roclark wrote:Thanks for looking into all that Jeff.
That's a bummer because a few years ago I used to use Cool Edit, an open source audio editing software, and you could very easily record both ends of the phone conversation. Adobe bought up Cool Edit, and now it costs a few hundred dollars.
For now, the Gizmo recording/Audacity edting combo will work. Maybe in the future, Audacity will be able to record both sides of the conversation at once without any manual fidgeting on the user's part.
I was hoping that one of the gurus here would have a nice simple solution, but perhaps not. There are a couple of work arounds that I can think of, but they wouldn't be the ideal and would require a bit of time & effort with each recording. Your easiest solution, if Audacity can't be made to work in its current iteration, might be to use other software.
Since Cool Edit was at one time open source, I imagine that the code that had been released under that license would forever be open source, regadless of whether or not Adobe bought up the rights. If you have an old copy around on another computer, and that copy had been able to do what you want, you could just copy it over to your new computer.
If you don't have a copy anymore, or don't know where to find it, you might be able to download a copy from someone else. Here is a site that has a couple of verisons. I don't think they charge anything for them, but I'm not sure:
Cool Audio Editor 2.3
http://www.safe-install.com/programs/co ... ditor.html
Cool Audio Editor 3.22
http://cool-audio.safe-install.com/
I suppose you could also experiment with some of the other free programs out there:
Kristal Audio Engine
http://www.kreatives.org/kristal/
Source Forge has others that might work:
http://sourceforge.net/softwaremap/trov ... rm_cat=115
---------------------------------
If they all have the same limitations and/or you just want to use Audacity, I think the best bet would be to use 2 different computers; one to record the interviewee's side of the conversation, the other equiped with just a microphone to record the interviewer's side of the conversation. You could then import the sound file from the 2nd computer into the audacity file on the other computer. You would then probably need to time shift it a little one way or another to synchronize them properly.
Another possibility is to take the beeps out of the Gizmo recording. Doing so manually would be quite a hassle, however, and if the beep occurred while someone was talking, you couldn't just delete that portion of the recording without also deleting that portion of the call as well. IF you could get an EXACT copy of the beeps, however, you might be able to "erase" them with a few minutes or so of your time. That's a pretty big IF as it has to be almost a perfect duplicate, but if that is possible, you can import, then invert that track, line them up perfectly, and presto, it's like the beeps were never there in the first place.
Here's a thread on that subject:
http://audacityteam.org/forum/viewtopic.php?f=12&t=1476
If Gizmo uses some sort of system in which the beeps are exactly and perfectly timed and use the exact same beep each time, I think that may well be possible. If you could make a call and mute the microphone on each end of the line and record it, you should be able to get a copy of the beeps alone, ideally without any other background noises etc. If the beeps are always an identical time period apart (ie. every 12.532 seconds) you should be in business. You would need to record for as long as the longest recording for the podcast you want to use.
It sounds like you would need to leave them all in uncompressed format (ie .wav file or audacity file) as a lossey format like .mp3 or ogg vorbis etc. would make them no longer identical.
You could then invert the file of beeps. From the menus at the top: Effect > Invert

- invert.jpg (88.04 KiB) Viewed 1529 times
Save that file.
Import it into any recording where you want to remove the beeps. You will then need to "time shift" the inverted beep file to line up PERFECTLY with beeps in the recorded interview. The time shift tool is circled in red in the above Audactiy screenshot. After selecting that tool drag the track you want to move back & forth until you get it lined up as you like.
When I experimented with that (see earlier thread) I couldn't get it to work until I had zoomed way in and got them exactly lined up. Real close didn't work. Don't forget too, that since the "beep track" is inverted it will be upside down with respect to the beeps in the file you are trying to clean up. It will be as if you are looking at a mirror image, with the mirror being upon that center line.
If the beeps are generated at random times or each beep is unique it won't work correctly. But if they are identical and are identically spaced one from another it should work like a charm. If you can line up a beep on the inverted line exactly with a beep on the recording, all of the beeps on the recording should disappear.