Yesterday I tinkered with increased bit-rate streaming, stepping up from 24kbps all the way to 40kbps, following Artisan Radio’s report that there is a detectable improvement even compared to 32kbps, which I also tried.
The stream sounded better.
Using Hans van Zutphen’s Stereo Tool for processing, I set the bandpass filter to a range of 40 to 20,000 Hz for the following reasons:
Using an article by Dave Moulten from TV Technology titled Thinking About Equalization: The Audio, I set the low end at 40 Hz because Mr. Moulten reports that the 1st octave of the audio spectrum, 20 to 40 Hz, contains little musical content and is not reproduced by most loudspeakers, so I turned it off;
Next the Nyquist Theorem applied (is that right), in which actual frequency response is 50% of the bandwidth. Therefore, at 40kbps the 50% frequency would be 20kHz. True?
The next day I got to thinking about the AM transmission, with bandwidth typically limited to the carrier width, residing on a 10kHz channel. To observe the modulation result using the upper audio band to 20kHz, I tuned to the channel adjacent to 1680kHz, namely 1670kHz, and noted active splash on sibilant sounds and piccolo notes.
SO I set the Stereo Tool bandwidth to a top limit of 5kHz and the modulation spillover at 1670kHz disappeared.
Always the quandry, that means either the stream would best match the situation by tuning down to, say, 16kbs, or…. I guess an outboard component bandpass filter for the transmitter would permit optimizing both types of transmission….
Join us again tomorrow.
Relaxing The Rules,
Back in 1972 I obtained a 3rd Class FCC license. Basically it was standard for anyone planning to work at a radio station to have one.
Studying the rules and such for the test, it was noted that with a 10 kHz AM channel width the highest audio frequency to be transmitted was 5 kHz. That was the rule. As you stated, double the modulating frequency is the bandwidth of the transmitted signal. Keeping the audio to 5 kHz would minimize adjacent channel interference.
Now days, it seems common place for AM’s to run the audio frequency wide open. I don’t work in commercial broadcast so I don’t know how the FCC feels about that. Now perhaps for stations going digital, they’ll have to reel that in a bit as the digital info occupies those upper sidebands.
A Survey
On the St. Louis AM dial there are four stations running HD digital hash, and the others were part of my dial twisting survey in which I checked their compliance with staying “on channel” with their audio bandwidth.
Three stations in particular had out-of-channel splash over, with one of those being extremely severe.
All the rest of the stations had zero out-of-channel splash.
Now I have been playing with going slightly over, with the following bandwidths:
7.5kHz, 10kHz, 12.5kHz and 15 kHz.
The shortwave station is an even more interesting case, since station separations are a mere 5kHz.
To be absolutely in control of all the different audio bandwidths, we need four distinct low-pass equalizers: for streaming, AM, FM and shortwave.
FCC Emission Limitations
I think the answer is here in 73.44 AM transmission system emission limitations:
http://www.hallikainen.org/FCC/FccRules/2010/73/44/section.pdf
Seems Clear..
Yep, for a licensed AM anything above 5 kHz better be cut.
Table Radios
The portable Grundig FR-200, which I carry all the time like it was Tinky Winkie’s “bag,” does a better job of locking to the AM signals now that I have the bandwidth limited to 5kHz. There is a built-in AFC which cannot be turned off, and with my previous wide-band audio signal the AFC would slip off channel and sometimes get the splash-side band signal and I’d need to retune.
All the radios in the house seem to tune better to the AM signals with proper audio bandwidth.
I have turned attention to the 5kHz channel width at 13.560, and there’s still plenty of “splash” going on over there, but I will need another audio band pass filter to solve that.
The FCC rules for International Broadcast are 73.701 to 73.788 but there’s no rule for AM modulation, although obviously many U.S. SW stations are AM….
The types of modulation discussed in the rules are
1) DBS – double-side-band
2) SSB – single side-band
3) Digital modulation.
NRSC
All licensed AM stations in the US MUST adhere to the NRSC curve..no exceptions..mono or C-QUAM…and a 5khz roll off sounds like the audio went through a meat grinder in C-QUAM.
I run my AM C-QUAM with a 40Hz to 15Khz curve, no splatter, no spitting, no sputtering. And I do not run modulation out to 125 percent either. No need to blast everyone’s ear drums or make everything sound so loud it takes away the natural sound of how the recording engineers and producers of music WANTED the material to be heard.
There is something about hearing an orchestra piece as it should be with the nice low ambient points and the louder points. Natural sounding, not blaring and jumping out at you.
Louder is NOT always better..even for AM.
RFB
Wasted Space…
Amazing! Let’s see, AM BCB channels are 10 kHz wide.
You have a 30 kHz wide AM signal that doesn’t interfere with adjacent channels?
Too bad all those cheap AM receivers can’t hear it. Not to mention the middle aged audience listening with a hearing range of about 5 to 8 kHz.
HF Modulation
The rules for international shortwave do not specify an AM emissions rule, but the DSB must be the next closest thing.
FCC Rules Part 73 Subpart F – International Broadcast Stations
73.701 to 73.788
73.756 System specifications for double-sideband (DSB) modulated emissions in the HF broadcasting service
(a) Channel spacing. The nominal spacing for DSB shall be 10 kHz. However, the interleaved channels with a separation of 5 kHz may be used in accordance with the relative protection criteria, provided that the interleaved emission is not to the same geographical area as either of the emissions between which it is interleaved.
(b) Emission characteristics—(1) Nominal carrier frequencies. Nominal carrier frequencies shall be integral multiples of 5 kHz.
(2) Audio-frequency band. The upper limit of the audio-frequency band (at— 3 dB) of the transmitter shall not exceed 4.5 kHz and the lower limit shall be 150 Hz, with lower frequencies attenuated at a slope of 6 dB per octave.
(3) Modulation processing. If audio-fre-quency signal processing is used, the dynamic range of the modulating signal shall be not less than 20 dB.
(4) Necessary bandwidth. The necessary bandwidth shall not exceed 9 kHz.
[70 FR 46677, Aug. 10, 2005]
73.757 System specifications for single-sideband (SBS) modulated emissions in the HF broadcasting service
73.758 System specifications for digitally modulated emissions in the HF broadcasting service
QUESTION: WHY is the low-end of the audio band rolled off below 150Hz?
For Now
The audio response curve being set for KDX in Stereo Tool Winamp Plugin, is now set at 150Hz to 4.5kHz, in accordance with the specification for the international shortwave station in the U.S. Therefore, the transmission at 13.560mHz is compliant.
However, the AM, FM and streaming signals are also using this limited frequency response by default, since we only have one bandpass filter.
The solution will be to obtain several distinct band pass filter “boxes” for each of the transmission mediums.
One attractive feature of this “tight” audio bandwidth is a sharp focus for small loudspeakers in noisy environments: the voice “jumps” out from the speaker.
The ideal world, described by RFB as hearing an orchestra in its true dynamic range, is not always the world we have. There tend to be all kinds of noises in the surrounding world, and the radio station sometimes needs to enhance itself with compression and equalization to be heard above the din.
Bottom line: there is no “one size fits all” solution to broadcast audio quality.
Yeppers
“Amazing! Let’s see, AM BCB channels are 10 kHz wide.
You have a 30 kHz wide AM signal that doesn’t interfere with adjacent channels?”
It’s amazing indeed when everyone has been programed AND forced to hear such narrow bandwidth at 125 percent + modulation when the format has been around since the hey day and no filtering or over driven modulation used back then…worked just fine…as it does now..when not over driven with loudness and brick wall filtering. You can measure 10 times more belch and splatter all over the frigging dial off KTWO AM 1030’s signal than you will EVER measure off mine.
And KTWO does not run HD digital IBOC..err iBelch junk.
“Too bad all those cheap AM receivers can’t hear it.”
Neat eh! Maybe those cheap radios are good for something after all! 😀
Neither does my Sony ST-JX220A C-QUAM receiver.
And that was no cheap radio.
“Not to mention the middle aged audience listening with a hearing range of about 5 to 8 kHz.”
Especially after pounding their ear drums with booming bass sub woofers pumping 120+db sound pressure inside the space of a small walk in closet….with the baby in the back all tucked in and secured to the car seat right next to or directly in front of that thumping trunk lid rattling noise….ya that hearing range would be somewhat diminished eh! LOL!!
OR from decades of listening to popping static and clicks and buzzes and spiking co-channel whines and whistles at levels that would equal the same as what is experienced by the baby from the sub woofer through the air…cept in this latter case…via close close range headphones…cuz the xyl is tired of listening to that noise after all these years!
LOL!!
I actually had a fellow engineer friend of mine who goes around the country contracting to AM stations to do their AM splatter measurements for compliance. He measured my signal’s splatter and it was considerably lower than KTWO’s. We measured NOTHING at all at 30khz from center frequency.
NOTHING!!
The FCC agent was also impressed.
RFB