Audio Quality: Dstar vs. P25 vs. DMR

[N9NRA]

When you hear P25 phase 1 on an actual subscriber radio, setup correctly, with current DSP and software, you will be ashamed of how lousy the D-star radio sounds. I think it is more the fault of the typical "cheap ham toy" radio design with a vocoder stuck on the end of the audio chain.

This of course, if your local P25 system is optimized and not some poorly maintained POS. We have one of those in my area- it's amazing no one has lost their life because of it either.

I get what ya are saying here, and yes, we have a system that behaves like what ya describe audio quality-wise, mainly due to problems with the link equipment at the sites. HOWEVER...in some cases DSTAR will be better than the other two, depends on the environment ya are listening to it in. I had a chance to demo (and do some comparisons between the three modes) some DMR radios at Dayton a year ago, and personally i found the DMR units to have a "smoother" sound quality, on the other hand DSTAR by comparison sounded kinda "rough" when heard next to the DMR radio. This may be `cause the throughput of the data is faster, i know that DSTAR is slow (sorry i don`t recall just what the throughput speed is right off), and P25 is the same (throughput speed wise) as DMR, so that has something to do with it too. N9NRA

 

[N8OHU]

I get what ya are saying here, and yes, we have a system that behaves like what ya describe audio quality-wise, mainly due to problems with the link equipment at the sites. HOWEVER...in some cases DSTAR will be better than the other two, depends on the environment ya are listening to it in. I had a chance to demo (and do some comparisons between the three modes) some DMR radios at Dayton a year ago, and personally i found the DMR units to have a "smoother" sound quality, on the other hand DSTAR by comparison sounded kinda "rough" when heard next to the DMR radio. This may be `cause the throughput of the data is faster, i know that DSTAR is slow (sorry i don`t recall just what the throughput speed is right off), and P25 is the same (throughput speed wise) as DMR, so that has something to do with it too. N9NRA

Narrower bandwidth for D-Star; 6.25kHz versus 12.5kHz for DMR and P25. Also, there are differences in the versions of the AMBE codec between the three systems as well.

 

[xmo]

If you are thinking that D-Star's poor audio performance is because of its spectrum efficiency (narrow bandwidth) - that theory doesn't fit reality because D-Star isn't really a 6.25 kHz format - it doesn't fit in the FCC 6.25 emission mask. It's more like 10 kHz wide - almost as wide as P25 Phase 1 CAI.

Look up the tests done by many hams that have posted their results on the internet - the Utah VHF Society is one example.

NXDN on the other hand, really is 6.25 and does fit in the 6.25 emission mask whereas DMR achieves 6.25 equivalency by fitting two simultaneous talkpaths into a single 12.5 kHz bandwidth.

 

[kayn1n32008]

Wirelessly posted (Mozilla/5.0 (BlackBerry; U; BlackBerry 9780; en-US) AppleWebKit/534.8+ (KHTML, like Gecko) Version/6.0.0.600 Mobile Safari/534.8+)

DStar is 4800bps. 2400bps for voice, 1200bps for FEC and 1200bps for lowspeed data/signalling. I believe DStar uses a 12.5KHz channel not 6.25KHz.

 

[N8OHU]

Wirelessly posted (Mozilla/5.0 (BlackBerry; U; BlackBerry 9780; en-US) AppleWebKit/534.8+ (KHTML, like Gecko) Version/6.0.0.600 Mobile Safari/534.8+)

DStar is 4800bps. 2400bps for voice, 1200bps for FEC and 1200bps for lowspeed data/signalling. I believe DStar uses a 12.5KHz channel not 6.25KHz.

Nope, but radios before the ID-31 and ID-51 don't have narrow enough filters to handle it properly. That's why people that built hotspots that worked with the older radios had troubles when using the ID-31; the narrower filtering meant that they had to learn how to properly adjust them to work with both kinds of radios.

 

[kayn1n32008]

Wirelessly posted (Mozilla/5.0 (iPad; CPU OS 6_1_2 like Mac OS X) AppleWebKit/536.26 (KHTML, like Gecko) Version/6.0 Mobile/10B146 Safari/8536.25)

Wirelessly posted (Mozilla/5.0 (BlackBerry; U; BlackBerry 9780; en-US) AppleWebKit/534.8+ (KHTML, like Gecko) Version/6.0.0.600 Mobile Safari/534.8+)

DStar is 4800bps. 2400bps for voice, 1200bps for FEC and 1200bps for lowspeed data/signalling. I believe DStar uses a 12.5KHz channel not 6.25KHz.

Nope, but radios before the ID-31 and ID-51 don't have narrow enough filters to handle it properly. That's why people that built hotspots that worked with the older radios had troubles when using the ID-31; the narrower filtering meant that they had to learn how to properly adjust them to work with both kinds of radios.

Oops slightly under 1000bps for data on V/UHF. Not sure what you are saying nope to tho. Care to explain ?

 

[N8OHU]

Oops slightly under 1000bps for data on V/UHF. Not sure what you are saying nope to tho. Care to explain ?

It's not 12.5kHz, although the earlier radios were likely using the 12.5kHz filters for D-Star. this has changed with the newest Icom D-Star transceivers.

 

[xmo]

So what are you really saying here anyway - that older Icom D-Star radios don't work very well with newer Icom D-Star radios?

The only things a narrower IF filter should impact are receiver signal to noise (sensitivity) and adjacent channel rejection although the earlier radios might be more tolerant of incorrectly adjusted (i.e. too wide) do-it-yourself repeaters, hot spots etc.

That's just another fundamental D-Star weakness - the lack of any published standard for testing and alignment and the complete lack of any available instrumentation to test D-Star radios' performance in digital mode.

 

[fineshot1]

When you hear P25 phase 1 on an actual subscriber radio, setup correctly, with current DSP and software, you will be ashamed of how lousy the D-star radio sounds. I think it is more the fault of the typical "cheap ham toy" radio design with a vocoder stuck on the end of the audio chain.

This of course, if your local P25 system is optimized and not some poorly maintained POS. We have one of those in my area- it's amazing no one has lost their life because of it either.

I use both on a regular basis and i think they sound the same.

They both have the same basic BER qualities once the receiver starts
getting noise into the data stream. Noise is the enemy of any digital
voice data as it distorts the perceived reception. If you are in a strong
receive signal area this is usually negated. If not the noise will hamper
reception as the BER rate goes up.

 

[N8OHU]

So what are you really saying here anyway - that older Icom D-Star radios don't work very well with newer Icom D-Star radios?

No.

The only things a narrower IF filter should impact are receiver signal to noise (sensitivity) and adjacent channel rejection although the earlier radios might be more tolerant of incorrectly adjusted (i.e. too wide) do-it-yourself repeaters, hot spots etc.

Indeed, and that was mostly what I was meaning. We also need to understand that even the Icom D-Star Repeaters don't properly support the intended bandwidth, for many of the same reasons.

That's just another fundamental D-Star weakness - the lack of any published standard for testing and alignment and the complete lack of any available instrumentation to test D-Star radios' performance in digital mode.

I would hope that any competent modern radio repair service would have the equipment and manuals to service any radio sent to them; as for home-brewed stuff, most of the proper settings have been worked out from experimentation, and the same equipment that can be used to set up a normal analog FM Repeater can be used to set the required deviation on a home-brew repeater.

 

[xmo]

Deviation is one thing - digital modulation fidelity is quite another.

If that home-brew repeater - or a factory made one - has poor frequency response, poor linearity, poor modulation compensation or any other kind of distortion of the digital payload - the deviation can be set to exactly the right value and the performance will still be poor - and there is no way to know why without a defined testing methodology and corresponding test equipment.

All the commercial digital formats have specific test methods incorporated into their respective standards documents - D-Star doesn't.

Likewise - test equipment is available to evaluate each of the commercial formats IN DIGITAL MODE. None exists for D-Star.

 

[DaveNF2G]

I agree with this, but I don't think amateur radio operators should follow this. We can't talk on their frequencies and they can't talk on ours. So the question of a compatible digital mode really doesn't matter.

The desire for compatibility arises from hams' legendary cheapness.

We want few radios that do many things at once. We want to buy surplus instead of new.
:twisted:

To answer another comment that did not quote through:

Everyone except hams (and GMRS) are affected by the narrowbanding mandate.

 

[n1ip]

hmm

I shudder to even get in this discussion...

Let's be clear about a couple of things... D-Star is NOT 6.25khz... It runs in a 12.5khz spaced channel and uses approx 10khz of bandwidth. It does NOT run in a 6.25khz channel and use 6.25khz.

Let's also get this baud rate discussion out of the way. Dstar does NOT use the FEC allocation of baud for VOICE FEC, it's only used in the high speed data usage. Why do you think that R2D2 exists, if it used FEC it would not have that problem.

NXDN - there are 2 modes. a 12.5khz version that only Kenwood supports, and a 6.25khz version that both Icom and Kenwood support. The 12.5khz version is 9600 baud with FEC included, and the 6.25 is 4800baud also with FEC included.... You'll never hear R2D2 on NXDN for example.

There is simply no comparison to the audio quality of NXDN vs. Dstar, NXDN wins hands down. Same is actually true of NXDN vs. DMR as well.

I have 6 NXDN repeaters, 2 Hytera DMR repeaters and a Trbo DMR repeater and of all of them I *MUCH* prefer NXDN.

Today, we have over 25 NXDN repeaters all linked up on the internet, both a mix of Kenwood and Icom. Its a simple process to get another repeater on the network and the cost advantage of NXDN is huge.... Where else can you buy a fully operational 25W 100% duty cycle repeater that will run in both analog and digital modes for around the $600 mark if you shop... And for a site that might want to upgrade all the other equipment that you have is usable... duplexer, antenna, amp if you have one, etc.

Anyway, I mostly lurk here, but saw this discussion and had to chime in...

my .02,
Alan

 

[newsphotog]

Everyone except hams (and GMRS) are affected by the narrowbanding mandate.

No... everyone except hams, GMRS, and anyone outside of 150-174 MHz and outside 421-512 MHz is affected by the narrowbanding mandate. Below 150 MHz is not affected.

 

[N8OHU]

Deviation is one thing - digital modulation fidelity is quite another.

If that home-brew repeater - or a factory made one - has poor frequency response, poor linearity, poor modulation compensation or any other kind of distortion of the digital payload - the deviation can be set to exactly the right value and the performance will still be poor - and there is no way to know why without a defined testing methodology and corresponding test equipment.

All the commercial digital formats have specific test methods incorporated into their respective standards documents - D-Star doesn't.

Likewise - test equipment is available to evaluate each of the commercial formats IN DIGITAL MODE. None exists for D-Star.

Can you read Japanese, or do you work for Icom? If the answer is no to either of these questions, then how do you know that there aren't test methods in the specifications for D-Star, since the only English translation that I know of does not contain the entire text of the JARL specification. And obviously the test equipment does exist, or so many of the existing digital communication systems that are built around GMSK/GFSK wouldn't be serviceable.

 

[N8OHU]

I shudder to even get in this discussion...

Let's be clear about a couple of things... D-Star is NOT 6.25khz... It runs in a 12.5khz spaced channel and uses approx 10khz of bandwidth. It does NOT run in a 6.25khz channel and use 6.25khz.

Let's also get this baud rate discussion out of the way. Dstar does NOT use the FEC allocation of baud for VOICE FEC, it's only used in the high speed data usage. Why do you think that R2D2 exists, if it used FEC it would not have that problem.

While I know the D-Star specification can differ from the reality and likely does for the older equipment, it does call for the system to be 6.25kHz OR LESS; the actual wording in the translated document indicates the target bandwidth was actually less than 6kHz. As for why there is R2D2, that's pretty easy and it doesn't matter if there is FEC or not on the DV Data (which the specification does call for); frequency being off on the radios from what the display shows, poor quality components used in the repeater (known to have occurred, by the way), data packet corruption in the controller, etc.

 

[JRayfield]

GMSK/GFSK are modulation types.
TDMA/FDMA are channel access methods.

A 'standard', such as D-Star or DMR, is a 'suite', describing the details of the standard, such as the type of modulation and the channel access method that is used. It also describes the 'protocol', the 'format' of the data that is sent and received over the air (typically referred to as the Common Air Interface).

So, there's much more involved with a 'standard' than just the type of modulation. For test equipment to work with a particular 'standard', that test equipment must therefore be designed specifically for that standard. That's why service monitors are available with options specifically for P25 Phase 1, P25 Phase 2, DMR, and NexEdge. So just because a piece of test equipment exists that works with a particular modulation type, does not mean that it will work with any 'standard' that uses that modulation type.

John Rayfield, Jr. CETma

Can you read Japanese, or do you work for Icom? If the answer is no to either of these questions, then how do you know that there aren't test methods in the specifications for D-Star, since the only English translation that I know of does not contain the entire text of the JARL specification. And obviously the test equipment does exist, or so many of the existing digital communication systems that are built around GMSK/GFSK wouldn't be serviceable.

Sent from my GT-P7310 using Tapatalk HD

 

[xmo]

N8OHU, please don't be concerned that we're trying to change your mind here - we realize that a true believer is never bothered by silly little matters like FACTS.

Unfortunately for amateur radio, the FACTS are that D-Star has issues - such as the FACT that D-Star doesn't fit in a 6.25 channel - it never has and never will regardless of new radio, old radio, home-brew radio - whatever; or the FACT that D-Star doesn't work very well at the air interface level.

We're just pointing out these FACTS to those who haven't yet drunk the Kool-Aid; to those who are willing to do a little reading of the considerable body of technical research that has already been conducted on the matter - such as that of the Utah VHF Society.

As far as a testing protocol or availability of test equipment, you should know by now how the internet works: regardless of the subject, if there's no verifiable reference - it doesn't exist!

So - prove me wrong - cite an available English translation of a testing protocol - a method of connecting test instrumentation to the radios for proof of performance in the digital mode. Such specifications are a part of all real digital formats. Provide a reference to a single instrument, available for purchase, that incorporates the D-Star format. None exists, whereas they do exist for P25. DMR, and NXDN.

Actually, Icom should have published a testing document along with the introduction of their D-Star line and they should also have offered a low cost testing adapter that would interface traditional RF instrumentation to a PC running Icom provided digital format testing software. The fact that they did not further insures that D-Star will continue to limp along with poor performance that encourages hams to switch to robust commercial formats that do perform well at the RF level.

 

[beischel]

Audio

The received DStar audio on my ID880 sounds just like the P25 and DMR audio I have heard. Very clear.

The comments about the R2D2 audio is when the transmitting station is on the fringe. FM would just break up and P25 and DMR would just drop out from our recent experimentation with the different digital modulation types.

Have not tried NXDN yet, but hope to soon.

 

[beischel]

N8OHU, please don't be concerned that we're trying to change your mind here - we realize that a true believer is never bothered by silly little matters like FACTS.

Unfortunately for amateur radio, the FACTS are that D-Star has issues - such as the FACT that D-Star doesn't fit in a 6.25 channel - it never has and never will regardless of new radio, old radio, home-brew radio - whatever; or the FACT that D-Star doesn't work very well at the air interface level.

We're just pointing out these FACTS to those who haven't yet drunk the Kool-Aid; to those who are willing to do a little reading of the considerable body of technical research that has already been conducted on the matter - such as that of the Utah VHF Society.

As far as a testing protocol or availability of test equipment, you should know by now how the internet works: regardless of the subject, if there's no verifiable reference - it doesn't exist!

So - prove me wrong - cite an available English translation of a testing protocol - a method of connecting test instrumentation to the radios for proof of performance in the digital mode. Such specifications are a part of all real digital formats. Provide a reference to a single instrument, available for purchase, that incorporates the D-Star format. None exists, whereas they do exist for P25. DMR, and NXDN.

Actually, Icom should have published a testing document along with the introduction of their D-Star line and they should also have offered a low cost testing adapter that would interface traditional RF instrumentation to a PC running Icom provided digital format testing software. The fact that they did not further insures that D-Star will continue to limp along with poor performance that encourages hams to switch to robust commercial formats that do perform well at the RF level.

DStar absolutely fits in a 6.25Khz channel. That's a FACT. Not sure where you got your belief that it does not, but you are not correctly informed.

Some earlier DStar radios possibly could not tune in 6.25 increments, but they could be memory programmed to store the newer channels. I have four DStar radios and have not found a repeater that could not be programmed into memory.

As for testing. Really who cares. You have to be the first individual in three years (since I have been using DStar who brought up that issue. It apparently is not an issue with the vast majority of hams.

DStar is THE standard for amateur digital radio at this point whether you like it or not. Everything else is a distanced minority. That is a FACT as you like to point out. Sure you can play for P25, DMR, and all the other commercial stuff that does not work as well in the amateur radio market, but it still is a commercial standard with commercial technology for a commercial market. Amateur radio does not have talk groups, radio IDs. We have call signs and multiple radios.

So if you actually want to talk to someone, get a DStar capable radio. Tens of thousands of users. Close to 2,000 repeaters worldwide. Get some other digital commercial radio, and you'll hear the sound of crickets, but only if standing outside.