Frage von Lindow:
Hi,
in a sound recording with lots of wind, I had stupidly my windshield (clip) forgotten. Now the "Windblubberer-noise 'between the language. They should be ironed out, or removed.
I work with Soundtrack Pro in FCS. My previous tests have shown that it mainly concerns the frequency of 125 Hz. With the EQ I have turned down this Frepuenz and it is much better, but could be even better because the wind noise is not completely gone. In the language is that frequency is not available.
Who can give me some good suggestions to improve the still. Is there an "anti-wind-Filter"?
Antwort von masterseb:
no, something like an "anti-wind filter" does not exist. what do large productions is: in all breaks when wind occurs, cut and remove ton. the need to mix it well so that sounds good.
Audition (Adobe) was in the cs2 version once a kind of "content-aware fill", one could in the waveform and select remove noise (supposedly little distortion / interference of the residual signal)
Antwort von Lindow:
no, something like an "anti-wind filter" does not exist. what do large productions is: in all breaks when wind occurs, cut and remove ton. the need to mix it well so that sounds good.
Audition (Adobe) was in the cs2 version once a kind of "content-aware fill", one could in the waveform and select remove noise (supposedly little distortion / interference of the residual signal) Hi,
on the wind-breaks, unfortunately I could not wait because the interviewee was quite nervous anyway and the recording was done quickly.
Antwort von Lindow:
no, something like an "anti-wind filter" does not exist. what do large productions is: in all breaks when wind occurs, cut and remove ton. the need to mix it well so that sounds good.
Audition (Adobe) was in the cs2 version once a kind of "content-aware fill", one could in the waveform and select remove noise (supposedly little distortion / interference of the residual signal)
Hi,
on the wind-breaks, unfortunately I could not wait because the interviewee was quite nervous anyway and the recording was done quickly. Hi I have just misunderstood, with the wind breaks. Did you wind breaks between words. Because stupid is that the wind synonymous within the words is, there is nothing to cut because there are parts of the word synonymous away then.
So I'll probably stick with my method with the 125 Hz frequency.
Antwort von thos-berlin:
I do not know if the breaks are long enough yet.
At least my audition can take a "fingerprint" and then "finger print" count out from the total signal.
So a point mark in a speech for use and as such fingerprint.
Whether this leads to a satisfactory result, I do not know but PC fan noise can be so reasonably (!) Remove well.
Difference between Lüftergeraäusch and Wind:
Fan noise is steady, no wind .....
Antwort von Lindow:
I do not know if the breaks are long enough yet.
At least my audition can take a "fingerprint" and then "finger print" count out from the total signal.
So a point mark in a speech for use and as such fingerprint.
Whether this leads to a satisfactory result, I do not know but PC fan noise can be so reasonably (!) Remove well.
Difference between Lüftergeraäusch and Wind:
Fan noise is steady, no wind ..... Hi Timm,
it is synonymous in SoundtrackPro. Just got this fingerprint tried to suppress noise and works well.
I knew not yet, thank you for your tip.
Antwort von stefangs:
caution with this fingerprint algorithms. which are fine, but in the first euphoria you can hear on the effect of the noise-excited and this ueberhoert the destruction of the desired signal.
I would, if you want the settings to use at all, take a combination of both (fingerprint and eq), both with mild. I say 'at all' at the fingerprint, because wind actually not so suitable for that is because it is a constantly changing signal is. you take a fingerprint usually better in uniform sounds like humming, air conditioners, etc.
the 125 hz, I would not take a 'notch', but rather a high pass with nerve ramp in the 125Hz-area sample and then, if it sounds better lower to start and for steeper (eg 36 db / oct) or higher to start and flat set (eg 12 db / oct). depends of the vote from the (male / female) and of the strength of the windgeraeusches. if the signal is in stereo, synonymous check in mono. steep filter slopes can disturb the phase and then suddenly it sounds weird in mono. in language you could but both channels * before * the processing in mono mix together and then adjust or equal to only take a channel as the spatial effect is possibly no preference.
Good luck,
stefan
