My hunch is that there wouldn’t be much difference with HD on or off because it’s the limited audio headroom in the *analog* side of the radio that is mostly at fault for the distortion problem heard on these radios. How excessive bandwidth caused by high analog modulation might affect the performance of HD or how the presence of HD in combination with high analog modulation affects the analog audio performance is something that I haven’t measured yet. But it is intriguing…

Whenever I’ve encountered a ‘where is it coming from’ distortion issue in the field the analog modulation has always been *well* north of 120%. So far, when the station is also running HD it’s only been the analog side that sounds distorted (most engineers know they dare not hammer away 0dBFS on the HD side!).

The most telling clue is that when modulation is backed down (for instance from 140% to 120%) without touching anything related to the processing the station suddenly sounds much better on radios that were previously distorted. Makes you go hmmmm……

Back in the day I was guilty of sneaking another 5% modulation (115% instead of 110%) if I could get away with it, and radios of the day didn’t seem to care. In the past few years however I’ve been to stations that are boldly modulating 130% to 140% and I think this is when and where the limitations of finite receiver headroom finally rears its ugly head.

Jeff

Thanks for writing! As you point out, most listeners aren’t using the kind of radios that we ‘radio’ people would use, which is why I enlisted the help of the station’s staff in loaning me a bunch of their ‘regular’ radios to see how those behaved.

Two of the 27 radios I tested were car radios – the one in my Mazda RX-7 and the one in the station’s engineering vehicle (our GM’s former vehicle), a fully loaded Eddie Bauer edition Ford Explorer. The vehicles were measured while parked in my garage at home, and both had a fairly easy way to grab pre-volume control line level output.

How the distortion sounded was a little hard to describe and it did vary by radio. As an audio guy I would describe it as a healthy mixture of harmonic and intermodulation distortion. It sure made an otherwise nicely processed radio station sound ‘broken’ on many of the radios. If you think about it, the nonlinear behavior measured in these radios at high modulation levels is just like adding a soft clipper (driven fairly hard!) on top of of the station’s on-air processing. This might also be why the apparent loudness of a station varied so much between the various radios (the ones that sounded louder were also the ones that sounded more distorted).

From scoping around inside a couple of the radios it appeared to me that they were mostly running out of audio headroom, not IF bandwidth, though that could be true to some degree too. In most cases it looked as though it was usually the radio’s stereo demodulator chip that was being overdriven, and since the distortion was being created ‘pre-volume control’ there is no way for a listener to turn the distortion down.

Jeff

Thanks!