Fax software tools

Does anybody know of a software tool or package that can take in an
audio  recording of a fax call, and extract the image that is being
faxed?

Thanks

Does anybody know of a software tool or package that can take in an
audio  recording of a fax call, and extract the image that is being
faxed?

Thanks

Mike wrote:

> 
> Thank you very much.
> 
> Monteria looks like they have what I am looking for.  I will have to
> contact them and see if the price is reasonable.
> (http://www.monteriallc.com/index.htm)
> 
> If anybody else knows of similar products, please let me know.

I never thought anything that simple would be so complicated.  I should
know better :-)

Patience, persistence, truth,
Dr. mike

-- 
Mike Rosing
www.beastrider.com                   BeastRider, LLC
SHARC debug tools

Mike Rosing wrote:
> 
> Mike wrote:
> > Does anybody know of a software tool or package that can take in an
> > audio  recording of a fax call, and extract the image that is being
> > faxed?
> 
> Why not just feed it to a fax machine?  Or a sound card with fax-modem
> software?  You already have the signal, just feed it to things that can
> deal with it.

That won't work.  Fax transmissions start with a handshake, including a
training sequence.  IIRC, the training sequence is there for the
answering-end to equalize the line.  If the training sequence fails, the
calling-end can choose another baud rate/modulation scheme.  In a
recording, the answering-end has no opportunity to influence the
calling-end.  

Also, IIRC, both the tx and the rx communicate in the same band, so
you'd need a four wire (2-channel) tap to separate the calling and
answering ends of the conversation.

If you have recorded both halves of the conversation seperately, you
MIGHT be able to play some games.  Basically, you have to tx some stuff
and listen for the answerer to finish responding.  When it finishes, you
can tx the next bit of the conversation.  If you blindly play the tx-end
back, you will find that the timing is off, and the whole thing falls
apart.

Third party processing is a LOT more difficult than when your equipment
is a party to the call.


-- 
Jim Thomas            Principal Applications Engineer  Bittware, Inc
jthomas@bittware.com  http://www.bittware.com          (703) 779-7770
Unix IS user friendly.  It's just more particular about its friends.

Jim Thomas wrote:
> 
> > > Does anybody know of a software tool or package that can take in an
> > > audio  recording of a fax call, and extract the image that is being
> > > faxed?
> >
> > Why not just feed it to a fax machine?  Or a sound card with fax-modem
> > software?  You already have the signal, just feed it to things that can
> > deal with it.
> 
> That won't work. 
> Fax transmissions start with a handshake, including a
> training sequence. 

Agreed.

> 
> Also, IIRC, both the tx and the rx communicate in the same band, so
> you'd need a four wire (2-channel) tap to separate the calling and
> answering ends of the conversation.

The fax protocols are communicating half-duplex. You don't have to
separate the transmit and the receive sides. 

> Third party processing is a LOT more difficult than when your equipment
> is a party to the call.

I can develop such software, however it is going to be relatively big
project.

Vladimir Vassilevsky

DSP and Mixed Signal Design Consultant

http://www.abvolt.com

"Jim Thomas" <jthomas@bittware.com> wrote in message
news:3F5F6564.F5F4FC90@bittware.com...
> Mike Rosing wrote:
> >
> > Mike wrote:
> > > Does anybody know of a software tool or package that can take in an
> > > audio  recording of a fax call, and extract the image that is being
> > > faxed?
> >
> > Why not just feed it to a fax machine?  Or a sound card with fax-modem
> > software?  You already have the signal, just feed it to things that can
> > deal with it.
>
> That won't work.  Fax transmissions start with a handshake, including a
> training sequence.  IIRC, the training sequence is there for the
> answering-end to equalize the line.  If the training sequence fails, the
> calling-end can choose another baud rate/modulation scheme.  In a
> recording, the answering-end has no opportunity to influence the
> calling-end.
>
> Also, IIRC, both the tx and the rx communicate in the same band, so
> you'd need a four wire (2-channel) tap to separate the calling and
> answering ends of the conversation.
>
> If you have recorded both halves of the conversation seperately, you
> MIGHT be able to play some games.  Basically, you have to tx some stuff
> and listen for the answerer to finish responding.  When it finishes, you
> can tx the next bit of the conversation.  If you blindly play the tx-end
> back, you will find that the timing is off, and the whole thing falls
> apart.

I haven't looked at fax modem standards for a while, but I don't think they
use the same band.  The transmission rates are asymmetrical, so they either
use a narrow band in the reverse direction, or half duplex, but I forget
which.

-- glen

Glen Herrmannsfeldt wrote:
> I haven't looked at fax modem standards for a while, but I don't think they
> use the same band.  The transmission rates are asymmetrical, so they either
> use a narrow band in the reverse direction, or half duplex, but I forget
> which.

Yeah, I think you're right.  I last looked at this in 1988, so it has
been a while.  Now that I think about it, I'm pretty sure they use
different FSK frequencies, and I THINK it's full-duplex.  I do know we
had to separate rx and tx in our recordings because the demod would get
confused if it was presented with both sides of the "conversation."  I
guess that could be done with a filter easily enough though.

We also had to demod enough of the answering signal so we'd know where
we were in the handshake, and then we could pause and resume the
playback of the calling signal at the right times.  Luckily, the
handshake is done with the least-common denominator modulation scheme.

But I distinctly recall that playing the signal back blindly did NOT
work, even if it was to the same fax machine that took part in the
original transmission.

-- 
Jim Thomas            Principal Applications Engineer  Bittware, Inc
jthomas@bittware.com  http://www.bittware.com          (703) 779-7770
Unix IS user friendly.  It's just more particular about its friends.

The machines that I have seen that do this from a recording required a
digital recorder. The wow and flutter of the analog recorders was too much.

Dirk

Dirk A. Bell
DSP Consultant


"Mike" <mike_dsp@hotmail.com> wrote in message
news:1911c5cf.0309100717.9d02fa8@posting.google.com...
> Does anybody know of a software tool or package that can take in an
> audio  recording of a fax call, and extract the image that is being
> faxed?
>
> Thanks

"Glen Herrmannsfeldt" <gah@ugcs.caltech.edu> wrote in message
news:rtK7b.411812$uu5.75263@sccrnsc04...
>
> "Jim Thomas" <jthomas@bittware.com> wrote in message
> news:3F5F6564.F5F4FC90@bittware.com...
> > Mike Rosing wrote:
> > >
> > > Mike wrote:
> > > > Does anybody know of a software tool or package that can take in an
> > > > audio  recording of a fax call, and extract the image that is being
> > > > faxed?
> > >
> > > Why not just feed it to a fax machine?  Or a sound card with fax-modem
> > > software?  You already have the signal, just feed it to things that
can
> > > deal with it.
> >
> > That won't work.  Fax transmissions start with a handshake, including a
> > training sequence.  IIRC, the training sequence is there for the
> > answering-end to equalize the line.  If the training sequence fails, the
> > calling-end can choose another baud rate/modulation scheme.  In a
> > recording, the answering-end has no opportunity to influence the
> > calling-end.
> >
> > Also, IIRC, both the tx and the rx communicate in the same band, so
> > you'd need a four wire (2-channel) tap to separate the calling and
> > answering ends of the conversation.
> >
> > If you have recorded both halves of the conversation seperately, you
> > MIGHT be able to play some games.  Basically, you have to tx some stuff
> > and listen for the answerer to finish responding.  When it finishes, you
> > can tx the next bit of the conversation.  If you blindly play the tx-end
> > back, you will find that the timing is off, and the whole thing falls
> > apart.
>
> I haven't looked at fax modem standards for a while, but I don't think
they
> use the same band.  The transmission rates are asymmetrical, so they
either
> use a narrow band in the reverse direction, or half duplex, but I forget
> which.
>
> -- glen
>
>

The ITU's T.30 Recommendation specifies the interaction between two Group 3
fax machines (which is the great majority of fax machines deployed in the
real world). For these "legacy" fax machines (V.17 14400bps modulation rates
and below), the protocol is entirely half-duplex, so they do use the same
band - the voice band (300Hz - 3400KHz). For the handshaking part of the
protocol, V.21 Channel 2 is specified (simpy FSK modulation at 1750Hz
nominal freq.), while the actual transmission of the page data is done using
whatever modulation was negotiated during the initial handshaking,
negotiation and training phases of the call.

To answer the original poster's question: yes, there exist companies out
there that market fax analysis software.  The two that I am familar with are
Genoa (now owned by Quality Logic - see
http://qualitylogic.com/genoa_test_tools/fax/products.html for their fax
offerings), and Telegra (not even sure if they're in business anymore).  I
believe that you can use two ChannelProbes by Genoa to build a call around
the page data you have captured, but I would contact their technical support
for more information.  By the way, this stuff isn't cheap.

You could also conceivably set up a fax modem to demodulate the data you've
captured, but you would have to issue it discrete commands to put it into
the correct mode, which requires that you have a good working knowledge of
T.30 as well as the command set for the modem you're using.

Good luck,

James

Vladimir Vassilevsky <vlv@abvolt.com> wrote in message news:<3F5F734B.96F9C9AF@abvolt.com>...
> Jim Thomas wrote:
> > 
> > > > Does anybody know of a software tool or package that can take in an
> > > > audio  recording of a fax call, and extract the image that is being
> > > > faxed?
> > >
> > > Why not just feed it to a fax machine?  Or a sound card with fax-modem
> > > software?  You already have the signal, just feed it to things that can
> > > deal with it.
> > 
> > That won't work. 
> > Fax transmissions start with a handshake, including a
> > training sequence. 
> 
> Agreed.
> 
> > 
> > Also, IIRC, both the tx and the rx communicate in the same band, so
> > you'd need a four wire (2-channel) tap to separate the calling and
> > answering ends of the conversation.
> 
> The fax protocols are communicating half-duplex. You don't have to
> separate the transmit and the receive sides. 
> 
> > Third party processing is a LOT more difficult than when your equipment
> > is a party to the call.
> 
> I can develop such software, however it is going to be relatively big
> project.
> 
> Vladimir Vassilevsky
> 
> DSP and Mixed Signal Design Consultant
> 
> http://www.abvolt.com

Thanks for the offer but I was hoping to find an existing tool that
runs under Windows and has a reasonable cost.

I actually have the expertise to develop this myself but as you said
it is a big project, especially if you want to support error
correction mode where portions of the image can be retransmitted.

Getting the bits is fairly straight forward...it's knowing what to do
with them that is the challenge.

Mike