[BC] Seedy Quality in IBUZ Ads

[Robert Orban]

At 09:43 AM 8/26/2006, you wrote:
>From: "Richard Fry" <rfry at adams.net>
>Subject: [BC] Seedy Quality in IBUZ Ads
>To: "Broadcast List" <broadcast at radiolists.net>
>Message-ID: <00a501c6c902$c7305970$6a65e242 at Insp1000>
>Content-Type: text/plain; format=flowed; charset="iso-8859-1";
>         reply-type=original
>
>Bob Orban:
> >However, I have to add that my brief audition of the
> >Polk radio made the crossfade between FM and
> >properly processed HD very noticeable -- when the
> >HD channel turned on, HF distortion decreased, the
> >highs opened up, and transient definition and impact
> >increased noticeably.
>_________________
>
>That is an interesting observation, Bob.  An analog FM exciter using a
>digital modulator and digital stereo generator has almost ruler-flat,
>demodulated frequency response to beyond 15 kHz, and harmonic, IM and TIM
>distortion levels so low as to be difficult to measure accurately.  OTOH,
>HD has such a high content of digital compression artifacts that
>conventional audio distortion measurements of it give very bad numbers.
>
>What do think accounts for what you heard?

This has been discussed at *great* lengths on this list in the past.

A properly operating analog FM exciter is not ruler-flat at all. It is up 
17 dB at 15 kHz due to pre-emphasis (or 13 dB in 50us countries). The 
receiver restores an ostensibly flat response, but, in fact, this 
pre-emphasis / de-emphasis process severely limits HF headroom in the 
transmission chain. 75us was great for 1948 recordings, but is a major 
constraint with day's recording practices.

The two advantages that the HD channel has over the analog FM channel are 
no preemphasis and 5 dB higher gain at the radio, so the digital channel 
requires at least 5 dB less peak limiting than the analog channel to 
achieve loudness parity when the radio crossfades between analog and 
digital. When you consider that the analog channel pre-emphasis limiting, 
the total difference in the amount of peak limiting between the analog and 
digital channels might be up to 15 dB. In the processing world, peak 
limiting falls in the "necessary evil" category. It never makes the audio 
sound better, only louder.

This of course assumes that the digital channel is being processed 
optimally. This is not hard to do is one is paying attention.

The reason that the digital channel sounds better is simple -- it requires 
dramatically less peak limiting than the analog channel. Moreover, in areas 
with lots of terrain and/or high buildings, you get even more of a 
difference thanks to the digital channel's multipath immunity,

Bob Orban