What can be done to distorted audio?

A

andrew00

Well-known member
I filmed a friend's band last night, to test out my new Canon 5dmark2, which performed really well.

I did as I often do, recoded the audio from the desk to my Zoom H4. I had the line sensitivity to low and moved the levels of the audio to about -24 to -18, which I've done many times before.

And yet the audio is distorted. Which makes me believe the sound man was giving me too hot a feed from the desk or something, as everything, from vocals to background sounds like it's clipping in some form.

Is there anything I can do to this audio? I know I can't remove the distortion, but I was wondering if I can do anything to minimise it - filters, audio effects etc?

(I'll post a sound sample later, have to pop out now.)
 
andrew00 said:
I filmed a friend's band last night, to test out my new Canon 5dmark2, which performed really well.

I did as I often do, recoded the audio from the desk to my Zoom H4. I had the line sensitivity to low and moved the levels of the audio to about -24 to -18, which I've done many times before.

And yet the audio is distorted. Which makes me believe the sound man was giving me too hot a feed from the desk or something, as everything, from vocals to background sounds like it's clipping in some form.

Is there anything I can do to this audio? I know I can't remove the distortion, but I was wondering if I can do anything to minimise it - filters, audio effects etc?

(I'll post a sound sample later, have to pop out now.)
Click to expand...

Very likely you were overloading the line input on the Zoom. The unbalanced input is rated for a -10dBm consumer line level signal on its lowest sensitivity setting. I'll bet the board gave you a +4dBu line level, 12 dB or 14 dB too hot. I'll bet the FOH guy was running his mix as loud as he could drive the speakers, clipping be damned, and you have a recipe for disaster. Lowering the audio recording level in the zoom doesn't help in that sort of situation as the clipping and distortion occur before the recording amplifier stage where the level control acts. All you're doing at that point is lowering the level of an already distorted signal.

It's not likely that there is much you can do to make it better - GIGO.
 
Thanks for your reply.

That's interesting as I had the Zoom on the lowest sensitivity, and I've recorded loads of gigs before on it where I've had no trouble with distortion - I know the H4's preamps aren't world beaters, but they've never caused me issues before. Hence why i thought it was a problem with the feed I was getting from the desk itself.
 
andrew00 said:
Thanks for your reply.

That's interesting as I had the Zoom on the lowest sensitivity, and I've recorded loads of gigs before on it where I've had no trouble with distortion - I know the H4's preamps aren't world beaters, but they've never caused me issues before. Hence why i thought it was a problem with the feed I was getting from the desk itself.
Click to expand...

Another case where the fact that there are two very different definitions of "line level" in the audio world cause problems. Pro line levels are +4dBu or 1.228 volts. Consumer line level is -10dBv, or 316 millivolts. That's a BIG difference in level. Some boards and mixers - my Mackie, for example - sends +4dBu pro line to the main, aux, and direct outputs but also sends the main mix at -10dBv consumer level to the tape outputs. It's not enough to know the board is sending you a line level signal, you have to know WHICH line level he's giving you. If I give you a line level feed from an aux bus it will be too hot for you; if I give you one from the tape out it will correct for your Zoom. Or if you add a 10 or 15dB pad to the line you can take the aux feed.
 
There are some apps like RX from iZotope that have "declipping" functions. They might help, it's hard to say till you try. I have had excellent results on totally blown out material and it's not worked at all on other things so YMMV. I think there is a demo.
 
To further copy Scott and Chadfish, I've seen Izotope RX do some amazing things with distorted audio.
 
Hey,

As an update to this - I ended up picking up iZotope RX, and on the whole it's been a good move, although it's unable to fix one problem, described below, I wonder if anyone has any suggestions?

- I recorded the audio at lowest pre-amp settings on my Zoom & w/low levels, but the signal was too hot.
- So I imported the audio into Sountrack pro, added a bass guitar compressor - which sounded the best at reducing the distorted audio, and increased the levels so they came closer to 0.
- I then imported that into iZotope RX, and, now that the volume was at a higher level, the de-clipper setting (which I subsequently realised also works as a plugin to Soundtrack!) did a good job of skimming off the distorted sound.
- That worked well for a quieter hip hop artist, who was using live instruments.

- However, the artist I'm now editing, from the same event, was using a backing track and the the beat from the backing track is too distorted, even after playing around with it as above.
- This means I feel I can't 'recover' it - but I was wondering if I can reduce it - i.e. what I can do to basically quieten the beat to the point where the distortion is background.
- I've tried reducing the bass in the EQ and various low cut's etc, but they're not having the effect.
- I'm concerned that the beat isn't actually 'bass', its merely a mid-range spike that is unrecoverable.
- So I was wondering if anyone could suggest anything I can do.

(Oh and to quantify, I can get rid of the bass rumble from the beat, with a low cut, but can't get rid of the distorted pop from it)

An alternative, thinking about it, is to get the backing track from the artist himself. In this case is there anything I can do to quieten the backing track whilst keeping his vocals on the live recording and add the studio backing track to that? I know the sounds will be all jumbled up, but I wonder if there's a frequency range or anything like that I can work with.

Cheers!
 
Last edited:
If you have a vocal track separate from the backing music, then get the CD of the backing music. It will sound much better.

As far as any major distortion is concerned: The saying goes, "You can't polish a turd". You can only do so much. Garbage in - garbage out.
 
Aye it's all in one stereo track unfortunately, no separate vocal track.

The one thing I can think if to s ync it to the studio recording, since the guy was following a backing track it should all be the same tempo etc. Not ideal, but perhaps better than mud sound.
 
If the studio recording has vocals on it, just use that! Then just fade him in when he says stuff like, "C'mon let me hear y'all SAY YEAH!!!"
 
I record concerts from the board as well - i keep the zoom input audio recording level to something between 2 and 5. which seems very low, but often times the feed from the board is to high, so just ask the sound guy to turn it down. Even show him the zoom, and show him that its clipping
 
Ah I recorded this with the H4, not the H4N so the settings were different.

The pre amps are controlled on the side and have a L M H settings, they were set to low. The levels are controlled in the menu's, which are the levels post-pre amps.

Does that mean the H4N have more control over the pre-amps then?
 
What about Soundtrack Pro? Is there anything in there that could help? I had the same issue with my H4N a few weeks ago on a shoot. I set my levels before the show with the soundguy and everything was good. He told me that once the show started he was going to boost his levels some. So I backed my H4N levels down to a tad above -24 in stereo mode. The pre show stuff that was playing while people were finding their seats sounded excellent. When the show started the emcee's audio sounded good too. The problem came when they potted up the audio for the music. A lot of my audio was distorted and in my opinion not useable. The bass killed me!

I am currently demoing several audio filters in FCP to see if I can at least get rid of the bass. I have no clue what filter is best for this. I did have a back up audio track from the camera location. For some of the songs, I am able to replace the distorted H4N audio with CD's the groups performed with and sync it up in FCP. I then mixed it with the audio from my camera location to give it that sound as if that song was the live version (had that track low but loud enough to hear ambient sound from the arena) and then mix in crowd cheers at the end of the performances from mics at the camera location. I know this sounds crazy but that seems to be working for now.

Next time I think I just need to hire my own sound guy to focus just on getting me good audio to use. This is my 3rd time shooting an event like this and working with the HOUSE sound guys and my audio isn't right yet! Thought sure the H4N was going to be a lifesaver but I fell victim to distortion too! LOL!

I just wished I was able to have someone stand next to my H4N during the show to monitor the levels and make adjustments when the audio got to hot. My camera location was no where near the H4N location so i couldn't do both. There were times when the audio was to low as well! I had to do an old trick and double up the audio track in FCP to make those tracks appear louder or normal volume instead of raising the levels on the low level tracks putting them in the red

I was wondering if I took my distorted audio and imported it into STP if there was anything I could to somewhat fix it. I know there is basically nothing that came be done with distorted audio but like the original poster asked, is there any filters to consider that might not be in FCP but in Soundtrack Pro?
 
Last edited:
DVX100Shooter said:
The pre show stuff that was playing while people were finding their seats sounded excellent. When the show started the emcee's audio sounded good too. The problem came when they potted up the audio for the music. A lot of my audio was distorted and in my opinion not useable. The bass killed me!
Click to expand...
Par for the course with live performances; every musician kicks it up three or four notches when they hit the stage for the performance. They're hyped up on adrenaline, booze, drugs or any combination thereof, plus the fact that a house full of people soaks up a lot of sound so the FOH mixer bumps it up by 10% to 20%. Next time drop your volume levels by at least a third, it will get louder as the night goes on. Trust me, I pounded stages for 20 years, starting in dive bars and working my way up to major concert halls and stadiums.
 
Well I bumped it down from the get go but looks like I will have to bump it down even further next time to avoid clipping or distortion....good advice nonetheless though!
 
One of the fundamental issues is treating the audio for video as a "set and forget" kind of thing. It isn't. A question such as "What are the right settings to record XXX with my YYY brand recorder?" just doesn't have an answer because there is no such thing as the "right" settings. Just like the FOH mixer needs to give his board and the mix constant attention and correction throughout the show, so recording the audio for a video of the performance needs similar constant attention to detail by someone focussed on the job (NOT the FOH mixer as he's mixing for the audience, not the video) and an equipment package that actually gives them the tools to do it.
 
You must log in or register to reply here.
Back
Top