Topic: EQing out wind noise?

[acidjack]

I had a recent recording that came out beautifully...except that, despite my use of windscreens, there are some spots where the A/C system in the venue is cranked extra-high and creates significant wind noise.  The worst of it is between songs, but it is audible during quiet passages.  It was a weird situation as I have taped from this spot many times and never had this happen before.  But anyway... has anyone tried to cut wind noise by reducing specific frequencies at which it occurs using EQ?  I did a light (-2dB) across-the-board low-frequency reduction which seems to have helped, but I'm wondering if others have had success targeting a specific frequency range.

[page]

Clarification, is it really wind noise or is it HVAC rumble? I've found the consistancy and frequency ranges are different.

[runonce]

If its just rumble - High pass filter (sounds like the situation)

Wind - hard to fix decently...IME

[capnhook]

I've used a turntable rumble remover (Groove Mechanic > Fix Rumble) to get rid of HVAC rumble before.....YMMV

[acidjack]

Although caused by an HVAC, this isn't a rumble - it's a distinct sound from the actual air it's blowing.  Damn thing was up HIGH especially considering it's the middle of winter...

[page]

Although caused by an HVAC, this isn't a rumble - it's a distinct sound from the actual air it's blowing.  Damn thing was up HIGH especially considering it's the middle of winter...

that will be harder. If what you're hearing is the motor of the fan or compressor, that's fairly easy, otherwise you are trying to remove irregular wind gusts (since wind isn't a constant noise).

Got a 15-30 second sample in flac you could sendspace?

[taperj]

I've been finding iZotope RX to be well suited for things like this(I haven't used it specifically for rumble but have removed intermittent artifact sounds etc from recordings), if you can isolate it in the spectral editor you can generally remove it. It also has noise removal and click and pop removal etc. YMMV but it's a place to start. (if it's a single sound or steady tone you may also be able to try a notch filter but it has to be pretty much dead steady tone wise)

J

[runonce]

I've been finding iZotope RX to be well suited for things like this(I haven't used it specifically for rumble but have removed intermittent artifact sounds etc from recordings), if you can isolate it in the spectral editor you can generally remove it. It also has noise removal and click and pop removal etc. YMMV but it's a place to start. (if it's a single sound or steady tone you may also be able to try a notch filter but it has to be pretty much dead steady tone wise)

J

I think wind physically shakes the mic diaphragm so much...that it distorts the signal in a way that cant really be fixed.

Generally, you'll end up making it "less worse"

[taperj]

I've been finding iZotope RX to be well suited for things like this(I haven't used it specifically for rumble but have removed intermittent artifact sounds etc from recordings), if you can isolate it in the spectral editor you can generally remove it. It also has noise removal and click and pop removal etc. YMMV but it's a place to start. (if it's a single sound or steady tone you may also be able to try a notch filter but it has to be pretty much dead steady tone wise)

J

I think wind physically shakes the mic diaphragm so much...that it distorts the signal in a way that cant really be fixed.

Generally, you'll end up making it "less worse"

^^ probably correct.

[exxile]

I had a similar problem from a recording I made not so long ago but have never been able to fix it. If anyone can offer any suggestions to try and improve this it would be much appreciated, it's more noticeable during the quieter parts & when the bass is increased during playback. Cheers
Here's a short sample: http://www.sendspace.com/file/7ofrsa

[runonce]

I tape/taped in a small room dozens of times where I was within 15' of two small fans used to augment the output of an aged AC system. They were mounted right in front of the grills and ran constantly, one in the forefield the other in the background relative to my mic placement.

It sounds like you'r hearing the mechanical fan noise rather than seeing/feeling the impact of puffs of air hitting the diaphragms, as others have said.

I was always pulling small jazz groups, every bass solo or quiet passage on the drums w/brushes resulted in an elevation of the the broadband noise produced by the fan blades. 

Using iZotope RX I'd isolate (or record specifically ahead of time) a passage of "silence", that is, the fan noise, and train RX to recognize it as such.  Then it was just a function of deciding if I could eliminate it or if it was better to just reduce it's intrusion.  Even with the ability to train a DAW to recognize a pattern as noise, there still seems to be some data that's removed that ends up leaving you with the babbling brook effect. 

Applying the tool to only the quietest ranges or the lowest frequencies is one aspect of the compromise needed to be reached when applying noise reduction.  I also am quite aggressive in cutting out banter and tuning if the result of leaving it would be to accent the fan noise who's level is now sharing the stage equally at those times.

RX is pretty good, IMO, for taking a good bite out of actual wind noise.  Using the spectral repair tool and marking the "hottest" bottom zones usually dampens the buffeting quite well.

Good Luck, YMMV

See reply 4

[Shadow_7]

The easiest way to remove it is to not record it in the first place.  OMNI mics help as they're less sensitive to it.  I live in a region windy enough that I've literally got three layers of fake fur over the mics to handle regional conditions.  And I still occassionally get some of it.  It's just faint enough to not need to be addressed most of the time.  As opposed to being 90%+ of the audible portion of the recording.

What's the subject matter?  My $100 DSLR is terrible with wind noise.  For one talking head video I shot on it (before getting better gear), I EQ'd out every frequency range not significantly relavent to the source(female voice).  Then amped it up with intentional clipping to bring the voice up to the audible range in competition with the wind noise, and clipping out most of the wind noise.  Surprisingly less worse.  As in still worse / horrid.  But you could make out what was being said better.  Better than you could hear it on location with the wind.  Or in the original unedited audio.

A/C vents can be influential.  It's why I always keep my furs on indoors and out.  Foam only just doesn't cut it IMO.  For the spoken word foam is fine.  But vents can pump out significantly more volume and speed of air per unit of time than a human mouth.  Having proximity away from those sources helps.  Having proximity towards the sound source helps.  In studio you can turn those units off while tracking.  On location, not so much as live human bodies pump out a lot of heat 98.6F in an enclosed space(assuming everyone is healthy and normal).