Talk:NXDN

First Section
The first section has this statement which not true. it can be used by anyone with a license or contracted with a radio service. "exclusively for use by police, fire, ambulance and other public safety organizations" mws72 —Preceding undated comment added 21:44, 15 September 2019 (UTC)

Audio Quality
I'm disputing the neutrality of the claims made in this about the audio quality being "acceptable".

Previously there was a section regarding the audio quality being poor, with two citations to back this up.

This has since been replaced with a section describing the quality as good, with three citations to back this up.

However, I'm disputing the neutrality of this, and I believe these three citations are biased because:

-They do not reflect "real life" situations and are all laboratory tests. The original two citations against the quality was based on real-life experience.

-The three citations do not specifically relate to the intended use of NXDN. One relates to codecs in Tandem environments, and as such the quality scoring is altered to reflect this (there is a large difference in the scoring between unprocessed audio and PCM audio, which should be largely imperceptable, suggesting there is something amiss). Another relates to the codec in aeronautical environments (which isn't the intended use of NXDN). The third relates to a TIA document that you have to pay to view, so isn't even accessable.

I believe that the section has worthwhile merit, but may also perhaps be better served by including the previous two citations and making mention that some people dislike the quality, etc, etc.

Sam, Mansfield, UK (talk) 14:12, 26 January 2010 (UTC)

Real-life tests are often quite subjective and hard to reproduce. Lab tests are more reliable and can be more reliably extrapolated to alternate environments. Designers of higher-level protocols such as NXDN and APCO-25 select an encoder based on a wide variety of criteria, not just on codec technical qualities. The relative merits of the AMBE encoder against other codecs should certainly be discussed, including concerns about audio quality, but it would be more appropriate to address those issues on the AMBE page - Advanced_Multi-Band_Excitation. —Preceding unsigned comment added by Hwardsil (talk • contribs) 00:43, 20 March 2010 (UTC)

Agreed, but the dispute here is that there was a claim that some percieve the audio quality of NXDN as poor, with two citations relating to the particular codec, which back up this claim. This obviously is completely contrary to the current point of view presented.

The point of view has been changed to remove the perception by some that NXDN has a poor audio quality to one which states that the AMBE vocoder used has been tested and is fine.

Removing the AMBE vocoder from the equation, some still find that NXDN has poor audio quality, and that may be worthy of mention? Again, one of the original citations related specifically to the use of digital PMR radio.

Sam, Mansfield, UK (talk) 17:45, 3 May 2010 (UTC)

Deleted the reference to ETSI TS102 490 as this is a totally different air interface protocol and is not the same as NXDN.--Vharywolf (talk) 07:40, 27 May 2010 (UTC)

"...some still find that NXDN has poor audio quality..."

Without citing specific tests, this is just anecdotal. Certainly, everyone is entitled to an opinion and there are plenty of forums for such opinions to be vetted before they should be posted in a reference article. Returning to the subject at hand, however, I do not know of ANY codec that doesn't reduce quality of the reproduced speech, so there will always be "some" that find the audio quality not to their liking. If there have been controlled tests that establish specific deficits in the reproduced speech sufficient to cause the codec to under-perform in its intended application, they would certainly be worth posting. I emphasize "in its intended application" - LMR/PMR - and "controlled tests", discounting informal reviews.

The quality of the AMBE vocoder has by far the dominant effect on reproduced speech quality - there is extensive information on its performance which should be discussed on the AMBE pages. The NXDN air interface, if it is able to carry the AMBE-encoded data from transmitter to receiver, has little additional effect on speech quality. At the radio level, specific models could have a bad microphone design, insufficient audio output power, excessive distortion or noise, speaker design problems - any number of things that affect speech quality which have little to do with either the vocoder or the air interface. All of this has to be "teased apart" by a competent lab that can determine if a problem exists at all and where it exists within the transmit-receive chain. If the tests showing poor audio quality are available to the public, please post a link and I'll be happy to read them and respond.

"Deleted the reference to ETSI TS102 490..."

The intent was to offer it as a related standard and not a defining document. I agree that it is a different air interface. Leaving that reference out doesn't damage the article.

--TiredAndConfused (talk) 16:57, 11 October 2011 (UTC) START: The audio quality of the AMBE vocoder is adequate. It makes users sound like a robot holding its nose, but they do tend to get used to it. The audio of all the systems using this vocoder are very similar. We have a test rig here with MotTRBO and NXDN VHF repeaters running side by side and I can't tell the difference, although our end-users tell me they prefer the NXDN voice quality. If there's a difference, it's minor.

What we *have* noticed, though is that users MUST drive the audio circitry adequately or the vocoder may misunderstand what it's hearing and the output will certainly still sound like someone is talking but you won't understand it. I also have reservations about the ability of the recovered audio to be understood in noisy environments as it simply isn't "Toppy" enough. Most salesmen of these systems seem to believe that audio quality equals lack of analogue noise and Rayleigh fade effects, although that's not my understanding. Voice quality should equal intelligibility, which is very different. We use a Motorola 12.5kHz UHF analogue FM repeater to control the digital trials and the users consistently find that it outperforms the digital systems. Part of that is the urban terrain, which suits UHF, and part of it is that the human brain is still better at recovering audio than a vocoder. FINISH: TiredAndConfused (talk) 16:57, 11 October 2011 (UTC)