HFU HF Underground
Technical Topics => Equipment => Topic started by: Kage on May 23, 2019, 2242 UTC
-
Found this (https://www.hfunderground.com/board/index.php/topic,21866.msg78748.html#msg78748) topic from 2015 on bandwidth but few replies and I never hear people discuss emphasis usage.
With mediumwave broadcasting it's almost necessary as listeners radios will have 75 microsecond de-emphasis built in, so the old standard NRSC mask 10db up near 10kHz with the sharp cut gets the job done like the big stations. Most audio processors will take care of this so no reason not to use it and I've noticed when not used the broadcast sounds muddy on AM radios able to pass the highs.
On shortwave I'm guessing there is no real need for it since standard audio bandwidth is out to 5kHz or so. What about some of you running higher bandwidth? I'd imagine most of the standard multi-band radios use de-emphasis on all the AM bands including shortwave since it's easier in circuit, and leave the FM broadcast section to do its own thing. Maybe some pre-emphasis on the shortwave transmission would help? Boosting up to 7.5dB at 5kHz and then brick walling it at that point to accommodate peoples average analog receivers using the mask on all AM bands internally?
Also wondering if anyone here has tried any practical bandwidth tests to see if lower/higher on both transmit/receive being equal made a difference in long distance listening?
I'd think that technically speaking lower bandwidths would conserve transmit power thus increase broadcast range slightly but the difference between 5kHz and 10kHz audio bandwidth is so small compared to something like FM upwards of 200kHz to AMs 10-20kHz occupied space that it probably doesn't matter so why not use higher fidelity audio on AM to make everyone happy from cheap radios to SDRs set wide open? We're talking pirating anyways, so no need to follow strict bandwidth rules unless the pirate band gets congested.
Just some caffeinated thoughts ;D
Edit: Discussion of SSB bandwidth welcome too! I'm so used to AM mode that I forgot this applies to SSB as well.
-
The NRSC pre-emphasis curve was implemented to stem the rising interference from stations continuing to crank up the high end in compensation for narrowing receiver bandwidths. In practice, most receivers are so narrow that no NRSC de emphasis is needed, the IF truncation does it for you. ITU rules that govern international broadcast service never had any such provision. The only specification was maximum 4.5 KHz audio without pre-emphasis. There is also some arcane rule in there as well about a minimum of 20dB dynamic range, perhaps to prevent a loudness war of sorts on HF.
Currently, my transmitter is set for 12.5 KHz audio bandwidth using emissions designator 25K0A3E. When the stereo exciter is back online, we'll be 25K0A8E. I run pre-emphasis to combat atmospheric noise, and like medium wave, to restore a sense of high end on narrow radios. The majority of the recording I get run around 3-5 KHz. Kilokat7, Mix and a few others will send me wide 15KHz recordings from time to time, largely dependent on propagation. Most of the areas we play in has a fair amount of elbow room, and bandwidth is not a problem. I like to give people the option to listen in higher quality should they choose to do so, and I monitor off air in full bandwidth as well. As the proliferation of SDR's becomes more widespread, the old receive bandwidth limitations are beginning to erode.
If you have the power (250W carrier and above) you might as well run higher bandwidth, as there is a better chance that someone will hear it. Low power guys probably are better off truncating the audio to 60-4500Hz to put the majority of the modulation power where it is likely to be heard. We are kind of the odd man out, and most stations who wish to run wide are more likely to run 10KHz as that is the AM standard and more likely to be obtainable with second hand gear. I would recommend running NRSC pre-emphasis as it does help the received audio quality with matching receiver de-emphasis.
+-RH
-
For SW broadcast I used to use a wider bandwidth, usually 8 KHz, High pass 100 Hz and 75 pre-emph.
Some tests proved that a narrow bw of 5 gave a much louder sounding signal and it broke through the crud so I guess distant reception was better. Checking my signal on the online SDRs, the narrow signal still sounded great.
Occasionally I run 80 Hz to either 8 or 10 Khz but that's more for my own pleasure, listening at home on a wide bw desktop radio.
I don't really believe that going for Hi-Fi on SW is that beneficial as most people are listening on comm receivers or SDRs anyway.
EDIT
I should mention that I run a digital AM Optimod with the added HF settings that allow me to modify my audio to almost anything I like.
-
I'm running the full CRL AM System 3 Processing chain using the AM NRSC standard pre emphasis I have the ability to adj the cutoff point and i'm set for 12khz I also use an SRS processor before the CRL chain I think that it's one of the best sounding signals on the air
tAC
-
Roughly 80 hz. to 6.5-7 khz. with some dicking around with the mid-bass and treble between 2.5-6 khz. for music. Moderate compression and limiting, especially on the i.d.'s which got a hard bass cut as I've got a deep voice, running a mid-boost between 200 hz. and 3 khz. for those. (They've got to be able to understand what you're saying to i.d. the station.) Standard pre-emph.
It's nice to see this generation of pirates sharing their audio secrets. A lot of us old farts wouldn't give away our hard learned secrets at gunpoint.
AM only, of course. I've never dealt with squawking sounds of SSB. Fearless Fred wouldn't approve.
-
I'm so cheap that I use a 20 band EQ. In the 1990's, that was used in FM to keep audio FREQ's above 15kHz at bay from messing with the 19kHz pilot carrier in an FM Stereo TX. Today? A 20 band EQ is still used to do a couple of things, (1): To give AM MW audio a boast in bass, MID, and semi-high, FREQ's, (2): To act as my ye poor man's De-esser to run various Podcasts that don't use De-essing when creating Podcasts. On a wide band RX, those S's and T's sound pretty harsh from these Podcasts. Yeah, I could use the EQ in Audacity to run those through and De-ess them that way, and I have to some, but time-wise, easier to do that at the TX audio chain. A sharp cut around 9, or 10kHz, usually does the trick, while a boast from 4 - 8kHz still adds to some nice brightness. One other advantage to this? No dedicated computer for this, but old analog tech. I do computer aided, but not computer dependent.
-
Yo, don't mix yo fricatives wit yo plosives, yo, uh, yeh yeh boiiiiii.
-
Awesome replies so far. My curiosity was peaked on this topic because of the analog audio processor I am designing and picking filter frequencies. Currently have an 8 pole overshoot corrected NRSC-1 compliant filter designed for the final stage which is flat that sharply rolls off around 9.2kHz and is down 15db at 10kHz, 25db at 11kHz and so on, but was thinking of adding a toggle to flip between that and something else which is part of why I asked.
Can't decide if I should add a 5kHz or maybe wideband 12kHz+ filter as the option. Guessing the earlier would be more useful for shortwave broadcasting but my ears say otherwise :-\
I guess as the old saying goes.. if they can't hear it with their radio why broadcast it? Going out past 5kHz is probably fruitless but for now I will stick to the mediumwave broadcast standard that is the NRSC mask and after looking at the replies will probably add the 5kHz filter as an option for restricted bandwidth 8)
If anyone is curious about the audio processor project I am working on you can find it here (http://darkliferadio.proboards.com/thread/1029/diy-audio-broadcast-processor-scratch?page=1&scrollTo=4587). It's slowly coming along and is a blog style post so always updating it. Only recently got the overshoot corrected filtering soldered up. Hopefully when it's all designed the info will be useful to other pirates that are electronics hobbyists too that want to get their feet wet in audio processing. In the meantime I will use Stereo Tool which is simply the best option out there on the digital side, I just prefer real-time analog.
-
Interesting observation on that 5kHz. cut. relative to radios being used to receive the signal. I knew an old jock from a large market AM station in the "Boss Radio" era. Band crowding was getting out of hand due to the wide open signals preferred then. He said holding it at 5kHz kept wide them from getting overrun by the splatter on either side and stand out at the same time. He said he was knocked out listening when the engineer made the switch at the way the signal stood out on the band plus the quality and clarity of the sound.
BTW, car and portable radios of the kind people listened to then tended to sound like crap. The goal was to get a signal out to them and make it as loud as possible.
-
Consider your likely listenership. Is it someone far aways tuning via a sw set with a fixed 6kHz bw, or a nerd with an sdr rig that can stretch out to 20kHz or more with synchronous detection? For US pirates, my guess is much more the latter than the former. In that case, open it up so the sdr nerds can enjoy hifi, but have the power to back it up. If not, 6kHz it is!
-
From what I've seen the SDR is replacing the traditional receiver so the home listener has a chance to hear a stations broadcast regardless of his or her location. so there is always a chance of an arm chair copy somewhere. regardless of brute force "carrier power" I only know of a few ops who can do AM 250W + carrier but if processed right even the 20w guys can sound like a powerhouse station somewhere on sdr so why not go the High Fidelity route. If your sound can't keep the audience tuned regardless of format then why bother. Spoken like a true FM guy..
tAC
-
I ran a short test last year with a friend who was listening on an online SDR whilst on the phone to me. The SDR was 100's of miles from my TX but with a good signal.
As soon as I increased the bandwidth to 10 KHz, he told me to put it back to 5 (actually I think it was 4.5) as he reckoned it sounded better. I think the HF setting I used also has a boost at the higher end.
I didn't ask him what he was listening on though. Probably laptop speakers or a phone.
On MW, I've always preferred the max bandwidth I could get. The max I could get from an analogue Optimod was 12 KHz which sounded great.