Topic: help with wind noise

[pwig]

How do you remove wind noise from a recording?

[morst]

Not to be a smart-alec, but the best way is to preempt it with a windscreen. Some people even use a gymsock if they don't have anything else available. Once it's on there, you're kinda screwed, though a carefully-applied bass roll-off could perhaps help a little. If it's on one channel at a time, you can try "monoing" the recording out in spots to attempt to save it. Why, what did ya get with rumble??

[stevetoney]

It's tough and requires alot of diligence, but I've had some limited success. 

Most recently for one of my recordings, I found that the rumble was due to some leakage around one of my screens (these particular screens didn't fit all that tight and the wind direction was staight from behind so the wind come up through the back hole in one of the screens).  Anyway, due to the wind direction and position of the other mic in the ORTF scheme, the sound was OK on the other channel. 

In that case for the short passages where the rumble happened, I duplicated the good channel and whacked the noise in the other channel by using the 'copy l to r' function or whatever it's called in Audition.  The problem then is that you change the dynamics of the short passages and there's a noticeable change in the dynamics of the sound, so you have to mess around with the dynamic controls a bit to get them to match as well as possible.  I mainly messed around with the delay so that one channel was very slightly off phase with the other.  It's not easy to get it sounding nice and transparent (with the rest of the recording) but in the end once you've got the mix right, it's the same set of changes for the whole recording so you only have to get it right the first time and then save or remember all of the settings...and it definitely is a sound improvement over leaving the wind rumble.  The bottom line though is that I've found that you're never gonna get a nice sounding end result that you're satisfied with.

Course, if you don't have one good channel to start with, then the above fix doesn't apply, but this is fresh on my mind since I just used this technique recently.

BTW, I suppose it would be theoretically possible to use the above technique if there was another recording of the show from another rig, but that sounds like too much effort to be worth it to me...and I kinda doubt the end result would be worth listening to (too much change in sound dynamics in your recording for it to sound homogeneous).