What's wrong with this audio

Jim H wrote on 12/3/2011, 7:35 PM
Can some one tell me what happened to the audio in this video? And if given the original file, could it be corrected? Hoping that some of you sound guys recognize the problem. I'm told the creators did "quite a bit of work" on the audio to get it to this stage. Not sure what that means but if they're asking ME for advice, I'm afraid to guess what they did to it. I've pitched in for this group on graphics before, but I'm no sound engineer. Thanks.


Comments

Steven Myers wrote on 12/3/2011, 7:54 PM
What's wrong is that it sounds like it has been subjected to heavy noise reduction. Can this one be fixed? I'd say no.
Can the original be fixed better than the example you've posted? No way of knowing without having access to the original.
Jim H wrote on 12/3/2011, 8:04 PM
Steven, I agree that's it's better to work on the original and I asked for it. I was just wondering if the problem was immediately recognizable. It almost sounds like they might have been recording from the lav radio and were getting some interference from the PA? Or if they're really novices, maybe they've got both camera's audio tracks on? Though to your point, that echo like effect may be a result of their post processing.
Steven Myers wrote on 12/3/2011, 8:14 PM
What's immediately recognizable to me is the noise-reduction artifacts. Unfortunately, I'm not good enough to hear through that, back to whatever it was they were trying to correct in the first place. :)
Dan Sherman wrote on 12/3/2011, 9:26 PM
The guy with the mask on sounds pretty good, but his voice is supposed to be distroted, right?.
Maybe reshoot the intro and have the host wear a mask?
Seriously. looks like an overzealous attempt at noise reduction.
Someone was trying to filter out an offending frequency.
Maybe that annoying elevator bell and the talking from the hallway.
If you can find someone with iZotope RX, that may help.
Or send me the original of the intro only and I'll see what I can do.

Dan
Jim H wrote on 12/3/2011, 10:17 PM
Thanks for the feedback. It was a live press conference so I don't think reshooting is possible - though with the spy from Iran it does look like a bad movie. If I can get my hands on the original audio track I'll post it.
BibbityBoo wrote on 12/4/2011, 1:14 AM
Do you have any feedback on whether these defects are present in the orignal upload? I ask because YouTube is going through some major mess lately with transcoding, some of which involves serious packet mangling that usually translates into some blocking artifacts, but could just as easily affect the audio stream as well.

This does sound to me like something one might have done my overdoing it with the Track Noise Gate filter, but as others point out, without the source audio and ideally the project file used to render, there's almost no way to pin it down to just one thing.

Have your contacts said anything specific about whether the dropouts are present in their originals? Sorry if I passed over something already covered in the thread to date.
Mike M. wrote on 12/4/2011, 1:22 AM
Is the original source this bad ? Like some others here have mentioned, it sounds like compression artifacts and noise...........typical of a low quality audio codec used from the original to YouTube.
PeterDuke wrote on 12/4/2011, 2:17 AM
Yes, it sounds a bit like the "talking under water" effect you get from some noise reduction methods.
JJKizak wrote on 12/4/2011, 7:00 AM
Noise reduction (too much) or out of phase international communication. (One countries comm system to another's.) These systems are usually phase connected with a 64.000 KHZ frequency which used to be generated in New York. Not sure anymore as this was 50 years ago.
JJK