Proper FCP7 Audio Filter to use for Distorted Peaks

Hello All,

And thanks in advance.

An actor friend of mine gave me a dvd of a performance from the Actors Studio.
Although the original videographers used three cameras, and have as good coverage as you can get for a theatrical piece, the audio they recorded is a disaster.

To be fair, they had a tough situation. The actor, (my friend), plays multiple characters conversing (with himself), on stage, and many times throughout the performance, one of his characters is screaming, and so they recorded a lot of blown out peaks.

Fortunately, despite this, the dialogue is audible.

He would like it if I can do something about the blown out peaks.

I am not an audio guy, (as you can sumise.)
But I can operate FCP7 pretty decently, and with the right direction, (where you all come in!), I could probably improve, (though not fix), this somewhat.

I am looking for direction as to what "audio filters," already present in FCP7, would remove these peaks, while still maintaining the audibility of the dialogue.

Perhaps it is a two stage process? Remove the peaks with "X," and then boost the overall gain?

I have both the AU filters, as well as the other audio filters that ship with FCS.

Any and all help is greatly appreciated.

Side note: It's 5:50pm Eastern Time, and I have to step out and will not be back till after 8pm. So I can't respond till then. Hopefully there will be a solution waiting for me when I get back!

> He would like it if I can do something about the blown out peaks.

Not much. ADR is usually the go-to solution. Once audio overmodulates, you can take it down 20dB and it'd still be soft overmodulated audio. Garbage in, garbage out.


www.derekmok.com

If the wave-form is topping out its ruined, it's distorted and you can't get rid of that. the best you can do is use the pen tool to stop it from peaking at least. Any audio filter applied to the clip will effect the whole clip and muffle the quieter dialogue, unless key-framed or you cut the peak either side and use dissolves.

Although if it's for a friend maybe he'd be available for some ADR? if he wants a decent end result it shouldn't be too much trouble, just record another partial take in a similar sized space (or add reverb) and mix it back in,

WM,

Hi and thanks for the replies.

BTW, I tried to reply to derekmok last night and was unable to connect to lafcpug. (I guess the site was down.)

ADR is totally out of the question.

It's a non-commercial theatrical performance, with one actor going in and out of several characters in real time, all interacting with each other throughout, and 75 minutes in length.
It's just a great piece of theater, (that he wrote and performed at The Actors Studio), and he wants it to be both preserved, as well as be more watchable.

Yes, in the trouble areas the waveform is totally topped out, flat across the board when his characters are loud. There are dozens of places where the audio peaks, and only a handful where it is too low. (20-1)

I know I can manually surf the levels through the entire piece, raising the barely audible levels up, and lowering the distorted peaks down, but, as little money is not involved, and time limited, I want to do it much more efficiently.

Last night experimenting, I, (think I), found that I could use the Compressor/Limiter, filter to reduce all the peaks to an acceptable level at once, and then only have to manually raise up the levels for the barely audible parts. Again, as the overdriven peak occurrences outnumber the barely audible ones 20-1, it'll save me a ton of time.

Is there any reason I should not use the Compressor/Limiter filter for this purpose?

And, is there another filter I am missing that would automate, ONLY, raising the lower levels up, while leaving the rest untouched?

Thanks again!

the filter effects the whole clip unless you cut the audio track and dissolve back to the normal audio. you could also try maybe lowering the mid and high-end frequency of the distorted sections with the HQ filter.

unfortunately this kind of thing is often chalked up to experience via feedback you give the director or sound dept, although irritating for the editor especially if your names going out on the work.

WM,

That is what a compressor does -- make the soft sounds louder and put a clamp on the loudest sounds. It also drastically increases the hiss, because hiss is part of what the compressor would make louder.

The order of the filters matters. You can try using EQ or high pass to drain out some of the low-end rumble, and then put a fat EQ to see if you can get a bit more low-end presence back to the audio. Or, noise-gate the hiss out first, and then compress what's left.

All this, however, hinges on the original signal. If there was no tone recorded to begin with, no amount of filtering will get it. And if you didn't have a proper microphone close to the actor, then chances are you're not just dealing with messed-up levels, you're also dealing with too much room reverb/echo, lack of presence, and too much distance. Practically none of these is fixable in post.

The best way to fix audio problems? Prepare for them before the shoot.


www.derekmok.com

I had nothing to do with the recording, and my names not going to be on it! I just learned from my friend that it will be screened in a college class.

I ended up doing what I previously wrote. Using the Limiter overall, and then raising up the low levels manually.

Though I had nothing to do with this recording, from previous experiences I have deduced that it is nearly impossible to get good audio recorded from a live theatrical production, that has such drastic level ranges, on a shoe string budget.

To record such audio decently, you'd need several mics placed in strategic places, multiple recorders, and someone familiar enough with the script who knows what's coming, to surf the levels.

Definitely not a low budget proposition.

Thanks again!

Izotope's RX offers extensive clean-up tools.
[www.izotope.com]

i havent used it so i don't know if what they offer is better than what you have already in Soundtrack Pro.
i suspect it is.

the advanced version is supposed to be able to remove unwanted reverb, which is some sort of black magic,
so maybe it can reduce actual distortion, if that's what you have.

i see there is a 10 day free trial, so you could give it a go.


nick

Thx Nick. I'll give Rx a go!