Modulation asymmetry for C-Quam AM Stereo

On my AM Stereo Forum one subscriber asked about using software audio processing on a Part 15 AM Stereo transmitter, and I’d like to recommend Breakaway Broadcast to him, but first I just want to make sure it can be set to optimally conform to the C-Quam specs.

At least according to CRL, the official Motorola spec is that single channel (L only or R only) negative modulation must be limited to -70%, in both the L+R and L-R domains. And since asymmetry is not applicable to the L-R subchannel, that means the modulation domain for single channel audio becomes -70%/+125% L+R and -70%/+70% L-R.

CRL does not mention any corresponding restriction on positive L+R modulation caused by a single channel, so that’s why theoretically -70%/+125% L+R modulation is possible during single channel audio, although I believe exceeding +100% single channel L+R modulation will cause crosstalk distortion into the opposite channel (sometimes audible as a slight crackling noise), so some purists like to impose a +100% L+R limit on single channel audio as well.

Note that CRL has always done single channel limiting in the L/R domain, after all of the matrixed limiting, clipping, NRSC filtering, and tilt correction has already been done. In fact, setting a CRL processor to output the audio as L+R/L-R instead of L/R will disable the single channel limiter, in which case you need to enable the single channel limiter built into your C-Quam exciter (and usually it’s not the best choice to have an exciter or transmitter do any kind of audio processing by itself!).

Orban took a different approach to this issue with their Optimod 9100 by adaptively narrowing the stereo separation during single channel audio so that the modulation limits are met, but this causes a rather annoying contracting and expanding of the stereo sound-stage during songs with extended single channel audio, such as "Year of the Cat" and "Here Comes the Sun." (Their current AM Stereo processor, the 9400, does not employ matrix limiting at all, and thus restricts all single channel modulation to -50%/+50%, causing a loss of loudness for mono listeners — the main thing that matrix processing was invented to prevent!)

So with all that said, I just wanted to make sure of what kind of single channel modulation limits Breakaway employs in its AM matrix processing, as the web site’s description of "70% negative, 105% positive (for C-QUAM stereo)" is rather vague, since it does not mention whether this applies only to single channel modulation (as the Motorola spec dictates) or to all audio regardless of single channel content (which would be far too conservative).