Best way to remove mic feedback in post?

[bob]

I recorded a live show where there are several instances of mic feedbacks. I
want to salvage as much as possible from those instances.

I have several audio/wave editors (adobe audition, soundbooth, audacity,
goldwave) at my disposal.

Most audio editors have notch filters, but to complicate things, the mic
feedback changes frequency slightly as it gets louder.

Is there a filter that automatically adjust to the frequency change and
remove the feedback?

[Don Pearce]

Use Adobe Audition in spectrum view. You can draw around the squeal
with the lasso tool and delete it. Works seriously well. Filters are
too coarse and clumsy for this job.

d

[Peter Larsen]

Probably not, this needs an operator. At a glance - using A3 - I would cut
up in segments and use FFT analysis and FFT eq. It may be possible to define
a few categories. It is possible to define very narrow filters, example for
50 Hz hum: 45 Hz 0 dB, 46 Hz 0 dB, 47 Hz 0 dB 48 Hz 0 dB, 49 Hz -36 dB, 50
Hz -36 dB, 51 Hz -36 dB, 52 Hz 0 dB, 53 Hz 0 dB, 54 Hz 0 dB, 55 Hz 0 dB. The
objective of the extra points is to tell the algorith that you want a sharp
corner.

The FFT equalizer is well suited for removing what you do not want because
it is phase linear. It is less well suited to compensate for minimum phase
deviations such as those caused by transducers. This is not a problem to
solve with the notch filter function. You must use 32 bit audio for optimum
sound quality, especially as it may end up being a multistage process. It
may be that you end up with the filters sharp enough, but have to do the
filtering twice.

Use splines, use dB display. Note: the software is broken by design, so you
MUST define filters in the actually used samplerate, I do not know whether
they fixed it in the new version. It reportedly doesn't have open append,
which I use a lot, so I skipped it. If you divide this in three piles of
numbered segments it may get batchable. Use - ta da - open append to
reassemble the audio, multiselect works, but select from last to first.

There is also frequency display editing in A3, it effectively does about the
same and is perhaps better suited for the actual problem, but you have
better control over the process by the above suggested method.

Kind regards

Peter Larsen

[geoff]

Or Sony SpectralLayers, or equivalent in Wavelab.

geoff

[Scott Dorsey]

No, there isn't. But I bet a nickel that the frequency change is
psychoacoustic and isn't an actual physical change. Look at a spectral
plot!

Years ago I would have recommended manual use of notch filters, but these
days there are a variety of retouching tools that will let you look at a
frequency/time/amplitude plot and rework it. CEDAR made the original one
but Izotope has one as well now.
--scott

--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[Mike Rivers]

On 10/13/2013 1:04 PM, bob wrote:
> I recorded a live show where there are several instances of mic
> feedbacks. I want to salvage as much as possible from those instances.

> I have several audio/wave editors (adobe audition, soundbooth, audacity,
> goldwave) at my disposal.

The tool for this job is a relatively new one (actually an old concept
with a new user interface) which in essence filters a defined frequency
range for a specific time period. It's usually called "spectral
equalization" and Adobe Auditon has a version of it.

You view the audio with a spectrogram display that uses color to
indicate amplitude and displays frequency as height along the vertical
axis. The horizontal axis is time. Feedback is pretty easy to see on
this dispay if you look sharp while the audio is playing. You "lasso"
the (usually bright red) blob using the mouse and press the "cut"
button. It takes a little finesse to take enough but not too much, but
with patience you should be able to clean up the recording pretty well.

Note that I said "with patience." This isn't an automatic process by any
means. You have to find the feedback and trim it out unobtrusively, but
these tools are capable of doing so much more and better than before we
had it available.


--
For a good time, call http://mikeriversaudio.wordpress.com

[Jason]

On Mon, 14 Oct 2013 17:54:30 -0400 "Mike Rivers" <mm1...@yahoo.com> wrote
in article <l3hp6r$ol0$1...@dont-email.me>

FWIW, you can increase the resolution of the spectrogram display by
right-clicking. There are 3 steps beyond the default. I guess this
reflects the fact that generating the display takes a fair amount of
computation. Right-clicking also allows selection of the range of
frequencies that is displayed - this came in handy when I was trying to
get rid of an ambulance siren in a live recording last year.

[Marc Wielage]

On Sun, 13 Oct 2013 10:04:59 -0700, bob wrote
(in article <l3ejr9$rpb$1...@speranza.aioe.org>):

> Is there a filter that automatically adjust to the frequency change and
> remove the feedback?

>------------------------------<snip>------------------------------<

Not quite automatically, but you can get some pretty good results with the
new iZotope RX3 standalone:

http://www.izotope.com/products/audio/rx/index.asp?hs13

Find the worst instances of feedback, let iZotope analyze it, and then you
can have it build some notch filters to knock it out as much as possible. My
experience is that you can generally remove at least 50% of problems like
this, then solve the rest just through careful level adjustment and EQ.

It's never perfect, but it'll be an improvement.

--MFW

[S. King]

another Plus-1 for spectral editing. The Spectral display and editing
feature in Audition makes cleaning up narration tracks easy compared with
wave-form editing. It is at least as big a jump in capability as moving
from analogue tape and razor blade to digital wave form editing. Various
mouth noises as well as extraneous noises like clothing rustles, stop
watch clicks, and script handling noise are easily removed... all not
really achievable or practical in waveform editing.

Steve King

[Trevor]

> another Plus-1 for spectral editing. The Spectral display and editing
> feature in Audition makes cleaning up narration tracks easy compared with
> wave-form editing. It is at least as big a jump in capability as moving
> from analogue tape and razor blade to digital wave form editing. Various
> mouth noises as well as extraneous noises like clothing rustles, stop
> watch clicks, and script handling noise are easily removed... all not
> really achievable or practical in waveform editing.

On the contrary I still find it better/easier to reduce plosives in waveform
edit, or fixed notch filter/s for any mains hum. Watch clicks or any other
short sharp spikes are probably easier done in waveform edit too IMO.
Spectral editing certainly works better for some things however.

Trevor.

[S. King]

I go back and forth depending on what works best for the problem.
However, I do find that certain clicks and mouth noises are much easier to
locate in spectral view--don't have to expand the wave form four or five
clicks to find the little nigglies. Hum and pervasive junk is definitely
better dealt with using notch filters or, for complex noise, the
restoration noise reduction tool. Then, there is the issue of maintaining
lip sync, where the spectral view and the lasso can remove the problem
without affecting shortening the audio track.

Steve King

[Mike Rivers]

>> another Plus-1 for spectral editing.

On 10/16/2013 1:29 AM, Trevor wrote:

> On the contrary I still find it better/easier to reduce plosives in waveform
> edit, or fixed notch filter/s for any mains hum. Watch clicks or any other
> short sharp spikes are probably easier done in waveform edit too IMO.

The better you can see what you want to remove, the better use you can
make of a spectral editor. That's the key to using that tool. One nice
feature of Spectral Layers is that if you identify the hum, you can tell
it to also process harmonics. This works better than multiple notches
with line buzz that always contains harmonics.

[Trevor]

"Mike Rivers" <mm1...@yahoo.com> wrote in message
news:l437sg$4jd$1...@dont-email.me...

>> On the contrary I still find it better/easier to reduce plosives in
>> waveform
>> edit, or fixed notch filter/s for any mains hum. Watch clicks or any
>> other
>> short sharp spikes are probably easier done in waveform edit too IMO.
>
> The better you can see what you want to remove, the better use you can
> make of a spectral editor.

True.

>That's the key to using that tool. One nice feature of Spectral Layers is
>that if you identify the hum, you can tell it to also process harmonics.
>This works better than multiple notches with line buzz that always contains
>harmonics.

Yep they usually do, but I can fine tune each notch better with a good notch
filter IME.
YMMV depending on the actual tools. Nice to have options in any case.

Trevor.

[Mike Rivers]

On 10/22/2013 4:19 AM, Trevor wrote:
>> One nice feature of Spectral Layers is
>> >that if you identify the hum, you can tell it to also process harmonics.
>> >This works better than multiple notches with line buzz that always contains
>> >harmonics.

> Yep they usually do, but I can fine tune each notch better with a good notch
> filter IME.

I bet you really can't, but Spectral Layers is complicated enough to use
so that if you have a way that works for you, you should stick with it.
And, really, there should be no reason to fine-tune each notch. Find the
right one (usually or 120.0 Hz will be close enough) and then just use
arithmetic to set the others. In Spectral Layers you can set the level
below which the "harmonic finder" will select harmonics, so if 480 Hz is
below -95 dBFS, it won't include it in the buzz "layer."

[Arny Krueger]

"bob" <nos...@hotmail.com> wrote in message
news:l3ejr9$rpb$1...@speranza.aioe.org...

>I recorded a live show where there are several instances of mic feedbacks.
>I want to salvage as much as possible from those instances.
>
> I have several audio/wave editors (adobe audition, soundbooth, audacity,
> goldwave) at my disposal.

The Very Narrow DTMF filter in Audition is my fave. Just plug in the
frequency of the feedback (obtain from spectral analysis with 65k points)
and go.

> Most audio editors have notch filters, but to complicate things, the mic
> feedback changes frequency slightly as it gets louder.

Do it in segments

> Is there a filter that automatically adjust to the frequency change and
> remove the feedback?

Never heard of it.

[Trevor]

"Mike Rivers" <mm1...@yahoo.com> wrote in message

news:l45o13$5vf$1...@dont-email.me...

> On 10/22/2013 4:19 AM, Trevor wrote:
>>> One nice feature of Spectral Layers is
>>> >that if you identify the hum, you can tell it to also process
>>> >harmonics.
>>> >This works better than multiple notches with line buzz that always
>>> >contains
>>> >harmonics.
>
>> Yep they usually do, but I can fine tune each notch better with a good
>> notch
>> filter IME.
>
> I bet you really can't, but Spectral Layers is complicated enough to use
> so that if you have a way that works for you, you should stick with it.
> And, really, there should be no reason to fine-tune each notch. Find the
> right one (usually or 120.0 Hz will be close enough) and then just use
> arithmetic to set the others.

Determining frequency is the easy part with any spectrum analyser, I still
want to fine tune notch width and depth.

>In Spectral Layers you can set the level below which the "harmonic finder"
>will select harmonics, so if 480 Hz is below -95 dBFS, it won't include it
>in the buzz "layer."

True, I obviously need more experience with Spectral Layers, but the
spectragram display already in SoundForge has allowed me to easily accertain
that for a decade or so. (and 3rd party tools before that)
But as I said, I'm all for options and new tools! One may work better for me
on one job, and the opposite for another.

Trevor.

[Scott Dorsey]

In article <l47kig$e64$1...@speranza.aioe.org>, Trevor <tre...@home.net> wrote:
>"Mike Rivers" <mm1...@yahoo.com> wrote in message
>news:l45o13$5vf$1...@dont-email.me...
>> On 10/22/2013 4:19 AM, Trevor wrote:
>>>> One nice feature of Spectral Layers is
>>>> >that if you identify the hum, you can tell it to also process
>>>> >harmonics.
>>>> >This works better than multiple notches with line buzz that always
>>>> >contains
>>>> >harmonics.
>>
>>> Yep they usually do, but I can fine tune each notch better with a good
>>> notch
>>> filter IME.
>>
>> I bet you really can't, but Spectral Layers is complicated enough to use
>> so that if you have a way that works for you, you should stick with it.
>> And, really, there should be no reason to fine-tune each notch. Find the
>> right one (usually or 120.0 Hz will be close enough) and then just use
>> arithmetic to set the others.
>
>Determining frequency is the easy part with any spectrum analyser, I still
>want to fine tune notch width and depth.

In the case of buzz, you want the deepest and most narrow filter you can get
away with. In the analogue world there was a lot of worrying about how tight
you could get because the buzz frequency would drift slightly with tape speed
and there was worrying about how deep you could get because the filter had
audible side effects.

But in the digital world where everything is rock-steady, the majority of
the fine tuning is deciding just how many of the individual harmonics of the
buzz you want to knock out.

[Blueberry Buddha]

I don't know if you have access to Izotop RX, but you should be able to get rid of that in stand-alone mode using "spectral repair". I hope this helps.

Ken / Blueberry Buddha (aka Eleven Shadows)
www.blueberrybuddha.com

(first time I've posted here in eons!)

[Blueberry Buddha]

Apologies, for some reason, I didn't see all the other replies at first, so I see this has been addressed quite well already.

Ken / Blueberry Buddha

[moto...@gmail.com]

Adobe Audition spectral editor all the way so easy just like lifting a layer off in Photoshop.
Takes just a few min to get the hang of it but the easiest quickest way to get rid of any feedback and no affect the underlying track.

By the time you use an FFT analysis and FFT EQ cut things up and batch em back i will have cleaned the whole track and have moved on.

Too bad you can't post a audio sample here just to show how easy and effective audition is but after all is it just 2017...

[geoff]

Or SpectraLayers

geoff

[jjaj...@netscape.net]

The new wave of 2017!!!

Actually, you can find Sony touting on YouTube about their Spectral stuff software. Shame Sony Music doesn't know how to use it!!

Jack

>
> geoff

[Scott Dorsey]

<moto...@gmail.com> wrote:
>Adobe Audition spectral editor all the way so easy just like lifting a layer off in Photoshop.
>Takes just a few min to get the hang of it but the easiest quickest way to get rid of any feedback and no affect the underlying track.
>
>By the time you use an FFT analysis and FFT EQ cut things up and batch em back i will have cleaned the whole track and have moved on.

The honest truth is that what the Audition spectral editor is doing IS an
FFT analysis and then FFT equalization, just with a convenient user interface.
It is in fact an idea that was pioneered by CEDAR, and Audition adopted the
basic design.

These days, years after the post you're replying to, there are an awful lot
of different options available. The days of riding the depth control on an
Orban 622 in realtime are long gone.