In this world where we have such amazing tools for compression and limiting, I continue to be amazing that virtually every radio outlet that takes listener calls or does interviews with guest over the phone has absolutely crappy levels.  Invariably you can hear the host at normal levels and any of the outside parties are less than half the volume -- often WAY less than half the volume.  It is so annoying that I can't stand to listen most of the time.  This is true of local sports stations, national talk programs, and even NPR.  I have observed this consistently across Sirius/XM and now IHeartRadio.  it is everywhere.
 
Why doesn't every radio station have software that automatically balances the levels for them?
 
And if they don't have that, can't they at least find an engineer that can turn the knob a little?

There are still radio stations? 

It's like those mixing tutorials you watch where the music volume is completely at odds with the instructor's voice. Where's the one place you would expect levels to be set right? 

craigb
There are still radio stations? 

I got it.  But if there is one thing they should be able to get right ...
 
 

Hi,
 
I would suggest that levels are tough when any group, or personality gets started, when they are trying to figure out what works or not, and how to talk and adjust things.
 
All in all, as time goes by, and one gets more comfortable on the air, the "mistakes" slow down and the levels tend to even themselves out.
 
However, there is a tendency, and in this case ESPN is the worst, where the commercials are intended to be louder than the voices of the announcers.

You do know that he's talking about audio levels, right? 
 
There should only be two factors:  Getting the volume to match with the correct amount of compression.

craigb
There should only be two factors:  Getting the volume to match with the correct amount of compression.

Right, and I would have thought this would all be automated by now.

craigb
You do know that he's talking about audio levels, right? 
 
There should only be two factors:  Getting the volume to match with the correct amount of compression.

That was an ooooppppssss, and it's cleaned up.
 
Thanks for the heads up

My take:
 
Those in charge of "critical systems" are insufficient due to those in charge of those systems being brought up on the wrong baby food (internet, video games, social networking & smart phones).
 
We went to the moon on vacuum tubes and duct tape.
Today we can seem to get anything right. Talk radio levels or otherwise, as old "Bert" has long since retired or died and nobody thought to care about what he did.
 
Has anyone noticed that GE is trying to get selfie kids interested in real science?
And the NFL is pleading with to go out and play.
 
Were so dead.

It's not impossible. Listen to the NPR or BBC podcasts and radio programs to hear how it's supposed to be done.
 
Those organizations have a history and culture of technical excellence. More importantly, most of their content is not live, but recorded and then carefully edited. No automated processing can possibly match competent hand-editing. They are also typically recorded in well-treated purpose-built rooms and attended by experienced engineers. You don't find that at local radio stations where most sports and political talk radio originates.
 
But I think the main reason talk radio sounds so bad is limitations of the telephone and Skype connections they rely on. More and more, interviewees are talking over cell phones and there's only so much an engineer can do to mitigate the dreadful audio quality of cell phones. Quite often, when you're listening and thinking "turn it up!" it's really because noise, excessive ambiance and restricted frequency response are interfering with intelligibility. Compression can't help with that.