K6JRF's Page
formerly W6FZC
WB Audio Techniques Page
(Updated: Sept 21, 2008)

This Audio Page discusses the requirements to transmit good, clean SSB audio by exploring fundamental radio and audio characteristics required to produce this kind of audio along with explanations of how-to-do-this with the FT1000D.

What makes good audio?
This is a difficult question because audio can be evaluated using many different criteria. If I had to pick out just one characteristic, I'd pick 'BALANCE'! It makes no difference if the audio is wide band or narrow band, if it's balanced, it'll sound good!

By this is meant, equal levels of signal from the lowest to the highest frequencies present in the output audio. Note that this does NOT have to be from DC to LIGHT! Your rig may only be capable of producing 300hz to 2.4Khz. That's ok. If the audio is 'balanced' across this spectrum, it can be very pleasing and 'articulate'. Certainly it will not sound as full and pleasing as a wideband 80hz to 3.2Khz audio signal would, assuming it's also balanced.

Ah, that word 'articulate'. What does it mean? Webster's dictionary says 'made up of distinct syllables and words that convey meaning easily and clearly'. I guess that's what good audio does, doesn't it? You betcha!

In wb audio terms, we try to get that quality over as wide as possible, from 50hz to 3Khz. And wider if your radio will allow. TS-950SDX can go from 50hz to 4.0Khz with a little educated push. My FT1000D, after mods, will go from 70hz to 3.2Khz. The FT1000MP and TS-870 are similar.

We will discuss equipment and techniques necessary to achieve wb audio.

How-to-do
There are no hard and fast rules per se but there are certain steps that are necessary to produce this. The first starts with the microphone. As I've said before, keep your old Shure 444 in the bottom drawer. It will NOT produce clean, articulate wb audio.

There are many choices for a quality mic which I mentioned in the AUDIO RACK and MIC section on the front page. Take the time to check these out to see if one might be right for you and your budget. This is very necessary step. Without a quality microphone, good Tx audio is not possible!

The next choice concerns the audio dynamics. The term 'dynamics' means all steps normally associated with processing of the audio signal. These are: gating, compression, limiting, de-essing and equalization. In order to more comprehend these, click on the many links in the aforementioned section. Take the time to read the details contained on these sites. It will help you understand the 'terminology' and give you ideas as to what is needed in your equipment lineup.

The chart below shows my current audio system equipment and the diagram shows all of the major interconnects between these equipments. These do change from time to time but the basic "ordering" concept has not changed. The numbers in blue refer to the input impedance of the stage except for the microphone where this is its output impedance. The balanced interconnections are employed between all units using either XLR or TRS connectors except for the Alesis Microverb. The output from the Alesis Microverb is unbalanced as is the FT1000D mic input connection so one channel of the Ebtech Hum Eliminator is used to ensure a ground loop free connection. The second channel is used for the Sony MD playback Eq, ensuring complete isolation between the audio processing system and the FT1000D.

K6JRF's Current Audio System Interconnection I firmly believe that the 'cleanest' sound is produced by;
1) attenuating any 'unwanted' frequencies before they enter the main processing chain,
2) using equalization (DSP1100) and special effects (Ultrafex) processing ahead of any Gating, Compression and De-essing (or Limiting),
3) always adding reverberation (MicroVerb III) effects as the last piece of equipment.


Therefore the my audio processing system reflects this philosophy.
1) For my voice, '400hz' is a no-no. Thus the Behringer MIC2200 preamp's eq stage is used to remove this frequency before it enters the processing.
2) The Behringer's EX3200 Ultrafex and DSP1100 are placed in front of the dBx DDP processor. This gives the DDP's compressor, limiter and de-esser the ability to compress, limit and de-ess the audio signal so that there is no chance of overdrive, harshness or clipping to the audio signal that drives the mic input of the FT1000D.
3) The Alesis MicroVerb III is at the end of the chain. The reason is simple; the gate in the DDP closes very quickly and thus the reverberation effect would be 'cutoff' unless the gate HOLD time is made much longer. This tends to defeat the action of the gate!


Real ESSB Examples
In order to see what's involved, what's better than a picture. These are spectrum plots made with SpectraPlus, a software program that turns your computer's sound card into a low frequency spectrum analyzer. The picture below shows the transmit audio response of my former radio, FT1000D taken by VE6CQ. The Tx bw is basically flat from 50hz to 3.1Khz. The key is the balance that it possess. The difference from top to bottom is less than 3dB - 4dB. Also the mid range frequencies are slightly attenuated so as to bring out the resonant bottom and top frequencies, where the 'articulation' lies.

The next two pictures show the most recent plot of yours truly taken by VE6CQ and W0WD. Note that with the new preamp (described below), I've reduced the heavy mid range frequencies and ended up with a more balanced audio from 'bottom' to 'top'. With the slight rise around 3Khz, brightness and clarity are present in the transmitted audio making it 'articulate'

K6JRF's spectral plot via SpectraPlus

K6JRF's spectral plot via SpectraPlus


FT1000D Techniques
The following sections are specific to the FT1000D radio and show my current alignments and adjustments along with the external audio equipment and its settings.

I've tried to maximize the TRANSMIT upper frequency point and did this by adjusting all oscillators and trimmers that control this in the radio. The 8.67Mhz IF (TC4001), 55.48Mhz (L4071) TCP and the two (2) capacitive trimmers (TC4006 for USB and TC4007 for LSB) were adjusted to get the highest possible frequency response from my InRad filters in the '2.4K' position (InRad #716 and #707C). The results are shown above by the frequency response spectrum plots.

TCP Adjustment for Highest Frequency Offset

Do not perform this adjustment if you are not sure of what you are doing!!!

By adjusting the TC4001, L4071 and the two (2) capacitive trimmers, TC4006 and TC4007, you can offset the 8.2Mhz filter (InRad #716) to the highest possible point that the radio will allow. This procedure is described here.

Refer to the FT1000 Service Manual, page 4-15 for details of the TCP adjustment. I use my Sony MD with a previously recorded set of tones: 100hz and 3100hz which I feed into the mic jack input. You can also use a signal generator but the MD technique is much faster and accurate. Also required is a output power meter that you can easily see while doing this adjustment. Set the FT1000D power output control to about 1pm and mic gain at 10am.

As you make adjustments, monitor the power output from the radio for each tone. This is a relative adjustment so trying to get the highest power out at BOTH frequencies (100, 3100hz) is not important. In fact when finished, you should see higher power out at 3100hz than at 100hz by about 3dB. If you don't, then the adjustment is not complete.

1) With the front panel hinged down, the radio at 14.2Mhz, USB selected, adjust the TC4006 for the highest offset frequency. Then adjust L4071 to move it up even further. Then readjust TC4006. Do this while exciting with 3100hz.
2) Select LSB; most likely the power is way down. Adjust TC4007 for the highest offset. Then adjust L4071 back a bit in the opposite direction from the first adjustment) to bring up the LSB output power. Readjust TC4007.
3) Select USB; Perform step 1) again but this time use an educated guess as to where to set L4071 to get a balance (ie, equal power out) point that BOTH sidebands can attain at 3100hz.
4) Select USB; Adjust TC4001 slightly to see if USB power goes up. If so, then repeat step 1) and 2) making small adjustments while switching back and forth between USB and LSB AND switching 100hz and 3100hz tones!

Put the metal caps back on to each section and raise the front panel. Let it sit that way until it fully warms up. The adjustments will change slightly and need to be lightly 'tweaked' again.

Results of TCP Adjustments
The following chart shows the results of the TCP adjustments on the final Tx audio when equalization is employed as detailed below. The charts show my current alignments and adjustments to my FT1000D along with the external audio equipment and its settings.

K6JRF Transmit Eq

The RED trace show the un-equalized response via a pink noise sweep into the front panel mic connection (Behringer MIC2200) without the Eq stage. Note the low frequency response has now fallen a bit since the filters have been 'pushed' to the highest frequency offset that the FT1000D can produce. This was accomplished by alternating adjusting the 8.67Mhz, 55.644Mhz oscillators along with the two (2) trimmer caps (TC4006 for USB, TC4007 for LSB) until the highest balanced frequency response for both sidebands was achieved [see Table Inset above].

The low frequency response 'peaks' around 400hz and if not attenuated, makes the resultant Tx audio very restricted sounding like Broderick Crawford of the now famous "Highway Patrol" series! Note that BOTH the low and high frequency response is at least 10dB down from the response at 400hz.

Now we add in the Eq stage in the MIC2200 preamp. The PURPLE trace shows the pink noise response with 400hz attenuated by -10dB. Removing this frequency (and adjacent frequencies) allows the higher frequencies to be heard.

Next the Behringer EX3200 Ultrafex response is shown in the BLUE curve. Now the low frequencies (50 to 100hz) have been eq'ed to the same level as the high frequencies.

The low and high frequency losses are typical all of ALL radios and would be very representative of what it takes to get flat transmit audio in most any radio! Note that it takes at least 10dB of additional high end gain to get flat high frequency (1K - 3Khz) response out of the radio.

The AQUA chart shows the response after the Behringer DSP1100 in included. Now the final 'polish' is added to the Tx audio. The 200hz area is attenuated by -8dB but the 80-100hz area is left flat. The gives 'balls' to the audio without the boominess that the 150 - 200hz frequency area gives. Next a rising response from 1Khz to 3.2Khz is added to give the Tx audio the 'clarity' that is so important to balance the heavy bottom response.

The chart below is a 'visualization' of all equalization currently used. Note that the highlighted section of the AQUA chart line compares almost identically with the RED enhanced portion of the BLUE line. The 60 hz notch is used to kill the 60hz compressor noise from my room air conditioner. The exact frequencies can be found in the Behringer DSP1100 section of the "Current Audio Settings" table below.

This software can be download from the Behringer web page by clicking on the Editor Software link. It's designed specifically for the DSP1100 but can be used to visualize any eq combination as long as the frequency, bandwidth and level are known.

K6JRF Transmit Eq

The Behringer DSP1100 is currently provides all frequency equalization including attenuation of room air conditioner compressor 60Hz noise. I've also transferred all equalization settings from the DDP into the DSP1100's memory, thus removing all of the task from the DDP's 3 band EQ. The 'key' audio frequencies for my voice are 70hz, 200hz, 400hz, 3.15Khz.

Behringer DSP1100 Destroyer

The DDP does not effect the equalization response since the eq section has been switched off. The DDP's response is the same as that appears in the RED chart; ie no coloration of the audio. This is an important feature of any 'GOOD' audio dynamics processor.
The following chart shows the FT1000D RECEIVER characteristics using the present filters (InRad #710 and #707C) in the '2.0K' position. The #710 is a 6Khz bw filter as is the CMF, #707C. The cascaded bw is around 5Khz. The AQUA chart shows the raw response of the FT1000D as it is currently adjusted. This is typical of most receivers with the exception of the TS-870!
The Yellow chart shows the compensation applied to SpectraPlus to allow accurate comparisons to other stations without introducing errors from your receiver.

K6JRF Receive Eq


Sweep Test Setup

Pink Noise Setup

The photo shows my method to sweep my audio rack and the inset shows the output into my laptop running SpectraPlus. The Sony minidisk (MD) recorder contains both the sine wave sweep (30hz to 5Khz) and Pink Noise (20hz to 20Khz) data. The MD outputs through a 40dB attenuator into the 'MIC' input of the Aphex 107 Preamp. From there, by activating the appropriate 'bypass' buttons on each equipment, it is possible to sweep each piece of equipment separately to determine its response. Then by adding each back 'inline', you can tailor the final response. Note here the emphasis was to get high end response (2Khz to 3.1Khz).

Audio Dynamics Settings:
Here are the current settings for all of my audio equipment. Today (8/25/00), I changed most ALL settings! Why? Because I've replaced the Aphex 107 Preamp with the Behringer UltraGain 2200. This new unit has a parametric stage (boxed in red) that allows attenuation of the heavier low mid frequency range (200 to 400hz) BEFORE it gets into the following stages. Also this frees up the DDP to do some other processing.

Behringer Pro Mic 2200 Preamp

The picture shows a closeup of the controls of one of the two separate preamp sections. The preamp features tunable low cut (from 32hz to 320hz) for removing low frequency noise such as air conditioners; Separate input and output gain controls; a parametric equalizer that's fully tunable from 20hz to 20Khz with Q and level control.

After installation of the Collins filters (see main menu, #7), the audio took on a heavier flavor due to the excessive mid frequencies. Couple this with the increased low and high frequency response, made the transmit audio sound mushy and heavy in the 'mids' with some 'tearing' on the top side.

[Current Audio Settings: Updated Jun 11, 2002]
Latest Changes In RED
Behringer UltraGain Pro2200 PreAmp:
Phantom: On (+48V)
Gain= +37dB
Low Cut: On, 50Hz
Parametric Eq: Fr= 400hz, BW= 1.8, Level= -10db
Output= +8dB

Behringer Ultrafex EX3200:
Bass Processor: ['Low'; Low Mix: '3.0']
Multiband Processor [NR Sens: 2.5; Tune: 2.75 Khz;
Process: +5.0; High Mix: 5.0]
Behringer DSP 1100: Filter Mode: PA (Parametric Eq)
[Engine R: Filter: 1; Freq=63hz; Fine=+6 (=68hz); BW=10/60; Level=+8db
Filter: 4: Freq=200hz; Fine=0; BW=20/60; Level=-8dB
Filter: 6: Freq=3.15Khz; Fine=0; BW=20/60; Level=+8dB]
Filter: 7: Freq=4.0Khz; Fine=-9 (=3.6Khz); BW=10/60; Level=+4dB]

dBx DDP:Com: Attck:0.1us, Hld:44ms, Rel:360db/sec, TCM:2.0ms
[Gate Th:-25 Rat:1:5] [Cmprsr OvrEsy:Off Auto:Off Th:-25 Rat:2.4:1 Gain:4.0db] [Eq Bnd1 Fr=63 hz, Q= 8, L= +6;
Eq Bnd2 Fr=---, Q=---, L=---;
Eq Bnd3: Fr=---, Q=---, L=---]
[De-esser: Not Used] [Limiter: Th= 0dB]
Alesis MV III:
Input:-2.5 Mix:-4 Output:+2.5
Low-Eq:1pm, High-Eq:2pm
Prgm: 'Plates'
Num: 7
K6JRF Audio Equipment

The picture shows my original settings. If you take the time to compare, you'll see that they all been changed in order to minimize the amount of compensation (in dB) needed in Filter 6 of the DSP1100. This setting was raised to +8db with a Q of 4.32 (BW=20/60)! This gives a nice boost from 2.0Khz through 3.1Khz that I need. Also the Compressor's Over-Easy function was turned OFF resulting in more audio 'attack'. The De-esser was substituted for the Limiter with the advantage of no low frequency clipping as does the Limiter. It limits ANY frequency when it reaches the threshold setting; the De-esser limits any frequency above the frequency set point (3.0Khz) by the amount indicated (40%).

The Aphex 250 AE was replaced with the Behringer EX3200 Ultrafex. This device is similar to the 250 but it is much more straightforward to setup and use. These devices are necessary to give clarity and brightness to the Tx audio since the radio's filters tend to attenuate higher frequencies and 'muddy' the audio clarity.

DDP Transient Capture Mode
Unlike analog technology with its response speed limitations regarding changes in amplitude, digital signal processing permits differences in amplitude to be identified in advance but you must use a bit of signal delay. Increasing this delay also increases the potential for the intelligent control. Even “looking ahead” by only a few samples is sufficient to ensure the intelligent application of dynamic processing – such as limiting, which ensures an absolutely reliable signal ceiling – without clipping.

The DDP's Transient Capture Mode (TCM) works on the principle of delaying the audio signal and letting a 'control' signal begin to activate the response of the VCA (Voltage Control Amplifier). The overall result is the perception of a very fast moving compressor that is able to catch the front of almost every transient signal, but not delaying the audio signal enough to produce any phase correlation errors. The DDP can vary the delay with the range from 0us to 3ms. This gives the processor enough time to react before the signal arrives at the point of processing.

The controls for TCM lie in the GATE parameter area. This effect is global but the GATE MUST be turned on in order for the TCM module to work.

The implementation in the current DDP allows the signal to sound smoother and makes it much easier for the DDP to process the COMPRESSOR, LIMITER and DE-ESSER functions because it catches ALL transients .

The DDP's TCM setting was changed from 500us to 1.5ms. This allows more time to set up the VCA gain and allows the processor to 'handle' the high level of energy at 3Khz that I'm using. Without this higher TCM setting, the transients were causing distortion and tearing. This tip thanks to N8QW.

Equipment Alternates:
Many say that the 'stack' of audio equipment is too much both in cost and physical size. Isn't there something smaller and more compact that does the job? The answer is Yes!

W2IHY markets an all-in-one audio processor called the 8 Band Equalizer. It's small and compact, has provision for different mic inputs including XLR connectors, can provide 5V phantom power and has a unique adjustable noise gate that cuts out background noise. It can purchased complete or in kit form including a cable to plug directly into your radio. This make a good alternate to going the 'full' audio route and may appeal to the majority of hams.

Send me Email Icon with your comments