[BC] Absolute Polarity (was XLR pinout debate)

[Robert Orban]

At 02:40 PM 8/13/2006, you wrote:
>From: "Mark Humphrey" <mark3xy at gmail.com>
>Subject: [BC] Absolute Polarity (was XLR pinout debate)
>To: "Broadcasters' Mailing List" <broadcast at radiolists.net>
>Message-ID:
>         <74b029b80608130643w1020fcdfpa606401c8ee52f4a at mail.gmail.com>
>Content-Type: text/plain; charset=ISO-8859-1; format=flowed
>
>On 8/12/06, Cowboy <curt at spam-o-matic.net> wrote:
>
> >  Oh, sure, we all know ( or we all should know ) that positive pressure 
> on the
> >  microphone diaphragm is supposed to produce positive going voltage at the
> >  output pin labled + but how do we know what the polarity was on the
> >  record ( that's CD for you kids ) when it was recorded ?
>
>This is a very good point -- and for that matter, how many receivers
>maintain "absolute polarity" from the detector through to the speaker?
>  Unless the entire system adheres to a standard protocol, I guess we
>always end up with a 50/50 chance of getting it right regardless of
>the way the connectors are wired.

At Orban, we have been aware of this for a long time and design our 
equipment to be non-inverting, input to output. Of course, this only 
applies to products using linear-phase crossovers; allpass crossovers will 
change polarity several places in the frequency vs. phase curve.



>While driving around, listening to my JVC "HD Radio" receiver
>transition from analog to digital, I've noticed the audio on some
>stations nearly cancels out during the overlap (apparently due to a
>polarity inversion), while on others, blending is hardly noticable.
>It's easy to understand how this could happen, especially if the
>station uses separate processing chains for the two components.  Some
>processor manufacturers (such as Omnia) have provided an AES polarity
>inversion switch, which makes it easy to fix  - as long as the
>monitoring system is accurate.

We added this feature to the 8500 last year per customer request.


>So this raises a few more questions:
>
>1)  Does iBiquity require receiver manfacturers to adhere to an
>absolute polarity standard on both the digital and analog sections?

Yes.


>2)  Is the HD codec phase linear?

I don't know for a fact, but it is very probable because phase-linear 
filter banks are also computationally efficient.


>3)  While we're on this subject, what about other lossy codec schemes,
>like MP3, AAC, etc.? Have any absolute polarity standards been set for
>them, or does anyone even care?

Offhand, I don't know of any codec that is not phase-linear (i.e., does not 
have complementary filter banks in the TX and RX sides). That being said, I 
have not investigated very low bite rate codecs for speech.

Bob Orban