-
Join the MyBroadband community
South Africa’s biggest forum. Discuss, discover, and connect with thousands of members.
-
Question of the Day: Did you buy your last car with cash or on a finance plan?
Audio Enhancement
- Thread starter WilmaG
- Start date
100dB gain with what noise floor? For example the dynamic range of CD is about 90dB and considered very high.
You get noise filtering, band limiting, side-chain EQs and specialist techniques to try and decode voice, but if you really require 100dBs of gain you will most likely not retrieve intelligible signal. The human ear can decode some things below a noise floor, but there are limits.
Noise cancellation algorithms introduce their own noise and you require a snippet of the environmental noise to begin with.
You get noise filtering, band limiting, side-chain EQs and specialist techniques to try and decode voice, but if you really require 100dBs of gain you will most likely not retrieve intelligible signal. The human ear can decode some things below a noise floor, but there are limits.
Noise cancellation algorithms introduce their own noise and you require a snippet of the environmental noise to begin with.
With what was it recorded? What Timber_MG is saying is, it is unlikely that you will recover it if you need 100dB gain, the theoretical max dynamic range for 16-bit recordings being ~96dB. If you cannot filter the conversation out with your ears, a computer will do no better.
copacetic
King of the Hippies
Can you provide a snippet of the audio? It's virtually impossible to offer advice without hearing anything.
A copy of the audio is here
https://docs.google.com/uc?export=download&id=0B5ingIYMt2rmWnFuZHI5OUZwQ1E
There is a lot of noise!
The actual conversation starts around 14 seconds.
https://docs.google.com/uc?export=download&id=0B5ingIYMt2rmWnFuZHI5OUZwQ1E
There is a lot of noise!
The actual conversation starts around 14 seconds.
copacetic
King of the Hippies
Yoh. I don't know if you are going to be able to extract anything useful out of that. On my ****ty headphones here I cannot even hear a hint of voice...
I'm not expert, but I'll take a look at home later.
I'm not expert, but I'll take a look at home later.
Electric
Honorary Master
You should give Reafir a try.
It's a Reaper plugin but freely downloadable as a vst.
You use a sample of the noise which the algorithm uses to remove similar noise spectrums.
[video=youtube;GOh04O6p6Vw]http://www.youtube.com/watch?v=GOh04O6p6Vw[/video]
http://reaperblog.net/2012/05/video-removing-noise-with-reafir/
http://forum.cockos.com/showthread.php?t=31762
It's a Reaper plugin but freely downloadable as a vst.
You use a sample of the noise which the algorithm uses to remove similar noise spectrums.
[video=youtube;GOh04O6p6Vw]http://www.youtube.com/watch?v=GOh04O6p6Vw[/video]
http://reaperblog.net/2012/05/video-removing-noise-with-reafir/
http://forum.cockos.com/showthread.php?t=31762
copacetic
King of the Hippies
That was precisely what I was going to try, hehe.
Electric
Honorary Master
Are you just here marketing a product?
Because it seems like you just baited us into responding and you in turn replied with a marketing ploy.
copacetic
King of the Hippies
Okay...
copacetic
King of the Hippies
Saying it's not impossible is incorrect - There is going to be a point where the degradation and background noise is such that it is impossible to extract enough meaningful audio to create an understandable signal.
Whether your clip is at this point is impossible to say right now.
And obviously, depending on the level of expertise and equipment involved, this point is closer or further away...
Whether your clip is at this point is impossible to say right now.
And obviously, depending on the level of expertise and equipment involved, this point is closer or further away...
Electric
Honorary Master
Of course it's not impossible but that is also to a point like copacetic says above.
I gave you a free alternative that works though.
Otherwise, go buy the software you linked.
Thanks for the assistance.
I appreciate it.
But I'm convinced there's audio in that recording.
The person spying on me is going to extreme lengths not to get caught.
The signal is passed down the telephone line very faintly.
I read this interesting commentary on the net:
There is actually another methodology which can be applied to
eavesdropping on room conversations using an unmodified telephone set.
Most ringers will function as a variable reluctance microphone, if the
line from the telephone is amplified to an extreme degree, along with
application of suitable signal processing to eliminate an incredible
amount of noise. As in the above methods, the necessary apparatus
must be within a few hundred feet from the telephone set, and the CO
pair must be broken during the operation (with circuitry to detect an
incoming call or outgoing call attempt and reestablish the CO line
continuity to avoid any suspicion on the part of the subject). I am
not claiming that a ringer is a *good* microphone, but under some
selected circumstances this technique can provide useful intelligence.
I may later regret this suggestion, but as an example to
illustrate this principle, here is an experiment that an enterprising
reader can perform using apparatus found in any well-equipped
electronics laboratory. Take a 500-type or 2500-type set with a
bridged ringer and connect its tip and ring directly to the input of a
low-noise amplifier providing say, 80 dB of gain in the voice
frequency range. A suggested approach is to cascade two
Hewlett-Packard 465A amplifiers, with each amplifier being set for 40
dB gain. Take the 80 dB amplifier output and connect it to the input
of a variable bandpass filter having at least 20 db/octave attenuation
(like a Kron-Hite 3100, 3500 or 3700). Take the output from the
bandpass filter and feed it to another amplifier providing 20 to 40 dB
gain and capable of driving a pair of headphones.
Tune the bandpass filter to reject powerline noise, and you have just
turned the telephone set into a crude microphone. At that point it
does not take much imagination to realize that given some competent
engineering resources and a commensurate budget, this technique can be
refined into a practicable eavesdropping device. The availability of
digital signal processing can also do wonders to eliminate the vast
amount of power line, impulse noise and other interference which
develops at the gain necessary for speech pickup sensitivity.
While electromechanical ringers are becoming somewhat a thing
of the past, many electronic telephone sets with tone ringers will
function as an even better microphone. Such tone ringers usually rely
upon a piezoelectric element as the loudspeaker, although a few
low-quality "drugstore-variety" one-piece telephones utilize the
receiver element as the ringer transducer. As most readers of this
forum are no doubt aware, piezoelectric devices will generally
function as both a microphone and loudspeaker. Even a piezoelectric
element optimized for tone ringer use, i.e., with resonance in the
range of 1.5 to 2.5 kHz, will still function as a usable microphone
for lower frequencies.
An on-hook telephone set with electronic tone ringer, if
isolated from the CO line and connected to an ultra-high gain
amplifier with suitable bandpass filtering, and if also subjected to
an appropriate RF bias to cause conduction across the initial
full-wave bridge rectifier and subsequent semiconductor junctions, can
in many instances be turned into a microphone. While this technique
will not work with all electronic telephones, it will work with a
significant number.
The above technique of compromising a telephone with an
electronic tone ringer was first performed almost twenty years ago on
the Ericophone. The Ericophone was an early one-piece telephone, some
models of which contained an electronic tone ringer. While the
geometry of the Ericophone defies verbal description in this forum,
the overall design scheme may best be described as phallic in nature.
Those readers who are familiar with the Ericophone will no doubt concur with this description
."
I appreciate it.
But I'm convinced there's audio in that recording.
The person spying on me is going to extreme lengths not to get caught.
The signal is passed down the telephone line very faintly.
I read this interesting commentary on the net:
There is actually another methodology which can be applied to
eavesdropping on room conversations using an unmodified telephone set.
Most ringers will function as a variable reluctance microphone, if the
line from the telephone is amplified to an extreme degree, along with
application of suitable signal processing to eliminate an incredible
amount of noise. As in the above methods, the necessary apparatus
must be within a few hundred feet from the telephone set, and the CO
pair must be broken during the operation (with circuitry to detect an
incoming call or outgoing call attempt and reestablish the CO line
continuity to avoid any suspicion on the part of the subject). I am
not claiming that a ringer is a *good* microphone, but under some
selected circumstances this technique can provide useful intelligence.
I may later regret this suggestion, but as an example to
illustrate this principle, here is an experiment that an enterprising
reader can perform using apparatus found in any well-equipped
electronics laboratory. Take a 500-type or 2500-type set with a
bridged ringer and connect its tip and ring directly to the input of a
low-noise amplifier providing say, 80 dB of gain in the voice
frequency range. A suggested approach is to cascade two
Hewlett-Packard 465A amplifiers, with each amplifier being set for 40
dB gain. Take the 80 dB amplifier output and connect it to the input
of a variable bandpass filter having at least 20 db/octave attenuation
(like a Kron-Hite 3100, 3500 or 3700). Take the output from the
bandpass filter and feed it to another amplifier providing 20 to 40 dB
gain and capable of driving a pair of headphones.
Tune the bandpass filter to reject powerline noise, and you have just
turned the telephone set into a crude microphone. At that point it
does not take much imagination to realize that given some competent
engineering resources and a commensurate budget, this technique can be
refined into a practicable eavesdropping device. The availability of
digital signal processing can also do wonders to eliminate the vast
amount of power line, impulse noise and other interference which
develops at the gain necessary for speech pickup sensitivity.
While electromechanical ringers are becoming somewhat a thing
of the past, many electronic telephone sets with tone ringers will
function as an even better microphone. Such tone ringers usually rely
upon a piezoelectric element as the loudspeaker, although a few
low-quality "drugstore-variety" one-piece telephones utilize the
receiver element as the ringer transducer. As most readers of this
forum are no doubt aware, piezoelectric devices will generally
function as both a microphone and loudspeaker. Even a piezoelectric
element optimized for tone ringer use, i.e., with resonance in the
range of 1.5 to 2.5 kHz, will still function as a usable microphone
for lower frequencies.
An on-hook telephone set with electronic tone ringer, if
isolated from the CO line and connected to an ultra-high gain
amplifier with suitable bandpass filtering, and if also subjected to
an appropriate RF bias to cause conduction across the initial
full-wave bridge rectifier and subsequent semiconductor junctions, can
in many instances be turned into a microphone. While this technique
will not work with all electronic telephones, it will work with a
significant number.
The above technique of compromising a telephone with an
electronic tone ringer was first performed almost twenty years ago on
the Ericophone. The Ericophone was an early one-piece telephone, some
models of which contained an electronic tone ringer. While the
geometry of the Ericophone defies verbal description in this forum,
the overall design scheme may best be described as phallic in nature.
Those readers who are familiar with the Ericophone will no doubt concur with this description
."
copacetic
King of the Hippies
Electric already linked to a solution you can try yourself, for free.
Take a look at that before you waste your money.
"You are convinced"... It sounded from the outset that you were absolutely certain there was speech in that audio - Are you saying that you know this for sure, or is it based on the assumption that someone is spying on you?
Take a look at that before you waste your money.
"You are convinced"... It sounded from the outset that you were absolutely certain there was speech in that audio - Are you saying that you know this for sure, or is it based on the assumption that someone is spying on you?