Recording dialogue in the wilderness

[cybrmarc]

How do you mean sounds outside the center? What is the center and how can it be located on the track?

For your first sample, the second speaker disappears after this effect. The first speaker is actually on the left side alone. Thus, instead of left + right, we can assembly the stereo track from left + center and delete the right side.

sample2 excerpt , 4stages.mp3

Trebor, and RHJ, thanks for the tips.
I tried the notching you recommended and the center isolation with your 2d stereo plugin.
I'm wondering how you can tell which channel what noise is on? Both the L and R are identical in audacity as I see them on screen.
Along the same lines: how do I remove the L or the R? With the above tool I've only seen the option to remove or isolate the center.

Also, Trebor, how did you know what frequencies to take out? And how wide? Do you look at a waveform, do you just know that 150hz and 5800hz are common noises for birds/main?

Learning to fish,
Marcus

[Trebor]

Trebor, how did you know what frequencies to take out? And how wide? Do you look at a waveform, do you just know that 150hz and 5800hz are common noises for birds/main?

I looked at the spectrogram display, rather than the waveform display, which showed a strong constant signal around 5800Hz ...

the bird noise is present throughout the recording and high-pitched

Audacity Spectrogram display , showing constant loud noise around 5800Hz.gif (229.45 KiB) Viewed 2165 times

The strong 150Hz signal, the 3rd harmonic from 50Hz mains hum, is more obvious on the frequency analysis than on the spectrogram ...

Large spike at 150Hz on Audacity 'frequency analysis' ('plot spectrum').gif (37.41 KiB) Viewed 2165 times

[Robert J. H.]

Trebor, and RHJ, thanks for the tips.
I tried the notching you recommended and the center isolation with your 2d stereo plugin.
I'm wondering how you can tell which channel what noise is on? Both the L and R are identical in audacity as I see them on screen.
Along the same lines: how do I remove the L or the R? With the above tool I've only seen the option to remove or isolate the center.
Marcus

(on Windows, there's a play only left or play only right option built into the plug-in)
We should first create the three channels.
Import the track and duplicate it.
On one track, remove the the center, on the other, isolate it.
Give them proper names such as "Sides" and Center".
Split both tracks to mono (Track drop down menu)
You can delete the last track because the center is always mono and the third and forth track is the same.
Listen to the tracks in solo mode. You'll perceive that the (second) right track has nearly no speech from the first speaker but some water gurgling.
You can in principle use the envelope tool to bring down the volume, where no speech is audible in the individual tracks.
Now, let's combine the left and center channel.
move the last track up, into second position (Track drop down menu, Move Up).
On the first track, open this menu again and choose "Make Stereo Track".
The second track can now be muted.
Listen to the result.

You can skip the "remove center" on the first track and just use the original left channel, this might even be better , you have to try.
I don't think that I have saved this project, hence I can't tell.

[cybrmarc]

Thanks for the walk-through you too. So now I can find the mains hum, or any big spike, and separate right, left, and center. I noticed when I removed the center on our digital recorder files it eliminated everything. So I'm guessing that doesn't differentiate as much as the camera (what was in the previous mp3)? Or am I missing something.

My latest challenge is wind noise.
Here's the clip:
http://www.teachingdrum.org/adjul/WindyDay.mp3
I ran a spectrum analysis on both the wind noise section, and a section without it:



What I don't get is why the file *without* the noise seems to be more erratic (with the wavy spikes). I'm guessing because the frequencies in voice is more nuanced than the wind. And when the wind dies down the voices become prominent. I tried running this through PluralEyes to sync it (we have the external audio which is virtually wind-less) but PluralEyes can't work with most of the files. I'm guessing because of the wind noise. So if I could mute the wind just enough for PluralEyes to kick in and pick up the voice, that would be great.

I can always manually sync the clips 1 by 1, but I figured this is an opportunity to learn what I can do with audio post.
And yes, we have a new camera on order with an omni-shotgun mic and professional windscreen.

Thanks,
Marcus

[kozikowski]

What you're listening to is the resonances of the wind screen, the microphone body and the capsule itself under extreme mechanical duress from wind. The rumble and slap is usually so much louder than the show it swamps it. It happens to your ears, too, but they have the advantage of interpretive wetware and multi-input stimulation to sort it all out. It's the same process that allows you to "hear" somebody with almost no actual sound by reading lips.

What are the holes in the spectrum again?

Koz

[Trebor]

... I noticed when I removed the center on our digital recorder files it eliminated everything. So I'm guessing that doesn't differentiate as much as the camera (what was in the previous mp3)? Or am I missing something.

If your digital recorder only has one microphone , i.e. mono not stereo, then the two tracks are perfectly identical, the difference between them is zero, ( two tracks looks like stereo but can be dual-mono where the tracks are absolutely identical).

What I don't get is why the file *without* the noise seems to be more erratic (with the wavy spikes).

The vertical scales on those two graphs are not the same, so the same spike will be different sizes on the two graphs.

Robert's plugin rjh-stereo-tool.ny on "isolate centre" works very well to reduce the wind noise on that ...

WindyDay, before-after Robert's plugin on ''isolate centre'' #.wav

(839.47 KiB) Downloaded 67 times

settings used on Robert's plugin in the ''after''.gif (10.2 KiB) Viewed 2143 times

[ You're fortunate to have made the recordings in stereo as it wouldn't be possible to reduce wind noise on a mono recording with equalization as wind occupies most of the sound spectrum ]