Telephone recording Advice Needed.

[philip888]

Hi I am editing audio files that were recorded from a conference call.

I want to fatten up the the phone call sound to help it not sound like a phone call.

Any suggestions?

Thanks

Philip

[daeron80]

Sorry, about the only thing you can do is re-record it through a system that has sufficient bandwidth. Telephone systems roll off lots of lows and some highs. Gone forever, I'm afraid. No way to get it back.

What we do when we want a good sounding phone interview is have the interviewee go to a studio and call us from there. The house engineer tracks him to a decent system and ftp's us a full res WAV or AIFF. We drop it into PT along with the person we recorded on this end, and, bingo, good sound.

[philip888]

Thanks. But this project I can't have them re-record it.

So that is it?

[Dism]

Originally Posted by philip888

Thanks. But this project I can't have them re-record it.

So that is it?

Indeed.

I've never heard of a phone recording that was made to sound not like a phone recording. Extremely low-fi, through a tiny microphone, with no detail at all. It can be cleaned up a bit with a LOT of work, but it's still going sound tiny and low quality.

[ondruspat]

Like D & D said you got what you got, there's no real way to "fatten" it up. It kinda reminds me of a time I got handed a video with horribly recorded audio. I could clean up the noise but it took away all the presence and strength of the voice. I ended up lowering the pitch of the audio and tweaking the EQ until it was intelligble. It still sucked but it worked for what the video was used for (it was played at a political convention, there was a lot of crowd noise and drinking, no one really paid attention to the audio, they just saw this fairly prominent mayor's face on the screen ).

If you can in the future though use a quality tap to interface the telephone audio if you can't do the studio suggestion. As with anything audio better equipment will yield better sound if used properly. I have a pretty decent phone tap but I haven't used it in so long I can't remember what it is!! If I dig it out I'll post it.

[philip888]

Thanks I appreciate it.

[nerhal]

use plugins. it will still be what it is but you can beef it up and clear it up. maybe EQ it as well.

[jazzdrumr]

Try an aural exciter or big bottom to add some low/high harmonics but your really stuck with 400-4k of bandwidth. Sorry...been there

[philip888]

Originally Posted by jazzdrumr

Try an aural exciter or big bottom to add some low/high harmonics but your really stuck with 400-4k of bandwidth. Sorry...been there

Thanks

[daeron80]

Well. I'm happy to say, I was completely wrong about this. Check out this wiki article. Scroll down to "Bandwidth Extension of Speech in telephone systems." I tried out what they mention there, and doggone it, it works.

I recently got handed a project of combing through phone messages consisting of questions for a theologian to answer on the radio (yes, really!). In addition to the problems of phone bandwidth, the system can only record to 8-bit µLaw @ 8 kHz. So every call is brickwall filtered at 4 kHz, even though some cell phone transmission is done at 16 kHz these days. So here's what I did:

• Use iTunes to convert the phone messages to uncompressed WAV files so PT can see them
• Audio track with phone message has 75 Hz hi pass and variable EQ suited to each call - there's tremendous variance in spectral balance among phones
• Bus to two Auxes
• Aux 1 is sub-synth: Recti-Fi followed by 63 Hz hi pass
• Aux 2 is a cheapo exciter using:
  • 3 kHz hi pass
  • Recti-Fi
  • Dyn3 compressor clamping down at 4:1 because some generated highs are way too loud.
  • 5 kHz hi pass
  • 8.5 kHz low pass bars the super-high trash that gets generated
• Those three tracks sum to an Aux for gain control
• The gain Aux feeds an Audio track to record the result

The sub-synth Recti-Fi plug uses Positive mode, and is pre-filtered at 500 and post filtered at 250 Hz. 100% mix, of course.
The exciter Recti-Fi uses Alternate mode. Pre- and post-filtering need to be adjusted based on the spectral balance of each call, the goal being to add a sense of brightness without getting harsh. 100% mix.

Here's a screenshot.
All the plugs on the master fader are just analysis tools except for the trim, of course. They have no effect on the sound.

It still sounds pretty nasty. I mean, GIGO, you know? But it's a marked improvement. I bet a really good exciter would work better for generating highs but I don't have one at work yet. I'll try the AIR one once we upgrade to 9, but my experience with it at home has not indicated that it yields stellar results. It probably does pretty much the same thing as my little fake-it arrangement.