Audacity Forum

I'm new at this. I recorded a 3-person Skype interview using a Yeti mic and the sound from my headphones was too loud so the mic picked up the other 2 voices and now I have a recording where 2 of the 3 voices have a strong echo. It's not reasonable to ask the interview guest to redo the interview, so I have to try to fix/improve it.

I tried the SPL Deverb plugin and I could not achieve any significant difference. I'm now in a pretty bad position. Under some pressure and a podcast partner who is not too happy since there was another problem earlier with the same interview (which I resolved, the problem audio is from part 2 of the interview).

I can supply a sample of the audio if anyone is willing to help me figure this out.

I can supply a sample of the audio

Yes, but you shouldn't depend on us performing miracles. Putting echoes and reverb in a show by accident is usually the kiss of death.

Forum postings are usually limited to 20 seconds mono (one blue wave) or 10 seconds stereo (two blue waves). WAV (not MP3) is recommended, particularly if we have to try surgery.

If you think more work is useful, post a larger chunk on an internet service and point us to it.

http://forum.audacityteam.org/viewtopic ... 49&t=72887

Koz

Thank you very much for your help.

I realize there are no miracles for this problem. Was mainly wondering if I could improve it.

I've attached a 10 sec WAV sample file.

I also put the full (part 2) of the interview in a public folder on Google drive here: https://drive.google.com/open?id=0Bwndy ... ndYczFieDA

While we wrestle with this, do you know what you did wrong so it never happens again?

Recording a Skype interview is not for the easily frightened. I did it with two computers and a small mixer.



This is an engineering test. Denise is four time zones away.

http://kozco.com/tech/audacity/clips/De ... ilyCut.mp3

A number of podcasts do it that way.



The broadcast people have production sound mixers with communications built-in along with telephone connections, but that's not particularly cheap.

We have been finding that all-in-one software solutions tend to change at the whim of Skype. Something that worked last week may not work today. The Pamela people recently got burned with that.

Am I being cheerful enough? As we go.

Koz

I don't think that's what you think it is. I think you were recording Skype echo processing instead of the show. That's fairly common. Usually you get mars voices with crunchy tails on everything. You got lucky in that you can still understand the performers.

As we go.

Koz

It's a mono show. You may have two blue waves, but they're the same and the bouncing sound meters bounce exactly the same. You could have split the show in two (left-hand drop-down menu > Split Stereo to Mono), deleted one of them and sent 20 seconds of mono.

As we go.

Koz

I may have gotten the slightest improvement through brute-force cancellation.

I'm working with the forum posted sample.

Drop-down menu on the left > Split Stereo to Mono.
That gives you two independent mono sound tracks. You can play both or one or the other with the SOLO and MUTE buttons. They naturally try to play at the same time.

Select the bottom track by clicking just above MUTE. Effect > Invert.

That could give you total silence. The top and bottom tracks are now completely opposed to each other and cancel. Now select The Time Shift Tool (two sideways black arrows.)

Click the bottom track blue waves and move to the right very slightly. Play it. Chances are the echo got much worse. Now carefully tune the time shift left and right and you may find a place where the echo isn't quite as bad.

You may need the Zoom tools. I only use three zoom tools out of the pile available.

Zoom (Windows)
-- Drag-select something on the timeline and zoom into it. Control-E
-- Zoom out a little bit. Control-3
-- Zoom out full. Control-F

-- Shift-ScrollWheel or Shift-TouchPadScrub will shift the timeline view left and right (sooner and later).

Change the timeline so it doesn't shift by itself:
-- Edit > Preferences > Tracks > [_] Update display while playing. (de-select)
Note: Update Display is good during recording.



Best of luck. If you change your recording technique for the better, post back how you're doing it.

Koz

If you get tantalizingly close, but no cigar, Select the bottom track and Effect > Amplify. Change the top number, Amplify to -1 > OK.

You can Edit > UNDO and change that number higher (-2) or higher until it stops improving.

Koz

I'm going over this in my head. I have done recordings like this in conference mode. You put the Yeti on the table and talk to each person on speakers. The Yeti is connected to a separate recording computer. It picks up you direct and clean, and it picks up everybody else in Skype mix from the speakers. That totally does work. I've done international production meetings like that. What you can't do is record on the Skype computer. Skype does Not Play Well With Others. It takes over the sound channels and there isn't much you have to say about it.

You can also use a Personal Sound Recorder on the table instead of the Yeti and computer.

Of course it's a given none of these computes can make noise. No fans.

Koz

I think you were recording Skype echo processing instead of the show.

That's interesting. I recorded the first half of the interview with Pamela and it crashed on me. So we did part 2 a few days later and I used DVDSoft Skype call recorder. I think it works okay if I have my mic set properly but I will do some more testing. If I am recording the "Skype echo processing" I guess you are saying that the recording software did that, because Skype changed something.

I have been considering getting a small mixing board (which small mixing board did you use?). But I wanted to make sure we were serious about this podcast before I took the next step equipment-wise. In the past my partner has used Google hangouts to record the show and I stripped mp3 from the mp4 and edited that. But Google hangouts has problems with lags and audio glitches, and in this case our interview guest couldn't use it, so we switched to Skype on the fly.

I'm looking at your suggestions/instructions and appreciate it. I saw a youtube video with a similar solution (I think) but I didn't understand the mono/stereo split and such. Your instructions are clear.