DV Info Net

DV Info Net (https://www.dvinfo.net/forum/)
-   All Things Audio (https://www.dvinfo.net/forum/all-things-audio/)
-   -   foiled again! (https://www.dvinfo.net/forum/all-things-audio/496109-foiled-again.html)

Kris Zimbelman May 18th, 2011 09:26 PM

foiled again!
 
I used a zoom H1 model audio recorder at a recent event. I've violated the cardinal rule of not using a headset to check the sound while he was recording. I had it set up on a Cabinet and behind it on the wall was an old-fashioned analog clock with a ticking secondhand. Are there any software solutions in Sony or Apple software that can remove this ticking? After I discovered this, I took a recording of the clock only thinking that having that sound distinctive may help me later go in and remove this anomaly. Thanks in advance for your help!

Greg Miller May 18th, 2011 10:38 PM

Re: foiled again!
 
I'd think it's a function of the frequency signature of the clock. If it's a fairly narrow frequency or frequencies, and especially if they are not in the most important part of the voice range, you might manage to reduce the clock noise somewhat. If it's a very wide frequency band, you will probably have less luck filtering it. I think it's very unlikely that you can eliminate it altogether. If the clock is as loud as the voice, you are probably SOL; if it's relatively quiet, you can try a few different techniques to reduce it.

As usual, I say: post a sample so we can comment on the actual problem, rather than just speculating about what might work.

Rick Reineke May 19th, 2011 09:12 AM

Re: foiled again!
 
I did this with a click-track bleed issue: If you have the ticking from the same source, location, placement and amplitude.. throwing it 180 degrees out of phase on another track should eliminate it. The problem will be the tracks drifting out of sync, and will need to be 'adjusted' regularly. You will need a DAW or NLE that allows moving audio clips/events down to the sample level and a phase invert option. In any case, there's no magic plug-in that I'm aware of and will likely take a lot of time.
Or you could try finding the precise frequency and bandwidth of the tick, and notch it out with a parametric EQ. The problem with this, it will also affect the other program material on that track.
A noise-reduction plug-in may also help, but I have never tried it.

Greg Miller May 19th, 2011 10:32 AM

Re: foiled again!
 
Rick, that sounds like a good theory. It would probably work quite well with a click track. I wonder if the tick-to-tick timing of the clock is as accurate and repeatable as a digitally generated click track. For that matter, does the clock's "tick" sound exactly the same each time? It's an interesting question and I'd love to play with a sample track...

Rick Reineke May 20th, 2011 07:56 AM

Re: foiled again!
 
Good point Greg. (the accuracy of the offending tick, tick, tick).
In my tenure, annoying clocks were always disabled prior to recording.. by whatever means possible, including hammers, or re-locating it out of audible range.

Tom Morrow May 20th, 2011 09:57 AM

Re: foiled again!
 
Adobe Soundbooth (just obsoleted I think) has a feature that lets you see an audiogram to visually find periodic noises, then you can drag a rectangle around them (in frequency and time dimensions) and delete. The help system had a really cool example of deleting a cell phone ring but that was at a single frequency; you could literally see the cell rings as bright circles on the audiogram.

Anyway, because it's obsolete and requires manual intervention this program probably isn't for you, but there must be something more sophisticated that will do this.

My guess is that because a clock tick comes from friction on a wide range of surfaces, that it will not have a narrow frequency range, and therefore will be difficult to seperate.

Greg Miller May 20th, 2011 12:37 PM

Re: foiled again!
 
Adobe Audition had the same feature. I very much doubt it would work for something that is not a narrow frequency band. And, even if it did work, imagine doing that process, manually, for one tick per second, for the duration of the entire file. ROTFL

As I said, if we had an audio clip, we could see what frequencies the clock occupies, and then could make more specific suggestions.

Kris Zimbelman May 21st, 2011 06:05 AM

Re: foiled again!
 
I appreciate all of your replies and helpful suggestions. I will post a sampling as soon as possible Greg.


All times are GMT -6. The time now is 01:29 AM.

DV Info Net -- Real Names, Real People, Real Info!
1998-2024 The Digital Video Information Network