Ham and audio tip for radio users

Most modern radios include some kind of speech processing to level and/or act as a audio limiter/leveller/agc to boost the average level of mic audio.

Better units have some parametric EQ based adjustable setup, and a compander unit for a better tonal and levelling AF adjustment including compression etc.

For reference, Parametric eq is a bit like a multi band eq/graphic eq, but where the common eq adjusts a whole band of frequencies equally per af segmented band, a parametric let’s you tweak the width of each AF sub band, how much post-filter makeup gain is employed, but also allows you to adjust the low pass and high pass curves creating a combo per sub-band of very tight high emphasis or lower emphasis wide band boost / attenuation.

Now you could spend a fair chunk of cash to improve using an external SP on radios lacking it or to use on an internal SP bypass.

But you could also mod a base mic and utilise an cheaper option made from a guitar/instrument compression pedal.

When I need that option, where using my Tascam mixer as processor and precise pre-process box control is impractical, I use an old Robotech modelling pedal of which had a lot of digital modes and adjusts and presets, so I keep broadcast and comms fine tuned custom presets designed to give preemphasis under lots of different environmental and propagation conditions, with SSB specific examples, FM and AM examples, DV specifics for each of C4FM and DMR. End result is you get the NB equiv of broadcast sound like spoken audio which can really cut through garbage on poor SNR analogue situations.

Cost, free as I had a spare Robotech unit, but the earlier examples (identifiable by have no uploading USB port) can be got for a fraction of external SP units and often walks over some very expensive items.

So, given it can be wired into a mic, it could make a good mod that doesn’t wreck a radios TA if the mic can be unplugged.

Just remember to adjust sanely, and the noise gate can make a pretty good mute between spoken sound - some units actually have noise cancellers in to compensate for single coil pickup noise.

If you’ve got something like that lying around, try recording and tweaking on a PC or Mac or Linux box - even if you don’t make a long-term fitting inline with a rig, the computer tests can tell you a lot about how people are going to potentially hear you and if you suffer a quiet voice, anything sane helps.

But be sane and remember, don’t be tempted to transplant into a radio that requires TA certification for legal use.

Continuing covering suitable ‘fx’ which can help with your output tonality and clarity -

a) Compressor - in principle, yes, but use with care. As a DRC method (Dynamic Range Control) style method, it works well when correctly adjusted in it’s attack, decay and sustained and level of compression, and makeup gain. A blended us of dynamic audio compression on say your actual voice frequency bands which have less emphasis to the degree that your audio out sounds weak and simple Parametric EQ to focus emphasis on your main voice characteristic frequencies is a good combo to start with. Potentially, with good mid-range emphasis and as full a natural reproduction of your particular ‘voice’ frequencies range within the accepted telephony range is your goal. It’s ok to sound ‘talk radio’ broadcast punchy where it benefits, within reason, but remember your QSO partner(s) are listening to a reproduction akin to the AM tonal response of a classic Roberts AM receiver speaker response profile. So if you can get your audio to reproduce well on that kind of speaker amp equivalence, you’re in the ball park comfortably.

b) Reverb - If mention i mention, reverb fx, you start having CB flashback nightmares about demented Echo mic users, that’s forgivable - no, i’m talking about positive use of reverb.

You can take an echo/reverb fx, reduce the repeat rate to zero, and adjust the rest of the settings to create a simple ‘double voice’ effect that when balanced wet to dry properly, gives you vocal singer’s recorded ‘airy feel’ to your audio which can make it a bit easier to listen to. It definately adds some emphasis and when set with a suitably short ‘tail’ (‘tail’ being the time the single reverb effect hangs on after you stop talking or take quiet breath pauses). So adjust it consideration of the main audio limits (you will go back and forth many times getting it spot on, comparing it’s added effect to what you’ve already done).

c) Parametric EQ - this is the big booted leader of your SP pack of effects. Get this side spot on, use the other two items for emphasis and punch.

A good Parametric has multiple bands (at least two, three is what i call a real minimum). Each band has, like with a graphic EQ, an associated freq range - but unlike a Graphic EQ’s basic boost/attenuate continous adjust setting, you’ll have a more moderate gain to attenuation range per band but you also have a band window adjustment (adjustable BPF) where you can move the center frequency of the band range. So by listening to your output, you can (per band) tweak this per-band window so the rolloff favours the more naturally emphasis of your actual input (mic) within the design response of the mic’s driver. Adjust this for each band (simple good ParaEQ’s have a basic Bass, Mid, and HF adjusts with a wider window adjust). The third major control alters the actual width of the window, so even a three band ParaEQ can make a useful tonal make up on a mic that maybe isn’t so great with your voice. It can, with sane adjustment, make you sound less over bright on an electret insert and even give the tone a more dynamic insert tone.

Moving onto more ‘modelling amp’ type useful FX, we start with -

Single/Humbacker modelling - designed to make a Humbacker type pickup sound more like a single coil pickup and the other makes single pickup guitar sound (example, you have two or three single pickups on most non-bass Fender guitars) sound more like the fuller less noisy Humbacker pickup tone. This, used with caution and care, can be a subtle but effective way of rounding out how full your mic audio sounds especially where it’s more where a non-reverb semi-doubling/harmonised blend is all you need to add emphasis to your spoken audio.

Gates - these allow you to set a pre fx lower threshold where if your input audio drops below the adjustment threshold, no audio passes into the effect chain. Some amp/guitar modelling ‘pedals’ or stacks give a you a post fx gate, so you can adjust to ensure only audio within a higher or lower threshold ever leaves the ‘device’, so combined with a low mic gain on the radio, means there’s a margin where the SNR at the mic amp isn’t into the noise floor, but the amp’s internal noise isn’t notably adding badly to the overall resulting received SNR.

But you must ensure your quieter audio, spoken audio, sits far enough above the noise floor and still have (uncompressed) a natural softness of softly spoken audio. Likewise, you need to ensure the peak–to-peak range sits comfortably under 0dB, with room to spare (headroom) to which i suggest a minimum of -3dB below digital peak (0dB) or better still, -6dB relative.

In digital voice terms, you are aiming for a) getting the most effective clear non-metallic and not ringing and non-clipping demodulated audio as heard. Metallic sounding audio is the DV equivalent to dropping below the threshold, where if you could see the received SQ, you are moving all over the place SQ wise, of stable SQ. Again, as with FM (and notably given DV on VHF/UHF is an fsk/fm hybrid - which depends on the actual DV mode), you only need to be comfortably avoid a threshold and steady in SQ to be the DV equivalent of ‘fully quietened’. Any substantial over-level on sig above that gives rapidly decreasing gains in non-existent added R in RST.

So a good jumping in point is to optimise your settings/setup (AF and RF) for analogue FM over a moderate distance where you are still ‘above threshold’ and not getting lost in the noise. Then alter it slightly, but carefully, to get a similar level of balance in DV but with an emphasis on being free of ‘ringing’ and ‘metallic’ in the resolved audio. Parrot/Echo test on a repeater or hotspot is good for this, hotspot for general eval and a nearby repeater, preferably one you’re borderline accesible to, and repeat your DV echo test. If you get mostly a ring and metallic free echo back, you’ve in the threshold - there’s really no reason, unless you’re on the fringes of a hotspot coverage, why you would sanely ever get an Echo back that’s got a ringing or metallic or both tone.

You’re definately trying to avoid Radio Luxemborg ‘fady fady’ AM effect where your audio is sounding like’s jumping between a bad day on 80m and at the other extreme, sounding like CB’er using max rig mic gain and the full kick of their power mic. After all, ou’re setting up speech processing, not trying to hit the crowd at the back of an arena on a PA system.

At the end of the day, you want your audio level as heard modulated, to be above the noise threshold and below clipping/distortion etc. So that could mean a set of adjustments where your quiet audio (even at extremes of distance and low sig quality) is heard at least a few dB above ‘the noise’ (in FM terms) and inside of threshold without being overdriven. Remember, with FM, you’re gains in ‘punch and clarity’ become very much a reducing loss of gains once you are comfortably ‘above the threshold’ to be ‘fully quietened’. In fact, a good report with tell you if you are hitting the ‘fully quietened’ level in FM use, as a lot of the other traditional descriptions are CW specfic (the T in RST), so a high R in RST is where you are fully quietened FM noise wise and clear to a natual conversation level. Likewise, you don’t need a needle banging S meter level (as heard) on FM usage - if your sig as heard remains steady (S meter reading) enough to keep you in the fully quietened range, your ‘S’ in RST report is good.

I aim, typically, to be able to leave the radio’s mic gain at unity and use makeup gain on my external kit and set up my externals setup to keep audio fed to mic socket comfortably within the correct mV range that prevents the FM limiter (analogue) from tripping and likewise creates pure digital silence level in where i’m not talking, natural pause and cautionary pre-response initial pause and is comfortably short of clipping to the point that if i knock the mic over or drop it, the resulting thump gets muted by the external audio setup’s limiters.

If you’re not using a base mic, on your indoor setup, you could always try any of the above ‘SP’ type methods, employing each to a different mic - say the equiv of a common Shure vocal/speech mic and a semi or highly directional electret. Adjust each, in how the level inputs to the chain and make adjustments to each version, find a balance if you have only one set of stuff to work with and use a two-channel mono mixer with attenuated master output (or line level if your audio kit can accept it), but ultimately, no matter what, you want to be feeding lowish mV level range into the rig’s mic input.

That’s partly why i utilise guitar/instrument stuff for ParaEq, compression etc, as the passed audio to each unit or module is kept within a HI-Z input range which, if the output was designed to go into a Hi-Z input of an instrument AMP or PA, is pretty ■■■■ close to what you want to feed a mic input on your rig. If it’s slightly too high, crack back the o/p level on the final stage of the external audio chain a touch and you’ll be comfortable.

It’s how, on base operations, i get a punchy but clear near-broadcast sounding spoken audio received, by taking audio mastering techniques used in pro audio and broadcasting and applying them within the realms of affordable audio kit. And given how modelling kit is rapidly making single effect guitar pedals unattractive and unnecessarily a liability for leisure musicians, you’ll find there’s a lot of old-school guitar pedals out there to exploit and modify to build your SP chain with. I happened to have a spare Digitech and a PC/Midi controllable more complex unit left over (i favour the Digitech for instrument use, hence owning a few duplicates) so they tailored my choices.

However, if you want a more focused example ‘off the shelf’, there is a multi-effect modelling unit by Digitech that’s specifically tailored to Vocal/Speech use, but typically (where you see them listed), they are boarding on low-mid priced ParaEQ SP’s sold to the ham radio market, so you’d see value in being able to repurpose an instrument version instead - but i would avoid a Bass focused pedal use, as it’ll be good at lower AF, but very limited in effect on the MF/HF unless it’s tailored for jazz bass use where the strung range of JG’s is a crossover between mid/high P Bass tunings and lower EG tunings.

Assuming that’s all way to much effort, there is an old trick that’s dirt cheap - take your mic level into a classic tape recorder with a good AGC auto record level. Then attenuated the DIN line out or the earphone socket level to mic input level range. If it passes a record monitor level, just defeat the write-protect lever and use the monitor’s amplified range for a crude pre-amp. That’s how i used to get around record level issues on cassette based computer storage systems, using a crossover between the in and out of the computers using the aux/Din and main sockets of a cassette deck created an effective preamp that made cloning code between computers easy and painless.

Remember, there’s little in radio that’s not solvable by ingenuity and a bit of lateral thinking - just remember to draw a line between sanity and TA restrictions.

There are very few people who know how to correctly adjust “audio and audio quality”. It would be wise of those people to set things at “standard” and then LEAVE IT ALONE! There are definite frequency range standards for amplitude modulated modes (which does include SSB). Most “audiophiles” exceed those standards and I’d be willing to bet that most everyone has heard them on the bands.

Well, it’s their responsibility to ensure, even just by use of onboard user level adjustment of mic amp gain and correct usage of an onboard ‘enhancement’ SP adjustment, to keep their set within operational conditions applicable to their license and not cause undue over deviation and other ‘over mod’ issues applicable to whichever mode is used.

It’s a ham operator’s personal responsibility as primary station operator and licence holder to be able to main, resolve and correct ANY issues they come across with due competency and if it’s beyond them, to seek knowledgeably sound assistance.

None of the above postings are in anyway irresponsible, or damagingly bringing any disrepute to the world of radio communications as it was for tips as reference as to how a user could set about improving/optimising input audio for best received audio potential (clarity and readability) without touching the insides of the radio.

I know the best option, for Ham users and certified engineers is to tackle the issue at the internal SP and modulation handling and if need be, upgrade the mic preamp to use applicablely suitable improved lower noise and/or higher gain devices, but in today’s VLSI+ radio equipment, even the knowledgeable can do more harm than good making legit internal adjustment without proper test gear. I’m in a position to do so, many Ham operators have some means to assess and measure output purity at some level beyond by-ear estimates, but the TA gear controlled users may find at least the concepts behind what’s going on in the parts of speech processing I highlighted useful insight.

And yes, everything I mentioned in FX terms exists at some AGC or preset level in embedded or daughter board factory SP circuits and modules to some degree of limitation, so it’s all valid.

Yes, it’s clearly better to leave well alone, and set to a ‘standard’ but given the average user wouldn’t be able to correctly assess the ‘standard’ or know if it’s their speech ruining clarity or deficiency in the audio chain, even simple things like replacing an element in a none-captive mic is worth looking into.

I can recall, for example, that 50% of my early Japanese built equipment was optimised around oriental voices which have a distinct register and speech characteristic in speech that Japanese equipment was tailored to, which required some substantial compensation, for my voice, using methods that roughly have the same enhancement affect that today’s SBR sub encoding has on digital P-A audio compression that codecs encode/decode.

Ultimately, since I did state fair warning about ensuring interested users did do any of the above within applicable TA and good practise consideration, their failures and lack of ability is their problem to fix.

After all, irrespective of what class/category we operate under, we are self-liable to ensure we operate according to legal requirement and known good practise that complies with legal requirements. The truest test of practicing good technical practice when optimising your station is being able to assess and utilise information relevant to the adjustment or modifications with understanding and due competency.

After all, if you can do that, you’re in the ‘blind leading the blind’ world on a tech level.

So ultimately, it’s the end user’s liability.

None of my personally applied methods had in anyway negatively affected my ops or others, because I was capable of meeting my station op requirements to assess and evaluate issues and had the background in audio and radio sufficient to tackle both sides of the equation to ensure a clean transmitted content that neither created undue artifacts, undue harmonic issues and modulation stayed comfortably within the the guard provision sub-bandwidth of the appropriate bandwidth of frequency as allocated.

I’m afraid I find this completely pointless I’m afraid. We have severely band limited audio and we have resolution limits to the audio A/D-D/A process. Applying full bandwidth audio sweetening techniques to already mangled audio is in my humble view, a pointless exercise that cannot do anything other than take the nastiness away a little. The digital audio that most radios produce is pretty horrible - and harmonically rich. Audio sweetening of low bit rate, less than idea audio is rarely successful - we can do it to higher bitrate audio with good highs and lows and we can make that sound nicer to some, but processing yucky audio I simply don’t think is worth the effort. I rather dislike the sound of digital radio, and certainly won’t waste any time attempting to enhance it in my studio!

I guess this really is down to the listener - if they like what they do to the audio, it’s fine with me but I won’t be joining the club.

Well, I was focusing on utilising aspects of audio tech to exploit the best of the limits

It’s value, whatever you are experimenting with for optimization or outright experimentation, is there - but if the whole idea of anything experimental optimisation offends, ignore it and mentally write it off, since such legal experimental use of controlled experimental signalling and modulation enhancement isn’t subject to the beliefs of others when the law is on your side supporting a licensed right to do so - and even when not strictly mandated as permitted, it can be no worse than the mere act of breaking open a mic to alter/replace the baffle on non-captivre mics or optimization with different pickup/transducer/mic elements.

As for pointless, well given how much actual harmful ‘pointless proving the principle’ research actually gets done legitimately these days, I’ll just say mine isn’t disruptive whilst controlled and moderately employed, certainly isn’t harmful and falls way short of some of the horrors deemed legit in the name of ‘necessary’ research by a dimension or two.

It’s no less pointless than when somebody experimented with multiplex NB stereo transmission on 2m quite a few decades back - the fact it worked enough to prove it could work and some QSO’s were undertaken in a serious of scheduled operations was a major step - I don’t think it went any further, as it was more of an idea that needed a reason to exist to be recognised, but it was NOV backed and legit and that was all that matters. What others thought of it’s use and presence and not being proper NB modulation changed nothing about it’s legitimacy.

Nothing is pointless when we postitively and productively learn from it , unless it’s the harmful stuff arrogantly and maliciously performed to the actual detrimental of others that’s as pointless as it is fundamentally wrong.

Oh yes - I remember very well using the entire 70cm band to send video to Holland. I just think that we just have to get used to the digital sound and not waste our time sweetening it. No the technicals, the aesthetics