TRX-1: Volume equalization between P25 Phase 2, Phase 1, and other modes

[dem1]

Now that I've begun listening to a wider variety of systems on my TRX-1, I recently experimented with various settings to try to reduce the volume level disparity between different modes, especially the notoroiusly loud P25 Phase 2 talkgroups. This combination of settings worked best for me:

1) Set DAC Gain to minimum (-10db) using EZ Scan (Advanced Features > Receive/PCIF settings tab) or via Global Settings on the TRX-1. This reduces the volume level of all digital objects. Be sure to record your current setting before changing it in case you wish to revert.

2) To compensate for the volume drop in step 1, turn on Audio Boost for all objects in all digital modes EXCEPT P25 Phase 2.

3) Turn on Audio Boost for all analog NFM objects. Leave Audio Boost off for all analog FM and AM objects, including most legacy 800 MHz analog trunking systems (Motorola, EDACS, etc.). While the TRX-1 uses a narrower IF filter when NFM is selected, it does not automatically boost audio output of NFM objects to adjust for reduced deviation.

These changes gave me more consistent volume levels between digital and analog. P25 Phase 2 is still louder than other objects, but the difference is less dramatic and no longer knock me out of my chair. While I leave AGC turned on globally, it seems to have little impact either way.

 

[jaspence]

Some of the problems come from the users. The distance of the microphone from the users mouth can make a noticeable difference, as can the individual's voice pitch and quality. Better radios have adjustments for this, but even the best cannot compensate for a microphone an inch away from the mouth of a user who has a strong voice versus a weak voice several inches away. With professional radios, adjusting each one for the user would be a time consuming task.

 

[seth21w]

Whistler has the best agc of all scanner makers. My whistler had no loud audio on phase 2 with agc on. I dont remember my dac adc settings.

 

[Wackyracer]

You can't increase analog volume, but you can turn digital volume down via DAC as described in the original post. Making your audio volume more balanced.

 

[DJ11DLN]

Going to give this a try later. The agencies I monitor on Hoosier Safe-T are all over the place on p.1...some you can barely hear, others seem as though they are gonna pop the speaker. Before the switch-over from Type II to P25 they were all pretty close to the same level, taking into account variances by actual operators. Indiana got jacked pretty good by big /\/\ this time it seems, at least on the set-up side.

 

[Ubbe]

When you listen live the agc seems to be doing its job but the recordings seems to add one additional agc function that are way too exaggerated and sounds terrible. Agc on an already agc corrected audio signal never sounds good.

Whistler uses an audio delay line to get rid of the squelch tail, but it is tailored for the speaker amplifiers slow reaction to unmute and mute the audio. When recording there is no slow reaction and each analog recording starts with a noise burst. When a mobile signal fluctuate at the SQ level the recordings are impossible to listen to. It would be so easy to fix in firmware to add a little time delay to the recording function and reduce the recording agc 3dB to make it more usable.

/Ubbe

 

[dem1]

When you listen live the agc seems to be doing its job but the recordings seems to add one additional agc function that are way too exaggerated and sounds terrible. Agc on an already agc corrected audio signal never sounds good.

Whistler uses an audio delay line to get rid of the squelch tail, but it is tailored for the speaker amplifiers slow reaction to unmute and mute the audio. When recording there is no slow reaction and each analog recording starts with a noise burst. When a mobile signal fluctuate at the SQ level the recordings are impossible to listen to. It would be so easy to fix in firmware to add a little time delay to the recording function and reduce the recording agc 3dB to make it more usable.

/Ubbe

I too hear the noise burst you mention at the beginning of analog recordings on my TRX-1, but only if in carrier squelch mode (CTCSS/DCS turned off). However, I have no speaker audio delay whatsoever on analog channels. I've verified this by listening to transmission on TRX-1 side-by-side with other receivers, and by listening to my own ham radio signals while transmitting. If there were an audio delay I'd hear a very distinct echo, even with delays as short as 25 milliseconds. Perhaps the recording captures audio a split second before the squelch opens in order to compensate for the CTCSS/DCS decoding delay - and in carrier squelch mode we get a split second of noise instead?

It does appear that for recordings, the unit has a separate AGC circuit with a slow recovery time. I can hear the audio gradually "ramp up" at the beginning of each transmission, which does not happen when listening via the speaker. This is especially noticeable with a constant audio source such as a PL tone. The same thing can be observed during a transmission after a loud audio peak.

I find that analog recording works best with a scan delay programmed. The recording continues for the duration of the delay, even if scan mode is not active. If delay is turned off and a mobile signal fluctuates in and out as you describe, the recording stops each time the signal cuts out, and a single transmission will span many very short audio files. Interestingly, if I pause the scan by activating "hold" (||) on a conventional analog channel with no delay, the unit ceases recording of subsequent transmissions until I release the hold.

My initial comment at the beginning of the thread about AGC having little impact refers to the "Digital Audio AGC Setting" which can be turned on or off globally or per digital object. All of the digital systems in my area have healthy audio levels, so the impact of digital AGC may be less obvious. But Phase 2 talkgroups sounded much louder than everything else (as noted in past threads), and that was primarily what I sought to mitigate.

 

[Ubbe]

Perhaps the recording captures audio a split second before the squelch opens in order to compensate for the CTCSS/DCS decoding delay - and in carrier squelch mode we get a split second of noise instead?

Yes! I had somehow always known that Whistler could predict the future!

/Ubbe

 

[Anderegg]

I've been using the same -10db setting for digital, with digital talkgroups set to audio boost. Maintains a pretty good match in volume for P25 and digital on my Whistlers. You can lessen the negative gain in steps if you wish to match further, but I find the -10db max to provide the cleanest digital audio to my PC.

Paul

 

[jonwienke]

Yes! I had somehow always known that Whistler could predict the future!

Time travel isn't required if you have a 1-second rolling audio buffer, so that if the CTCSS decoder indicates squelch should open, you still have the beginning of the transmission in the buffer, and can play the entire transmission, albeit with a slight latency.

That would require adding the audio buffer hardware to the scanner design, but that would be a lot cheaper than a time machine.

 

[dem1]

Time travel isn't required if you have a 1-second rolling audio buffer, so that if the CTCSS decoder indicates squelch should open, you still have the beginning of the transmission in the buffer, and can play the entire transmission, albeit with a slight latency.

That would require adding the audio buffer hardware to the scanner design, but that would be a lot cheaper than a time machine.

Taking this idea one step further, digital audio recorders by nature need to utilize a buffer due to the time it takes to process and save incoming audio. So let's assume that when the TRX-1 scans an analog channel, its recorder begins buffering raw audio from the discriminator even before the squelch opens. If a transmission is detected, the open squelch signal triggers the recorder to immediately begin saving the audio it has buffered. Since the currently buffered audio was actually received before the transmission began, it is saved as an initial burst of noise. Or in CTCSS/DCS mode, it is saved as a small slice of the actual transmission received during the CTCSS/DCS decoding delay interval.

To test this hypothesis, I recorded a very brief ham radio transmission utilizing both the TRX-1's recorder and an independent unsquelched receiver with its audio output fed to Audacity. I then imported the TRX-1's audio file into Audacity and aligned the two tracks. The two second audiogram below shows the results. The independent receiver is on top and the TRX-1 on the bottom.

The top track begins with open squelch white noise, followed by silence as the transmitter is keyed up, along with a brief "click". The audio segment in the middle is a quick touch tone, and there is more open squelch white noise after the transmitter unkeys. The lower TRX-1 recording captures the same signal, plus what appears to be a brief burst of the white noise buffered by the recorder right before the squelch opened.

You can also see the effect of the TRX-1 recorder's AGC on the touch tone. The quick audio spike at the beginning of the tone triggered the AGC to reduce gain, which gradually recovers as evidenced by the tone's widening envelope. By contrast, the tone remains at constant amplitude in the top track.