How to remove this kind of clunk noise?

[Peter Berger]

I use Izotope RX3 which is great for removing noise, but I cant figure out how to remove this kind of problem (i dont how to call it in english). Hear it in attachement. I tried spectral repair, but without succes. Thanks for any help!

[Adam Bauser]

I'm not aware of any software that can repair an audio hit like the one in your clip. Is that a microphone bump?

Only options I can think of:

1) When I shoot with a boom or lavaliere, I also run the camera's on-board microphone. Sometime I can use the audio from the camera's microphone to replace a mic hit. If it's just a few frames of audio, you won't notice the difference in tone so much. Even if you do, it's better than the hit sound.

2) Replace the dialog. Have the speaker/talent say that part again and edit it in. Sometimes you get lucky and can steal the same word from another part of the speech or interview.

3) do a quick dip down on the volume at that spot or even cut the frames out. Sometime what's a really obvious edit to you isn't as obvious to the audience, especially when there's picture.

[Peter Berger]

Thanks. Yes, microphone bump.

I just managed to partially repair it with spectral repair.

[Jim Andrada]

Hi Peter.

I'm traveling this week. I'll get home on Friday and see what i can do with the clip. Izotope is great and I've been able to rescue stuff that I wouldn't have thought could possibly be repaired with it.

I'll try to remember to have a shot at it when I get back to Arizona

[Greg Miller]

The simplest thing I could do was to remove the samples with the worst noise. In this case that was 2420 samples (roughly 50 msec.), beginning with sample number 37081 (see step1). I just cut on a negative-going zero crossing so there will be no abrupt transition at the cut point.

That will shorten the entire clip a bit (in this case from 80641 samples to 78220 samples (see step2). And then, if you need to maintain sync, you can either add a tiny bit of silence somewhere nearby, or else do a slight stretch (see step3) to get the length of the file back to the original number of samples.

Of course that's quick and dirty. It has taken much longer to type this and grab the screen shots than it took to fix the audio.

And sometimes it helps to "double up" the good samples just ahead of, and just after the cut, rather than fixing the sync issue by stretching. In the present case, I think it goes by pretty fast and doesn't sound too bad. attached file v01.wav below

EDIT:
There is also a bit of high frequency raspyness after the place where the "clunk" was removed. In fact I think perhaps a preamp somewhere is clipping or something is in saturation as a result of that overload, and that's causing the raspyness after the "clunk."

I have fixed the raspyness with a gentle HF-rolloff over the syllable following the edit. The final version is attached file v02.wav below

[Battle Vaughan]

Another approach -- which is perhaps similar to the one you used, Peter, --- is to use an envelope tool (this one is is Audacity, excellent freeware) to lower the amplitude of the offending sound to a very small (but not-silent) level. One advantage is that the synchronization of the audio is not affected, and it sounds, if anything, like the speaker took a breath there....

[Christopher Young]

Waiting on a 90 min render so had a go at it. It's a 'fudge' but better than the hit. How many have you got?

Chris Young
CYV Productions
Sydney

[Graham Bernard]

It does sound like an Tx/Rx, transmitter/receive send, battery going down?

My take on it.

http://www.dvinfo.net/forum/attachme...1&d=1394444702

G

[Graham Bernard]

And after a bit more . . . .

http://www.dvinfo.net/forum/attachme...1&d=1394457823

[Peter Berger]

I had similar results with the spectral repair, but it stll is not perfect :)

[Graham Bernard]

Let's hear it Pete?

G

[Peter Berger]

I will upload it.

BTW, I figured out what the clunky sound is. It appears allways in the moment when speaker uses a projection remote control for a PowerPoint slideshow (she also used some kind of a laser pointer). Is this possible that the remote control could cause this problem with my wireless lav (Sennheiser G2)? Would changing the frequency on my tranciever/reciever helped?

[Graham Bernard]

Audio loops. Passing Taxis. Not heard of IR nor Laser pointers affecting UHF? I hazard a guess that they aren't in the same "Wave Band". But Sennhieser do have these babies:

Sennheiser Canada Inc. - SZI 1029-120

G

[Greg Miller]

Laser obviously wouldn't affect an RF device.

However, there are a huge number of RF-based remote controls for projection, and for whole-room control (lights, curtains, screens, audio, etc. etc.) for that matter. (Manufacturers AMX and Crestron come to mind as just two manufacturers.) It is quite possible that an RF-based media control system is bothering a wireless mic system. Check RF frequencies, obviously, but even that might not be enough. If the remote-control transmitter is powerful enough, it might cause the mic's receiver to de-sense, and you might need to go to a drastically different frequency band. In fact, since a lot of wireless lav microphones are unbalanced (between the capsule and the transmitter), it's even possible that the remote-control RF is getting into the wireless transmitter via the unbalanced lav wiring.

[Graham Bernard]

Greg, I agree.

G