-show hosts remind callers to turn their radio volume down, and
that seems to suffice. Can you tell us what specific problem needs to be
addressed?

Jerry
--
Engineering is the art of making what you want from things you can get.
¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯ ¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯

[Pat]

">Pat wrote:
>> Thank you ladies and gentlemen for your responses.
>>
>> Well I will try the adding of 1Hz to 5Hz and also will try the phasing
>> ideas. In this case all ideas are worth trying out. As for the problem
>> being delay, I have heard feedback coming out when a caller calls the
>> radio station while the volume of their radio receiver is high. That'
my
>> concern.
>>
>> Some callers can call from long distances while some would call from
>> within the same town with the radio station, how do I manage th
feedback
>> control and callers are from different places, say the distances of u
to
>> 1600km. We know the delays of telephone lines and microwaves depend o
the
>> distance. Theproject is manageable for the same distances like in a
>> conference room but I am stuck in this case.
>>
>> I thank you once again
>
>The feedback that disturbs you is not entirely the same as what limits
>the gain in a PA system. Many advances have happened since I built the
>frequency shifter more than 50 years ago, and now there are better ways

>Most talk-show hosts remind callers to turn their radio volume down, an

>that seems to suffice. Can you tell us what specific problem needs to b

>addressed?
>
>Jerry
>--
>Engineering is the art of making what you want from things you can get."



My objective is to completely do away with the talk show hosts asking th
caller to turn down their volume. I want the caller to be free to talk an
have their volume relatively up without causing negative feedbak.

Thank you

[Pat]

">Pat wrote:
>> Thank you ladies and gentlemen for your responses.
>>
>> Well I will try the adding of 1Hz to 5Hz and also will try the phasing
>> ideas. In this case all ideas are worth trying out. As for the problem
>> being delay, I have heard feedback coming out when a caller calls the
>> radio station while the volume of their radio receiver is high. That'
my
>> concern.
>>
>> Some callers can call from long distances while some would call from
>> within the same town with the radio station, how do I manage th
feedback
>> control and callers are from different places, say the distances of u
to
>> 1600km. We know the delays of telephone lines and microwaves depend o
the
>> distance. Theproject is manageable for the same distances like in a
>> conference room but I am stuck in this case.
>>
>> I thank you once again
>
>The feedback that disturbs you is not entirely the same as what limits
>the gain in a PA system. Many advances have happened since I built the
>frequency shifter more than 50 years ago, and now there are better ways

>Most talk-show hosts remind callers to turn their radio volume down, an

>that seems to suffice. Can you tell us what specific problem needs to b

>addressed?
>
>Jerry
>--
>Engineering is the art of making what you want from things you can get."



My objective is to completely do away with the talk show hosts asking th
caller to turn down their volume. I want the caller to be free to talk an
have their volume relatively up without causing negative feedbak.

Thank you

[Bob]

"Pat" <[email protected]> wrote in message
news:[email protected]...
>
> My objective is to completely do away with the talk show hosts asking the
> caller to turn down their volume. I want the caller to be free to talk and
> have their volume relatively up without causing negative feedbak.
>
> Thank you
>
Pat,
Why re-invent the wheel? A solution already exists. As I mentioned elsewhere
in this thread, you can buy a box for <$100 US that contains 12 banks * 2
channels (stereo) of parametrically defined filters. The primary purpose of
the box is to suppress narrowband energy that rises above a (user settable)
threshold of the total energy in the signal. In its default mode, it
automatically detects these feedback tones and quickly configures an
extremely narrow notch filter to kill the system gain at the offending
frequency. The effect on the overall audio quality is said to be nearly
imperceptible because the filter is only a few cents (100ths of a semitone)
wide. This is one brand; there may be others:

http://www.behringer.com/DSP1124P/index.cfm?lang=ENG

The box is designed for sound reinforcement (concerts) but I see no reason
it wouldn't work for what you want (live radio call-in).

Bob

[Jerry Avins]

Pat wrote:

...

> My objective is to completely do away with the talk show hosts asking the
> caller to turn down their volume. I want the caller to be free to talk and
> have their volume relatively up without causing negative feedbak.

Negative feedback is no problem at all. The problems arise from positive
feedback and disorienting delay. Many "live" talk shows use a delay of
several seconds to allow offensive words to be bleeped out. I don't see
how you can get around that.

Jerry
--
Engineering is the art of making what you want from things you can get.
¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯ ¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯

[Pat]

>Pat wrote:
>
> ...
>
>> My objective is to completely do away with the talk show hosts askin
the
>> caller to turn down their volume. I want the caller to be free to tal
and
>> have their volume relatively up without causing negative feedbak.
>
>Negative feedback is no problem at all. The problems arise from positiv

>feedback and disorienting delay. Many "live" talk shows use a delay of
>several seconds to allow offensive words to be bleeped out. I don't see
>how you can get around that.
>
>Jerry
>--
>Engineering is the art of making what you want from things you can get.
>¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯ ¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯

Thank you very much. I want to come up with a design of my own. If th
wheel was invented already then it's pointless to reinvent it. Man
publications that I have read do not have a solution to that problem.