PA8W Amateur Radio

Wil, PA8W,  E-mail: PA8W@upcmail.nl           
Back

Analyzing and adjusting your stations audio

So here's the hard part; now we know what's important for intelligibility, it is time to implement this in our station.

Lots of modern transceivers offer enough control over the frequency response and over the compression ratio (often called processing) to optimize your audio.
In these cases, you don't have to add additional devices to get things right.
However, proper adjustment is easier said than done, allthough often a necessity, since a factory default is not likely to bring the desired result. 
I've heard plenty of brand new, high tech transceivers that sound absolutely crappy right out of the box.
They really need some attention and tweaking to get them right.
The audio report of the other stations is important, but generally too poor to really refine your adjustments. 
A good way to go is: Monitor yourself using good headphones on a second receiver.
Do not only use your transceivers monitor function for your judgement; 
your audio does not pass all stages in a normal radio chain, and therefore it will yield different results.
No, take a second radio and good headphones, listen for a while to stations with the best audio you can find, and then listen to your own transmitters signal.
If you repeat this a few times, you will get a good impression what your audio is like, and what is needed for further improvement. 
Take plenty of time for this, and divide your judgement into two parts, the dynamic behaviour, and the frequency response, as discussed further:

Dynamic behaviour:

Lots of transmitters can be overloaded quite easily, with loud but unreadable audio as a result. 
This is absolutely to be avoided! Check if you can find a setting of audio gain and compressor level to avoid overload, but at the same time the less loud parts of speech must be boosted a bit to keep them audible. Take care to use the microphone in the right way; do not hold a handmike right in front of your mouth at only a few centimeters distance as you see in movies; you will surely "pop" the mike with every letter P or F  you say. Here's a bad example: sound/popping.mp3
It's better to hold a handmike in a 90 degree angle beside your mouth, so the top or side of the mike is towards you, in this way you're less likely to blow into the membrane. Be sure to keep a constant distance, and train yourself to use a steady volume of speaking.

Note that a battery fed mobile transmitter may initially sound great with a full battery, but may overdrive easily when the battery voltage slowly drops after some time of operation!
This is a commonly heard problem, since the mobile operator has no or little means to properly monitor his signal.

If your radio appears to be easily overloaded, you could decide to add a limiter-compressor to your audio chain, but take time for proper adjustment of this device as well.
Read the page on tayloring your audio in the audio menu.

Avoid unnecessary high settings of audio gain even if your set shows excellent dynamic behaviour; it will emphasize room reverberation and the noise of fans, and even your breathing noise will become audible.
Also, your microphone input will be more likely to pick up RF radiation. So, just don't overdo it.
Here's an example of way to much compression, the S-meter showed S9 throughout the recording, during speech as well as in speech intervals. 
sound/compressed.mp3
And here is an example of far to much audio gain.
sound/compressed2.mp3

Frequency response:

The frequency response is highly depending on your specific microphone. Some handmikes really sound poor, but if your handmike uses a electret element, it should be quite useful. The electret has a nice flat frequency response, and is quite immune to RF radiation.
If you are thinking of a different mike, read the page on microphones in the audio menu.

Listen carefully to your audio, and compare it to other stations with clear audio. You will almost certainly have to take out some bass, and turn up some treble in most sets. 
(often called Microphone EQ or Transmit EQ in the rigs menu)

Just experiment and see how much low end you can take off before you start to sound "thinny". The low end is just moving S-meters on the other side, activating theirs AGC as well, causing your audio to sound less powerful.
Taking off the excess low will make your signal sound considerably louder on the other side, because all of your power is used for the functional part of the audio passband.

If your set has no proper tone control, but plenty of excess audio gain, you may put a small series capacitor in your mike to get rid of the excess low. The roll-off should begin to set in below about 500Hz, depending on the roll off already present in your set.
At 250Hz your audio should have lost a few dB already.
You may also improve the high end:
Most radio-microphones use capacitors running from the element to shield to take away RF. They may be chozen quite big, and a reduction of that capacitance may get you more brilliance and still provide good RF filtering.
In my Icom handmike I reduced the capacitor to 1/5th of its original value. And this was a big improvement.

The last step is to get some external tone-control.

Read the page on tayloring your audio in the audio menu.

Some examples of good audio and some retransmissions: 

Note that considerable changes in frequency respose will have effect on dynamic behaviour as well, and vice versa. 
So if you change one, check both!

After your tests and tweaking, ask a few stations with good audio themselves about their opinion, you may get something like this:
sound/g4awj01.mp3
(I was testing a new mike with the settings of the old mike, which resulted in a little too much high frequencies...)
 A small real-time adjustment resulted in this:
sound/g4awj02.mp3 
And no, I am not using external audio processing, just a carefully adjusted, fine transceiver:
sound/g4awj03.mp3

Note that any comments on your audio, good or bad, are rather useless unless the other side explains why he has this opinion!

 Another on air check: 
sound/2m0cdo01.mp3
sound/nice ssb audio.mp3
And a spontaneous comment on my audio in another QSO:
sound/lovely audio.mp3

Here's some examples of very good and loud audio from II5GAL:
sound/excellentaudio1.mp3
sound/excellentaudio2.mp3

My QSO with Gordon, MA0GPZ/p who was operating on 40m from the beach of EU010, so no fancy audio equipment there, but nevertheless very functional audio: 
sound/ma0gpz1.mp3
sound/ma0gpz2.mp3
sound/ma0gpz3.mp3
And again positive feedback on my audio from Gordon as well.
 
In QSO with David, G8KAP: Listen hoe well he can be copied in spite of all the QRM. His signal was hardly moving the S-meter over the QRM.
sound/dave.mp3
 Now that' s functional audio!

Here's a sample where I am not very strong near Quebec, but my audio pulls me through:
 sound/va2pw01.mp3  
And a similar case: 
sound/but very nice audio.mp3


And here's how my (modified handmike) audio is recorded and re-transmitted by Tony, IK1JUO: 
sound/tony.mp3

 Note that my audio here has too much compression, since my signal has been processed twice, once during the original transmission, and then again during the IK1JUO transmission. This will also affect the frequency response. Still nothing to be ashamed of. Thanks Tony!

And another re-transmission, this time by Fred, PA0PAF, who has a nice, relaxed sounding audio thanks to his software defined radio:
sound/pa0paf.mp3
The first 10 seconds is my voice re-transmitted, the rest is Fred. Did you notice I've got the bigger mouth? Also my thanks to Fred.


Conclusions:

If you want your small station to sound big, don't waste your transmitters energy on low audio frequencies.
Reduce the low end and emphasize the high end to put most energy in speech intelligibility.
But don't overdo it though, try to stay reasonably close to the natural sound of your voice.
The emphasized high end will make your voice less difficult to copy in noisy conditions,
With the reduced low end your transmitter will have 6 to 10dB (!) more power available for the important part of the spectrum!

Once I experienced a striking example:
 A station with an awful lot of low end in his audio, running 100w and a lot of additional audio devices and a super microphone, responded to my CQ calls on 40m.
I heard him call amongst other responding stations after at least 4 of my CQ-calls, but every time other guys with "economic" audio beat him in the small pile-up, even one station running 10 watts...

So at last, I called for this specific station only, just to give him a chance, and had a QSO with him.

I told him how he was surpassed in the pile up again and again because of his audio properties, 
which sounds great by the way, only as long as no other signal was spoiling the fun...

Then he increased his power up to 400w and only then he was a match for the 10w station...
This guy would have got through much, much better if he would have cut off everything below 250Hz...

However, there's nothing wrong in restoring a little low end when having a QSO in perfect conditions.
When you're constantly 59+20dB, there's plenty of energy available to "dress" your voice with a warm sounding low end.
And you may take back some of the compression as well since there's plenty of headroom to keep everything audible.


Pre-emphasis / De-emphasis:
Wouldn' t it be nice if all hams using SSB would:
Emphasize the high end up to 10dB on transmission, using a standard R/C time,
and reduce the high end 10dB on reception, using the same standard R/C time.
Because then, our signal / noise ratio could improve up to 10dB with almost no costs...

 Revolutionary? Not at all.
The same trick is done in FM technology, to improve the already very good signal / noise ratio there.
And in much more (ancient) audio applications, where the audio has to pass some kind of noisy process. 
As far as I know, this technique is not used in SSB, where 6 to 10dB improvement really would make a big difference.

How about some experiments?

Enjoy!

73, Wil.


Back