Make boomy audio more intelligible

[[email protected]]

Every once in awhile, I get in a recording (for news purposes) in
which the voice is swamped in natural "reverb", such as a person
speaking from a podium but being picked up by an audience mic. I
sometimes use a parametric eq to find the most intelligible
frequencies, then boost them. The goal is to get it as AUDIBLE as
possible for quick turnaround use on a news radio program.

With today's technology, is there some more effective way to do this?

[Hueyduck[_3_]]

a écrit :

> With today's technology, is there some more effective way to do this?

Not very new, and a bit obvious, but if the voice is always louder than
the ambiance, could you try an expander?

Ratio 2:1 below the level of the voice, for instance.

Huey

[[email protected]]

On Apr 6, 1:48*pm, Hueyduck > wrote:
> a écrit :
>
> > With today's technology, is there some more effective way to do this?
>
> Not very new, and a bit obvious, but if the voice is always louder than
> the ambiance, could you try an expander?
>
> Ratio 2:1 below the level of the voice, for instance.
>
> Huey

No, it's often buried in it. Our recordings come from all kinds of
sources over which we have no control. I just wondered if there's some
new technology that can pull the voice out through some phase trick or
something.

There is some software that allows one to go back into a mono musical
piece and correct a bad note on one instrument, so I was hoping that
some similar magic could solve my problem.

[Arny Krueger]

> wrote in message


> Every once in awhile, I get in a recording (for news
> purposes) in which the voice is swamped in natural
> "reverb", such as a person speaking from a podium but
> being picked up by an audience mic. I sometimes use a
> parametric eq to find the most intelligible frequencies,
> then boost them. The goal is to get it as AUDIBLE as
> possible for quick turnaround use on a news radio
> program.

> With today's technology, is there some more effective way
> to do this?

The technical term for what you are looking for is blind dereverberation.

Google will get you info about the SOTA in this area.

The demos I've seen are not all that impressive.

[Ty Ford]

On Wed, 6 Apr 2011 12:19:41 -0400, wrote
(in article
>):

> Every once in awhile, I get in a recording (for news purposes) in
> which the voice is swamped in natural "reverb", such as a person
> speaking from a podium but being picked up by an audience mic. I
> sometimes use a parametric eq to find the most intelligible
> frequencies, then boost them. The goal is to get it as AUDIBLE as
> possible for quick turnaround use on a news radio program.
>
> With today's technology, is there some more effective way to do this?

No.

Regards,

Ty Ford



--Audio Equipment Reviews Audio Production Services
Acting and Voiceover Demos http://www.tyford.com
Guitar player?:http://www.youtube.com/watch?v=yWaPRHMGhGA

[[email protected]]

On 2011-04-08 said:
>> Every once in awhile, I get in a recording (for news purposes) in
>> which the voice is swamped in natural "reverb", such as a person
>> speaking from a podium but being picked up by an audience mic. I
>> sometimes use a parametric eq to find the most intelligible
>> frequencies, then boost them. The goal is to get it as AUDIBLE as
>> possible for quick turnaround use on a news radio program.
>> With today's technology, is there some more effective way to do
>this?
>No.

I have to agree with Ty here. That's when whoever makes the
recording needs to learn how to boom. Ask guys on ramps,
some of them boom and handle the rest of the audio chores
simultaneously. They can give you some hints and kinks, and
they could use a good educational thread over there lately.




Richard webb,

replace anything before at with elspider
ON site audio in the southland: see www.gatasound.com

[Richard Webb[_3_]]

FOllowing up my post for Blackburst:
> I have to agree with Ty here. That's when whoever makes the
> recording needs to learn how to boom. Ask guys on ramps,
> some of them boom and handle the rest of the audio chores
> simultaneously. They can give you some hints and kinks, and they
> could use a good educational thread over there lately.

Ramps btw, though I think you've been around here a few
years, so probably know this, is
rec.arts.movies.production.sound newsgroup.

THere are guys over there who do this so-called reality tv
stuff, and often they do the run and gun booming when
appropriate and/or necessary. THat's your key to good audio capture in those situations imho.
Ask some of those guys for some hints and kinks.

Regards,
Richard
.... Remote audio in the southland: See www.gatasound.com
--
| Remove .my.foot for email
| via Waldo's Place USA Fidonet<->Internet Gateway Site
| Standard disclaimer: The views of this user are strictly his own.

[[email protected]]

Thanks everybody for the help.

I don't think I made clear my situation. I am currently the producer
of a politically-focused radio program. Our "open" to the program is a
montage of the most significant political audio clips of the day.
These come from a variety of sources in the US and around the world,
over which I have no control. Most are reasonably well-recorded, but a
few are recorded by amateurs. I was simply wondering if there is a
way, beyond parametric eq, to make the voice more intelligible when it
is badly recorded with lots of natural reverb. Apparently there is
not. Thank you all!

[Scott Dorsey]

> wrote:
>
>I don't think I made clear my situation. I am currently the producer
>of a politically-focused radio program. Our "open" to the program is a
>montage of the most significant political audio clips of the day.
>These come from a variety of sources in the US and around the world,
>over which I have no control. Most are reasonably well-recorded, but a
>few are recorded by amateurs. I was simply wondering if there is a
>way, beyond parametric eq, to make the voice more intelligible when it
>is badly recorded with lots of natural reverb. Apparently there is
>not. Thank you all!

Multiband compression can be a useful tool for quick and dirty unattended
equalization. You can set an 8-band compressor, for instance, to always
shoehorn any incoming signal into a particular frequency envelope.

This would be a pretty horrible thing to do to music, but it can be a
useful tool for enhancing speech. It's nowhere near as effective as
having a skilled operator do the job, but it's something you can put in
your signal chain when you don't have a skilled operator available.

One common application for this sort of thing is telephone feeds on
call-in shows. If the incoming call is bass-heavy or bass-light, it
always gets poked and prodded to have about the same balance of
frequencies.

It won't remove reverb, it won't clean up echoes. But it can eliminate
tubbiness on tubby signals while at the same time pumping the low end
on thin signals too. It can always add a little presence peak when it's
appropriate and leave it unchanged if there is already one there.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[[email protected]]

On 2011-04-09 said:
>I don't think I made clear my situation. I am currently the producer
>of a politically-focused radio program. Our "open" to the program
>is a montage of the most significant political audio clips of the
>day. These come from a variety of sources in the US and around the
>world, over which I have no control. Most are reasonably
>well-recorded, but a few are recorded by amateurs. I was simply
>wondering if there is a way, beyond parametric eq, to make the
>voice more intelligible when it is badly recorded with lots of
>natural reverb. Apparently there is not. Thank you all!

Understood. DIdn't know because of some of the other stuff
you do if it were content captured/generated by you and your
staff. IF so wanted to help you educate the folks doing the
capture if that were the case. SOmetimes all you can do is
reduce the suck factor as much as possible and go with it.
SUch is life.

REgards,




Richard webb,

replace anything before at with elspider
ON site audio in the southland: see www.gatasound.com

[Bill Graham]

wrote:
> Thanks everybody for the help.
>
> I don't think I made clear my situation. I am currently the producer
> of a politically-focused radio program. Our "open" to the program is a
> montage of the most significant political audio clips of the day.
> These come from a variety of sources in the US and around the world,
> over which I have no control. Most are reasonably well-recorded, but a
> few are recorded by amateurs. I was simply wondering if there is a
> way, beyond parametric eq, to make the voice more intelligible when it
> is badly recorded with lots of natural reverb. Apparently there is
> not. Thank you all!

Sometimes it is possible to make a recorded voice more intelligible by
cutting off the low frequencies and boosting the highs. This will make it
squeeky and probably unrecognizeable, but it may become more intelligible.

[Mark]

On Apr 9, 11:14*am, " > wrote:
> Thanks everybody for the help.
>
> I don't think I made clear my situation. I am currently the producer
> of a politically-focused radio program. Our "open" to the program is a
> montage of the most significant political audio clips of the day.
> These come from a variety of sources in the US and around the world,
> over which I have no control. Most are reasonably well-recorded, but a
> few are recorded by amateurs. I was simply wondering if there is a
> way, beyond parametric eq, to make the voice more intelligible when it
> is badly recorded with lots of natural reverb. Apparently there is
> not. Thank you all!

if you have ANY control at all over the recordists, one thing you can
suggest is that the original recordings be made with MANUAL level
control and not AGC.

Misuse of AGC makes reverberation worse.

Mark

[[email protected]]

On Apr 9, 2:04*pm, (Scott Dorsey) wrote:
> > wrote:
>
> >I don't think I made clear my situation. I am currently the producer
> >of a politically-focused radio program. Our "open" to the program is a
> >montage of the most significant political audio clips of the day.
> >These come from a variety of sources in the US and around the world,
> >over which I have no control. Most are reasonably well-recorded, but a
> >few are recorded by amateurs. I was simply wondering if there is a
> >way, beyond parametric eq, to make the voice more intelligible when it
> >is badly recorded with lots of natural reverb. Apparently there is
> >not. Thank you all!
>
> Multiband compression can be a useful tool for quick and dirty unattended
> equalization. *You can set an 8-band compressor, for instance, to always
> shoehorn any incoming signal into a particular frequency envelope.
>
> This would be a pretty horrible thing to do to music, but it can be a
> useful tool for enhancing speech. *It's nowhere near as effective as
> having a skilled operator do the job, but it's something you can put in
> your signal chain when you don't have a skilled operator available.
>
> One common application for this sort of thing is telephone feeds on
> call-in shows. *If the incoming call is bass-heavy or bass-light, it
> always gets poked and prodded to have about the same balance of
> frequencies.
>
> It won't remove reverb, it won't clean up echoes. *But it can eliminate
> tubbiness on tubby signals while at the same time pumping the low end
> on thin signals too. *It can always add a little presence peak when it's
> appropriate and leave it unchanged if there is already one there.
> --scott
> --
> "C'est un Nagra. C'est suisse, et tres, tres precis."

Thanks for this suggestion, Scott and Flatfish. I hadn't thought of it
(because I've never used it in this way) but it just might help!

[Ty Ford]

On Sun, 10 Apr 2011 11:05:49 -0400, wrote
(in article
>):

>
> Thanks for this suggestion, Scott and Flatfish. I hadn't thought of it
> (because I've never used it in this way) but it just might help!

usually the cure is as bad as the disease.

or like I said before, no.

Regards,

Ty Ford


--Audio Equipment Reviews Audio Production Services
Acting and Voiceover Demos http://www.tyford.com
Guitar player?:http://www.youtube.com/watch?v=yWaPRHMGhGA

[[email protected]]

On 2011-04-10 said:
>> Thanks for this suggestion, Scott and Flatfish. I hadn't thought
>>of it (because I've never used it in this way) but it just might
>help!
>usually the cure is as bad as the disease.
>or like I said before, no.

But which way would you rather have it? Pick your poison in
other words. Shoot for intelligibility, get rid of the most
undesirable characteristic(s) as much as possible and call it a day.




Richard webb,

replace anything before at with elspider
ON site audio in the southland: see www.gatasound.com

[Scott Dorsey]

In article >, > wrote:
>On 2011-04-10 said:
> >> Thanks for this suggestion, Scott and Flatfish. I hadn't thought
> >>of it (because I've never used it in this way) but it just might
> >help!
> >usually the cure is as bad as the disease.
> >or like I said before, no.
>
>But which way would you rather have it? Pick your poison in
>other words. Shoot for intelligibility, get rid of the most
>undesirable characteristic(s) as much as possible and call it a day.

When I started in radio, it was normal for there to be an announcer,
a disk jockey, and a control engineer on duty at the same time. That
was on top of the guy on duty at the transmitter facility. The
control engineer just rode gains and equalized program feeds to make
everything sound uniform.

Now you're lucky if there is a DJ on duty rather than an automated
playback server. So, with only one person doing the work of three
or four, something has to give, and this is one of the things that
gives. The machine isn't as good as a real human being doing it,
but it's better than nobody doing it at all.
--scott

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

[Scott Dorsey]

In article >, > wrote:
>IN the controlled world of the recording studio or the
>soundstage you've got that "do over" if you need it.
>THis obviously isn't NPR with decent telephone hybrids and a
>board op who sees it for air after an editor has already
>seen and worked with it.

Do they? Morning Edition drives me up the wall because they don't seem
very careful with intercutting field recordings. I mean, really, at least
notch the hum....
--scott

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