AGC for telephone recording

Go To Last Post
11 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

A couple of times a year I have a telephone consultation that I record and transcribe for both of us (with the permission of the other person). Over the years I have used a couple of different pickups/recorders but while both recorded me just fine they were weak recording the other person. When I try to transcribe the conversation it's difficult to hear the other person.

So I was wondering whether there are any inexpensive alternatives. I looked at some better equipment, but it looks like several hundred dollars to do it right (tape machines with AGC). Since all I need is the pickup (or direct connection) and a smart amplifier (not the recorder) I was thinking under $50 would be nice.

So if anyone knows a source for (inexpensive) existing equipment or a circuit for something that would actually work I would really appreciate it. Remember you're talking to a software guy here, so talk real slow.

Thanks!

Chuck Baird

"I wish I were dumber so I could be more certain about my opinions. It looks fun." -- Scott Adams

http://www.cbaird.org

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hi Chuck,

Use Skype and its recording add-on. Works for me.

Cheers,

Ross

Ross McKenzie ValuSoft Melbourne Australia

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Throw more bits at the ADC, then do the AGC in software.

Audacity leveler should do the trick:

http://forum.audacityteam.org/viewtopic.php?f=28&t=8826

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Use the soundcard in the PC hooked up the the line and record 2 channels.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hi Chuck,
I am not sure at which point you are recording, but I gather at the 2 wire part of the line. At this point the TX level is much higher than the RX level. If you monitor at the 4 line at the hybrid you can pad the TX level back and then recombine them back in a summing amplifier.

Charles Darwin, Lord Kelvin & Murphy are always lurking about!
Lee -.-
Riddle me this...How did the serpent move around before the fall?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I agree with the above suggestions, and won't expound on them...

I will add several additional thoughts, however.

If you chose to grab the signal right from the Tip & Ring lines make sure you go through a DAA device. This protects the phone line from your equipment, and protects your equipment from the phone line's FLOATING -48V amd potentially several hundred volt Ring Signal.

Remember that in the old days if you got caught hacking the phone Ma Bell offered you a job. These days not the case.

(Recording, smart answering machines, etc., do not count as hacking...).

Second thought to add to the list above:

If you use a cordless phone then the AGC has already been done for you for its own wireless link from the handset to the base. Grabbing that signal, from either end, would save you the trouble of doing it all yourself. Older units, (big, clunky, through hole designs), obviously have more circuitry exposed for hacking. You could talk on one phone, and use another one for your recording.

Option #3:
(I want my cut)
Hire Uncle Bob to write an Android App to automatically record the call, then feed it to Google's voice recognition engine, then eMail you the transcript. Uncle Bob gets his next "Killer App", you get your problem solved, and we can all retire.

JC

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Wow

Imagecraft compiler user

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

One assembly solution :
The main parts would be
SLIC , CODEC , Memory and TinyAVR .
SLICs have a complete control over Tip and Ring .
Setting their inputs , you could simply check different states of line .
RingTrip , Ground Key and Loop Detect .
Normally , they've got one Det (detect pin) [1] getting activated whenever the states are met .
By the way , SLICs would extract voice and send it to CODECs or get Voice from CODECs and send it over line .
PCM modulation and demodulation are accomplished by them .
Micro could communicate with CODEC .
Regarding One Frame Synchronizer (Normally 8 Khz) , you could save PCM data coming out of CODEC on memory and Demodulate it whenever interested .
Whenever the PCM data would be ready (Ready to come out of CODEC) , One pin would indicate it . you could connect this pin to external interrupt pins of micro .

[1] cantankerous clever detective

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I appreciate the comments so far. Another thought I had - the real time conversation is fine, it's just transcription that's a problem. Does anyone know any software I can feed the recording to which can do equalization of the two sides of the conversation? That would solve the problem as well.

Chuck Baird

"I wish I were dumber so I could be more certain about my opinions. It looks fun." -- Scott Adams

http://www.cbaird.org

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

audacity effect->leveler might do what you want.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

How about this one?

Should cost around $3 and can be built haywire or Manhattan style in half an hour or so.