Personalized Equalizer Settings for K3S and Modern Radios

Modern transceivers frequently allow you to set your receive and transmit equalizer settings. Setting the receive is fairly easy since you just have to listen to your own speakers. On the transmit levels, I'm sure other hams have had the same problem I have. How do you pick?

Initially, I set them according to the settings recommended by Heil Sound, the makers of my microphones.Now, that seemed better than my K3S's default of no modification. All seemed well, but I still had problems. Frequently, I had to repeat the Yankee portion of my call. My wife, KC0RBV, also had problems with the same portion of the call. Watching the SWR meeting while my daughter talked on Kid's Day really gave me a shock. Her voice was able to keep the meter pegged at 80 watts almost the entire time. Her voice appears to be made for SSB. Then, I wanted to improve my setup and maybe use the equalizer (EQ) settings to improve my transmitted power and ineligibility.

So I read up on cardoid microphones and how to generate the most intelligible signal. It seems cardoids often add a little bass to a persons voice. And that bass does not improve intelligibility. So I decided that removing the lower end frequencies would help improve my transmitted voice and decrease the bandwidth so I could improve my transmitted power. I also do not have a terribly high voice so I could probably cut-off the upper frequencies. The question was how to do this. I could, and did, look through online forums for other settings to see what was better. The problem with this approach and using anyone's defaults is that our voice is unique to us. So EQ settings for one may not be the best for others. A second approach, often suggested, was to do on-air testing with another ham listening or another radio. I didn't have a second radio to listen with, and I have a schedule that doesn't work well with many of my ham radio friends in the area. Plus, I didn't like the idea of spending a lot of time going through all the EQ settings and taking up valuable frequencies for testing. I wanted another way. What I am going to cover in this post is a way to help select EQ settings using freely available software.

In my search for a way to do this, I realized that Audacity has an equalizer panel within it. Audacity is an opensource audio editor. I also had a USB microphone head set. Between the two, I thought I may be able to do something. The first step was to download and install Audacity and insert the USB headset. I then made a recording with no changes in equalization to hear how my voice sounded. You can listen to that file here. Figure 1 shows the recording.

Figure 1. Initial recording in Audacity

Next, I wanted to here the effect of Heils Sound's recommended settings. I did a file Save As so that when I applied the equalization I didn't loose the original. Then, I clicked Ctrl+A to select the entire sound, clicked Effects, Equalization (See Figure 2). Audacity actually has more equalization ranges than the K3S or other modern radios (See Figure 3). To replicate what I think was going on in my K3S, I set all of the frequencies below 50 Hz and above 3200 Hz as low as they would go in Audacity to suppress them. Then, I used the following settings recommended by Heil Sound for contesting and DXing. I set the Audacity equalization to -16 dB for 50 Hz, -16 dB for 100 Hz, -10 dB for 200 Hz, -6 dB for 400 Hz, +4 dB for 800 Hz, +6 dB for 1600 Hz, +8 dB for 2400 Hz, and +12 dB for 3200 Hz. Audacity does have some additional frequencies between these so I would set their value to be half of the adjoining frequencies. My assumption being that is how the K3S was behaving. Remember, this has been how I have had my radio set for some time.

Figure 2. Menu selection for Equalization under the Effects menu

Figure 3. Equalization window in Audacity

Figure 4. Recording after recommended Heil EQ settings

Listening to the result made me immediately realize what was wrong. The Yankee after the eight was very weak. If you listen closely, you can hear the issue, too. I also noticed that the highs on the recording became much larger as shown in Figure 4. The Heil settings were amplifying the high portions of my voice too much and suppressing the lows more than they should. Hearing and seeing this really embarrassed me. It was definitely my signal that needed clearing up. All those poor South American stations that had so much trouble! (Sorry guys and gals!) And I now know why the audio engineers are needed so much in the recording industry. Again, I hit Save As.

Now the question was how to fix it. From my reading across the internet and from things I learned from some YouTube videos I took the following approach. First, lower frequency sounds do not tend to significantly contribute to intelligibility, and in fact reduce the average transmitted power. Another consideration was to keep the bandwidth down to around 2 to 3 kHz to match SSB. I reopened the original file since applying equalization to an already equalized audio will give you funny results. Next, every thing below 400 Hz was set to -16 dB. 400 Hz was set to -3 dB. My voice does not have a lot of highs so I set the 3200 Hz range to -16 dB. I listened to these settings. I repeated the process over and over until I thought I had a setting that replicated my voice well and didn't have any extras. You can hear the results here. You can also see in Figure 5 that the audio looks more like my original recording, but with less variation.

Figure 5. Recording after custom equalization

My final equalization settings were as follows: -16 dB for 50 Hz, -16 dB for 100 Hz, -16 dB for 200 Hz, -10 dB for 400 Hz, 0 dB for 800 Hz, +1 dB for 1600 Hz, +3 dB for 2400 Hz, and -16 dB for 3200 Hz. Very weird and unlike anything I had read on the internet.

Remember, the purpose of this is not to give another set of recommended settings, but a process for determining the settings you should use for your voice. Each person's voice is a little different and may need different settings. Mine look really funky compared to what I have seen online. Your equalization may also be drastically different.

During the remainder of the contest season, my wife and I noticed far less repeats. Particularly for the Yankee portion of my call.  So it seems to have helped. It also seemed to help bust through a few pileups, but that may just be wishful thinking.

There are some things that I would do differently. First, I probably should have connected my Heil microphone to the computer for recording. Using my actual mic would provide a better representation of what is going on in the shack. I could have also determined a couple settings for on the air comparisons with a buddy. I am also thinking of creating custom EQ settings for each operator since our voices are so different. This style of recording allows you to have the person talk once in the mic and then you can equalize without them. Thus, this process would enable multi-op stations to find settings best for each operator.

I hope this post helps you improve your station. It is a low cost way to improve intelligibility and signal strength by finding better settings for your radio.