Performance expectations when transitioning from analog to digital systems

[KevinC]

Yeah, let's all continue to operate on ancient 50 year old equipment that is hardly fit for hambone usage anymore. Give me a break.

We've had GTR's dead keyed as a control channel 24/7/365 for 11 years and counting. APX 7000's that were manufactured in 2010/2011 that are still dead balls on frequency operating on TDMA-exclusive systems, which requires a high degree of accuracy as one might imagine.

In conclusion, modern digital equipment that is performing as good as the day it was put into service a full decade later with no tuning or other maintenance required or performed. Buy good equipment, get good results; cut corners and buy junk, wind up spending more in the long run.

I still use my APX7000 from 2011 daily. I'm not super rough on it, but it's been in temps from 8 to 116 degrees, flown all across the country, come flying out of my backpack and skidded across concrete a couple of times and still works just fine.

Not sure flying all across the country means anything, just thought I'd throw that in.

 

[ElroyJetson]

You understand that my comment was facetious in nature...but still pointing out some truths. We often get rid of the old, simply because it is old, not because it has lost its utility. I don't have that wasteful mindset. I keep cars and maintain them, which always costs less in the long run than buying a new one every three years. Things that do their job well with no or little attention required stay in service. My PC will likely remain in its place and serving its purpose until it is so out of date that it has been rendered obsolete to the point of being non functional by "upgrades" to software that didn't NEED to be upgraded, except as a means of forcing people to retire older but working computers.
(HP pulled that stunt on me by refusing to offer Windows 10 compatible software for my older HP printer that worked much better than the newer printer that replaced it....yet the newer printer's software and drivers do work under Windows 10. Still, the newer printer is quite BAD. )

 

[GTR8000]

Okay.

 

[Ubbe]

If you you were buying a system and the vendor's sales team showed this to you, what would you ask them to provide in the way of proof?

Let them demonstrate using two portables that can be programmed with low power, 0,1W and have one channel in analog and another in digital and try them at a distance where the coverage starts to be sporadic. Have one person walk around while talking and going between the two channels. Analog will always win.

Those diagram that show digital works better at lower signal strength assume that it is a steady non moving signal. In real life it's a fading signal going rapidly between low signal a no signal and when listening in analog you only loose some syllables but you understand everything but in digital it takes a while to resync when the signal are lost and it will be either totally quiet or just some occasional burst of digital Donald Duck audio. A digital system always needs an average higher signal strength to work properly.

If a new digital system uses the same antenna system and single receiver configuration as the old analog then it will be a worse audio quality and a worse RF coverage. What usually happens are that a customer informs about the issues with the old analog system and the sales team of the digital system then makes a total overhaul of the antenna system that includes higher gain antennas and directional antennas and dual or triple diversity receivers and possible a few extra sites. If those actions where done to the old analog system it would have solved all issues.

But what you get with a digital system are that the data stream can not only carry audio information but also additional info like text and system info and also encryption keys and control over that only valid users are on the system. Users that switch from analog to digital will always complaint about the bad audio quality. An analog system have a good audio processing as the modulation limiter at transmit and the analog FM demodulator in a receiver will keep the level more constant and you will also hear low modulation audio without any degradation in quality. In digital systems the audio level are much more critical to be able to use all of its limited dynamic range or it will either overload leaving a square wave type of sound or too low level and there's only a few bits of resolution left in the audio, you get the audio to expand more in the dynamic range while an analog system compress audio to make it sound more even in level and are more easy to listen to.

/Ubbe

 

[MTS2000des]

Total B.S.

Our system is miles above the ancient 1989 era Smartnet .89x it replaced. The GTR8000 and GCM8000s are amazing and blow away the old combo of MSF5000/SpectraTAC/AstroTAC stuff. I can login to the comparator and watch portables on the other side of our diverse landwise county ping RFSS. The audio stream of recovered audio is perfect. Subscribers are constantly moving, on portables inside vehicles. Rarely do I hear any BER. Usually it's one or two known locations (like the WalMart in Union City) that are problematic.

Compared to the old system where you could constantly hear voter "swaps" and Rayleigh fading on every other transmission. Static pops and clicks from the dying MSF receivers made painful experience for our dispatchers.

Analog FM can sound great when perfect conditions are met. Rare is that ever the case. P25 overcomes noise well. I can regularly have usable comms on my standalone VHF Quantar at -115dbm and still get a usable DAQ of 3.4. On FM, you're listening to hash and trash trying to pull out words.

 

[GTR8000]

Total B.S.

💯

I often get the sense that some people have spent way too much time in a classroom/library, and not nearly enough time in the field/real world. 🥴

 

[xmo]

Considering that the topic of this discussion is: Performance expectations when transitioning from analog to digital systems, and given that future interested parties may locate this discussion with an internet search, I think is would be advantageous to clear up a few points.

First, the digital systems discussed in the context of North America are commercial and public safety systems based on formats such as DMR and P25 rather than other formats such as Tetra or D-Star (which has an air interface performance in faded environments that is substantially inferior to any other digital format)

Second, the use of the term "audio quality" leads to a lot of confusion. Audio quality is intrinsically subjective. In the context of land mobile two-way radio - and especially in the context of the discussion topic: "Performance expectations when transitioning from analog to digital systems", it is crucial to understand that audio "quality" is not based on subjective measures such as: likeability, fidelity, pleasantness, or naturalness of speech.

In the world of land mobile two-way radio, delivered audio quality (aka DAQ) is defined by the TIA and expressed as a number on a scale of 1 to 5 where the definitions of the ratings are exclusively based on speech intelligibility - how well speech can be understood.

After all, that is the fundamental purpose of the system - to convey speech from one person to another in separate physical locations.

There have been some excellent observations posted here based on actual experience by knowledgeable members. These discussions also inevitably receive responses based on opinions, often worded as fact, but without citation of credibe references. These should be considered strictly as that - opinion.

I have considerable interest and experience in the subject, both in the lab and in the field, and I will attempt to respond to some points of interest and answer any questions people may have.

 

[xmo]

Ubbe posted a number of opinions to which MTS2000des took exception, as do I. Let's look at a few unsubstantiated claims and misleading opinions.

"In real life it's a fading signal going rapidly between low signal a no signal and when listening in analog you only loose some syllables but you understand everything but in digital it takes a while to resync when the signal are lost and it will be either totally quiet or just some occasional burst of digital Donald Duck audio."

This sounds exactly like a comparison of analog to D-Star for which it is totally accurate. However, this is not the case when comparing analog to P-25. Certainly, a faded signal environment will poke holes in either format, but unlike D-Star which has very poor resync performance, P25 is comparable to analog.

Re-entry or late entry for both analog and P25 is specified in TIA standards documents. Analog CSQ or basic P-25 must transition from start of reception to useful audio in less than 200 milliseconds. Analog CTCSS/CDCSS or talkgroup P-25 must do so in less than 400 milliseconds.

Another factor for consideration in comparison of analog to digital in faded environments is that hams or 'radio guys' will back off the squelch control so they miss less during dropouts. Commercial and public safety users typically don't do this so for them, audio holes due to sinal drop out are just as bad with either format.

Ubbe also said: "A digital system always needs an average higher signal strength to work properly."

In fact, the opposite is true. The TIA publishes documents used by engineers and communications professionals when designing systems to meet customer requirements. These documents, based on extensive lab and field tests as well as user experience, show that to meet a public safety grade delivered audio quality of 3.4 or better in real-world faded multipath environments, wideband analog needs roughly 5 dB stronger signals and narrowband analog requires 8 dB stronger signals.

 

[Ubbe]

This is a test from a Youtuber that isn't a fair comparison, not using the same radio and antenna for the analog to digital comparison but using different radios that have different antennas. But I have gotten the same results from numerous tests when using the same hardware in both analog and digital. When the RF signal are strong all systems work fine but digital always have some voice quality degradation due to the reduce of the number of databits over air to make the signal fit in the bandwidth. The difference are when a radio are moving, like in a vehicle or a person moving around while transmitting, like they do in 90% of transmissions.

/Ubbe

 

[Project25_MASTR]

This is a test from a Youtuber that isn't a fair comparison, not using the same radio and antenna for the analog to digital comparison but using different radios that have different antennas. But I have gotten the same results from numerous tests when using the same hardware in both analog and digital. When the RF signal are strong all systems work fine but digital always have some voice quality degradation due to the reduce of the number of databits over air to make the signal fit in the bandwidth. The difference are when a radio are moving, like in a vehicle or a person moving around while transmitting, like they do in 90% of transmissions.

/Ubbe

Still a very subjective/opinionated test for various reasons.

• Video fails to explain if the analog is testing 5 kHz deviation (wideband) or 2.5 kHz deviation and judging by the UK model numbers and British accent pretty safe to assume 5 kHz deviation. Keep in mind, we had a mandate to narrowband in the US by January 1, 2013 for most VHF/UHF commercial services. The reality was, significant performance degradation for analog operations.
• Video doesn't provide any info on the alignment of the radios. DMR specifically is not very tolerant to frequency error exceeding 100 Hz (as anyone who has ever serviced first gen TRBO radios in 800 MHz radio can tell you) where 400 Hz of error on wideband analog is still totally usable (and within 0.5 ppm at 851 MHz).
• Not knowing firmware revisions is problematic as there have been significant DSP advances introduced in firmware updates over the last two decades. Compare a Motorola Astro Saber (introduced in 1996) with R07.xx.xx (last major release in 2008) firmware to a Motorola XTS5000 (introduced in 2004) with R20.50.07 (last major firmware release in 2016) to a Motorola APX6000 (introduced in 2012 but still in production) with R29.00.00 (which was released less than a year ago). Vastly different results.

 

[GTR8000]

This is a test from a Youtuber that isn't a fair comparison

"Test"? That's about as far from an objective test as you can get. It's a casual observation at best, and none of the radios are receiving public safety grade P25 transmissions which is the gist of this thread, not DMR or dPMR. You've done nothing to reinforce your original claims by posting that video, except maybe to further weaken your assertions by presenting unscientific "evidence" that analog is superior.

 

[KevinC]

"Test"? That's about as far from an objective test as you can get. It's a casual observation at best, and none of the radios are receiving public safety grade P25 transmissions which is the gist of this thread, not DMR or dPMR. You've done nothing to reinforce your original claims by posting that video, except maybe to further weaken your assertions by presenting unscientific "evidence" that analog is superior.

Pretty sure I've posted this before, but here goes...

I've implemented numerous mixed-mode (analog and P25) systems. I always have the customer compare analog and P25 in weak coverage areas, P25 always wins. Same subscriber, same repeater/coax/filtering/antenna, only difference is changing channels in the subscriber.

Oh yeah, my BCT15 works just fine wirelessly (without any developer Apps/tablets/phones/whatever). Sorry, couldn't resist.

 

[Echo4Thirty]

Pretty sure I've posted this before, but here goes...

I've implemented numerous mixed-mode (analog and P25) systems. I always have the customer compare analog and P25 in weak coverage areas, P25 always wins. Same subscriber, same repeater/coax/filtering/antenna, only difference is changing channels in the subscriber.

Oh yeah, my BCT15 works just fine wirelessly (without any developer Apps/tablets/phones/whatever). Sorry, couldn't resist.

Way back in the day, I ran a mixed mode simulcast TRS. Digital ALWAYS outperformed analog using the same radios. The only literal difference was the modulation and TGID hex. Some instances the analog user could bring up the system but not hold it then switch to a digital system and have just enough BER to not only keep the system, but pass audio, robotic VOCODER is guessing its tail off audio, but the message was received at the console.

The important thing is one cannot lump any analog vs digital debate into something so binary as that. Even in analog what bandwidth/devation? PL? DPL? CSQ? ALL of those can affect decode range. Then you get to the many many variations of digital, some of which are horrible and some are better than analog. Then we get to what receiver are you using? One cannot expect the same performance on a couple of hundred dollar consumer receiver vs a multi thousand dollar one. I used to take complaints from scanner users back in the day telling me my system sounded like crap on their uniden/radio shack wizbang digital jobber scanner. I would bring them an XTS3000 which was our worst sounding digital subscriber at the time and it would blow the doors off their scanner. Cant compare an APX to an SDS.

"When we went from Analog TV to Digital, the signal sucked worse!" ok yeah ATSC blows, but do not extend that to AMBE P25.

 

[xmo]

Ubbe, please do not consider the comments in this discussion as a personal attack. It is a discussion board and you are certainly entitled to your opinions. That said, a number of your comments are not expressed as opinion but are worded as statements of fact.

A statement of fact must be:
1) Common knowledge
2) Easily verified - e.g. by an internet search
3) Supported by citation of a credible reference.
-or-
Alternatively, a statement of fact can reference original research by the author in which case said author must have available evidence of the use of accepted methods and accompanied by sufficient details and data to enable other researchers to duplicate the process.

None of these are offered support your statement:

"If a new digital system uses the same antenna system and single receiver configuration as the old analog then it will be a worse audio quality and a worse RF coverage."

KevinC and Echo4Thirty each gave examples of experience with mixed-mode systems. Kevin stated: "I always have the customer compare analog and P25 in weak coverage areas, P25 always wins. Same subscriber, same repeater/coax/filtering/antenna, only difference is changing channels in the subscriber"

I'll add my experience to theirs. I was directly involved with a mixed-mode Motorola Smartzone simulcast system from initially working with the design engineer and proposal team through installation, optimization, turn up, and years of support including numerous upgrades.

The digital audio intelligibility was consistent across the entire system service area with no degradation due to noisy signals, picket fencing multipath flutter, or simulcast overlap artifacts.

This experience confirmed that the system exceeded RFP requirements. It also easily passed the coverage acceptance test.

The modern coverage acceptance tests are hard evidence of the exceptional performance of the digital format. A computer controlled process acquires multiple samples of signal bit error rate by a drive test through a statistically determined grid overlay using a calibrated example of the actual subscriber device employed on the system.

The data acquisition of each sample is initiated (and logged) based on GPS location. The samples are acquired over a specific time based on vehicle speed and a number of wavelengths in accordance with Lee's method. The BER samples are then mapped to DAQ through an industry standardized, format specific, BER to DAQ equivalence.

RSSI data can be acquired concurrently by the same process so that signal intensity could be mapped for comparison to the coverage prediction (heat map). The RSSI data could also be mapped to analog DAQ, however the complexity and cost of peforming these tests means that they are typically only included in sales of large systems and today no one is selling or buying large analog systems.

 

[MTS2000des]

I'm sorry, but a video of simple non-scientific radio hobbyist simplex using low tier CCRs and an XPR of unknown origin is not even close to anything of value. Real testing of coverage on modern LMR systems is software driven. When we did our acceptance tests on our P25 simulcast system, the vendor had a "Voyager" kit consisting of a tuned and aligned subscriber with appropriate attenuators connected to magnet mounted quarter wave antennas, simulating an actual subscriber as specified in the RFP response. The Voyager consisted of a computer that read raw BER test pattern put up over the air and recorded every test point on the map. DAQ testing consisted of 4 experienced dispatchers on dispatch consoles using headsets and Harvard phrases used to test DAQ. Everything captured by our AIS based logging recorder.

That's real testing using industry standard methodologies. Hobbyists often forget that "just because it sounds like crap on a scanner" doesn't mean anything from an actual end user standpoint. Comparing results using cheap Chinese junk radios and simplex is like comparing a Big Wheel to a Ducati.

 

[krokus]

At my work site, we made the change from a high band VHF FM repeater (with PL) to two time slot MotoTrbo, on the same freqs. This was replacement of just the repeater and the handhelds. (PR-860 to XPR-7550.)

There has been a significant improvement in coverage, as some areas that are in RF shadow now have useful signals. We also have full site coverage, when we have to use the radios in LOS mode, which we did not with the prior radios.

 

[xmo]

I don't have a problem regarding the test video as long as it is recognized for what it is - an informal test conducted by a person who was curious and wanted to see a comparison. Lacking the resources available to a communications professional, he used what might be called the MacGyver method. In other words: "do what you can with what you have where you are"

There are a couple of valid take-aways. The video shows that analog audio gets progressively noisier and cruddier as signals get weaker and it shows that something is seriously wrong with the dPMR pair. Based on TIA data for signal level to BER/DAQ, dPMR performance (as a format) should be similar to DMR.

 

[KE4ZNR]

I'm sorry, but a video of simple non-scientific radio hobbyist simplex using low tier CCRs and an XPR of unknown origin is not even close to anything of value. Real testing of coverage on modern LMR systems is software driven. When we did our acceptance tests on our P25 simulcast system, the vendor had a "Voyager" kit consisting of a tuned and aligned subscriber with appropriate attenuators connected to magnet mounted quarter wave antennas, simulating an actual subscriber as specified in the RFP response. The Voyager consisted of a computer that read raw BER test pattern put up over the air and recorded every test point on the map. DAQ testing consisted of 4 experienced dispatchers on dispatch consoles using headsets and Harvard phrases used to test DAQ. Everything captured by our AIS based logging recorder.

That's real testing using industry standard methodologies. Hobbyists often forget that "just because it sounds like crap on a scanner" doesn't mean anything from an actual end user standpoint. Comparing results using cheap Chinese junk radios and simplex is like comparing a Big Wheel to a Ducati.

Yep. I was here when we switched from a mixed analog/digital 3600 to straight P25 4 site 800Mhz simulcast system and while some old timers whined at first ("It sounds different") now I would have a revolt if I tried to switch users back to analog. Our users LOVE P25 digital.

 

[GTR8000]

One important point to make that might be lost in all of this...P25 voice sounds noticeably different than analog voice, and takes some getting used to. Unfortunately that is far too often mischaracterized as "P25 is not as good as analog!" The fact that the voices you're familiar with sound different when digitized by the vocoder does not equate to P25 being inferior to analog. In fact one significant benefit of P25 is that it's full quieting even on the fringes of coverage. As long as there is enough signal to correctly relay the 0's and 1's, you're full quieting whether the signal is -40 dBm or -100 dBm. Yes, it can experience intermittent bit error at the extreme fringes of coverage, but an analog signal of the same weakness would be so far down in the noise as to be unintelligible and mostly just white noise.

When music was first digitized en masse in the 80s, everyone complained that the CD was woefully inferior to an analog album. To this date there are still "purists" who argue that point vociferously (sound familiar?), however the overwhelming majority recognize that digitizing media (music, photos, video, etc.) was an inevitable and necessary evolution. RF is no different.

 

[KevinC]

One important point to make that might be lost in all of this...P25 voice sounds noticeably different than analog voice, and takes some getting used to. Unfortunately that is far too often mischaracterized as "P25 is not as good as analog!" The fact that the voices you're familiar with sound different when digitized by the vocoder does not equate to P25 being inferior to analog. In fact one significant benefit of P25 is that it's full quieting even on the fringes of coverage. As long as there is enough signal to correctly relay the 0's and 1's, you're full quieting whether the signal is -40 dBm or -100 dBm. Yes, it can experience intermittent bit error at the extreme fringes of coverage, but an analog signal of the same weakness would be so far down in the noise as to be unintelligible and mostly just white noise.

When music was first digitized en masse in the 80s, everyone complained that the CD was woefully inferior to an analog album. To this date there are still "purists" who argue that point vociferously (sound familiar?), however the overwhelming majority recognize that digitizing media (music, photos, video, etc.) was an inevitable and necessary evolution. RF is no different.

2.4 dB above the noise floor and P25 works. Try that with analog. 😉