Batboard

Anyone have any experience with this issue? I'm part of a 700 P25 implementation and we have discovered that when the Scott SCBA Vibra Alert activates your audio is no longer able to be understood. It's as if the vibra alert nullifies your voice. Simplex and duplex.

You are not trying to use IMBE for fireground comms are you?
You just found out why it is not recommended.

Chris
N9LLO

I know this horse has been beaten several times, but just what's so bad about P25 in public safety or fireground ops?

There's a fire service north of the GTA (Barrie) that uses P25 for their fireground ops, and I've not noticed any severe issues. A lot of it is training. I know it's not recommended and personally I think it's a bad idea. But, why is it that the experts advise against P25 for fireground communications? Is it just because it has a tendency to have other sounds drown out your voice?

It's because it's a lot easier to destoy a P25 CAI signal to a point where it's no longer recognisable by a receiving radio.

If someone gets into serious trouble on a call, and talks over the top of someone else on the radio on simplex, chances are, someone will hear them. With a digital system, if they walk over someone else, neither signal will be readable by any radio.

As I said the issue is duplex and simplex, on and off the trunk system. Using XTS2500. The question is not if we should use fg ops on or off the trunk but if anyone else has experienced issues communicating when the Scott SCBA is in vibra alert.

Thanks!

What happens with the vibra alert comes on? Do you completely lose comms, or get broken audio?

Nexrad16 wrote:Anyone have any experience with this issue? I'm part of a 700 P25 implementation and we have discovered that when the Scott SCBA Vibra Alert activates your audio is no longer able to be understood. It's as if the vibra alert nullifies your voice. Simplex and duplex.

Heh, you can't understand half the transmisson in analog when that trips nevermind in digital.

FWIW, have you tried it with the mask voice module? That seems to bypass most of the vibra noise (at last in analog, haven't played digital with it yet).

bellersley wrote:I know this horse has been beaten several times...why is it that the experts advise against P25 for fireground communications?...

Basically, because P25 offers no clear benefits over "plain vanilla" FM simplex in a fireground scenario, while exhibiting several odd (dangerous) failure modes such as the one described by the OP.

The problem of unreadable audio has to do with the way the XTS, Astro radios process the TX audio. NO way arround it.......


VOCoders just fail trying to process continious tones, beeps, ect.

The fire service that bellersley speaks of in Barrie is P25 VHF, and it's on 141.xxxx & 141.xxxx

Would it make a difference if Nexrad16's fire service was on VHF??

Okay, got more info and it is the vocoder as someone noted. The issue exits with or without the use of the voice amplifier that Scott makes.

That being said can we expect that an EF Johnson radio might work?

Thanks to everyone for your valued input.

The DVSI IMBE vocoder is a basic part of the P25 (Phase I) standard. The IMBE vocoder was designed to do one thing and one thing only- transmit human voice. Background noise, continuous tones, etc. will drive it nuts. You will see the same problems with any P25 radio. The issue is not the audio circuitry on the radio, it is the encoding used to convert and compress the signal to and from the digital domain.

When you transmit analog audio on a 12.5KHz channel width, you have the full channel for the audio signal, and minimal compression is required to achieve good fidelity. (A 25KHz channel is better still, and there's nothing quite like having 60KHz, except for possibly having 100KHz ) However, if you want to put a P25 voice signal into a 12.5KHz channel you have to support the voice signal data, plus error correction data, plus associated signaling data, and other data overhead. That means that the vocoder has to really compress the voice data, using a very limited set of codec entries to reproduce it. (Admittedly, I'm simplifying a little here).

Wowbagger posted a great example in http://batboard.batlabs.com/viewtopic.php?t=62528 of what happens when you ask P25 to process a "dirty" voice signal (in this case a harmony vocal track). The results, even under laboratory conditions, make the point pretty clearly.

Thanks! This is very interesting. I'm somewhat confused why /\/\ never mentioned this knowing fire was going to use the system.

I can't help but question corp. America. We actually have reached a business practice in which money means more than the lives of the people who will committ, under the most risky environments, to rescue them and their families? This sort of omission puts everyone at risk, the public, police officers, EMS personnel, and firefighters.

Sad...

Anyway, thanks to all who provided a response.

We were wondering about this very issue in the thread that HumHead referenced in his post. At the time when that thread was started, I didn't have any P25 equipment to play with. Now I have access to some gear, and when I have some free time I may do some testing with various PASS alarms to see how the different alarm tones affect the P25 vocoder. If you have any audio clips you could post or send, that would be great. Here's a quick overview of the testing I'd like to do:

1) Have someone speak a predefined sentence into a radio in analog mode with no background noise. Record the received audio.
2) Repeat (1) in digital P25 mode
3) Repeat in analog with a PASS alarm sounding in the background (one the waist of the person with the radio)
4) Repeat (3) in digital P25 mode
5) Repeat (3) and (4) with different PASS alarms

My feeling is that until this background noise issue is addressed at the vocoder level (may be never), we should probably stick to analog for fireground communications.

Andy

I guess this issue has made its way up to the IAFC, and they're requesting data from the field. If you have data, it would be a good idea to send it to them:

http://www.iafc.org/displaycommon.cfm?a ... clenbr=719

Andy

Yes. That came from some of the testing done here in Boise.

What is going to happen when there are no more analog radios available and ALL audio regardless of who's radio is processed by vocoders?

Should not someone be working on different ways to make the PASS alarms compatible? Like changing the sound the alarms make?

com501 wrote:What is going to happen when there are no more analog radios available and ALL audio regardless of who's radio is processed by vocoders?
Should not someone be working on different ways to make the PASS alarms compatible? Like changing the sound the alarms make?

When that time comes, there will be no more fire

The 'Scott' pass alarm has had no effect during our testing.

The PASS and an open bypass doesn't affect the audio nearly as much as the Vibralert. Be prepared to hold your microphone at arms length with a voice amp. This isn't due to the effect of a public safety microphone feeding back either (which we disposed of for anyone with a voice amp long ago). Engine noise at the pump panel, rescue saws , and operating handlines interior also generate considerable artifact. Personally I consider the results thus far to be best case with existing equipment and I have endorsed my results as such. There is no real test that can duplicate the noise on the fireground like the fireground. Digital is great for dispatch, unit movement, and first aid calls but not ready for the fireground.

Agreed. We have done much testing using the saws, fans etc. all with miserable results and all forwarded to IAFC & soon the IAFF.

Huston FD just experienced a 'close call'. Short story is that those involved made it because they were able to communicate. I question what may have happened if they were using the P25 Digital standard. Some follow-up this week with HFD.

Be safe!

Next $500 accessory.... the throat mics for firefighting...

IAFC wrote:
Digital Radio Test Protocol

Download these instructions (pdf)

The following is a set of instructions to test a digital portable radio for interference in the presence of common fireground noises:

Test A:

1. Use a portable radio on a digital channel. Please identify the radio type.
2. One firefighter will operate the power tools listed below at full throttle while another firefighter stands directly next to the one operating the power tool with the portable radio (approximately 3 feet between the radio and the device, apparatus and/or tool).
3. A third firefighter, in a remote location, will monitor the radio transmission.
4. Once the power tool is operating at full throttle, the firefighter with the portable radio will repeat, “MAYDAY, MAYDAY, MAYDAY.”
5. The firefighter monitoring radio traffic will document the audio quality for each test as:

I – intelligible, able to understand the voice audio message
U – unintelligible, unable to understand the voice audio message

Record the mode as:

D – digital
A – analog

Use the following power tools, devices or apparatus as appropriate. We recommend you test as many sources as possible:

K12 circular type saw
Chain Saw
PASS Device
High Apparatus engine idle

If your department uses Scott SCBA, perform the test while the vibra-alert is activated in low air mode (in a safe environment).

Test B:

Repeat Test A using the same radio in the analog mode.

[rant]
This retarded test actually made it to the IAFC, which cripples my faith in them. Using the above procedure, can you answer the following?:

Is it the VOCODER or digital signal itself?
Is it simplex or through a crappy coverage system?
Is it DSP revision issue? (makes a huge difference!)
Is it AGC settings? Are they set per manufacturer specifications?

Didn't think you'd be able to answer any of them. All of the above are crucial to figuring out a problem. I'm not saying there ISN'T a problem, only your test proves nothing.

Way to add confusion to an already hot topic....

[/rant]

tpg

com501 wrote:What is going to happen when there are no more analog radios available and ALL audio regardless of who's radio is processed by vocoders?

This will not happen... not as long as you can buy 10 analog radios for the price of 1 digital radio anyway

Yes I too believe some of the testing criteria is a bit non specific. In my testing I made all transmissions in a simplex mode to ensure that it is not a system related issue. My AGC settings are per the manufacturers recommendation, and my FW/ DSP revision is the most recent. From what I can tell it is just limitations of the technology operating in a high noise environment.



[rant]
This retarded test actually made it to the IAFC, which cripples my faith in them. Using the above procedure, can you answer the following?:

Is it the VOCODER or digital signal itself?
Is it simplex or through a crappy coverage system?
Is it DSP revision issue? (makes a huge difference!)
Is it AGC settings? Are they set per manufacturer specifications?

Didn't think you'd be able to answer any of them. All of the above are crucial to figuring out a problem. I'm not saying there ISN'T a problem, only your test proves nothing.

Way to add confusion to an already hot topic....

[/rant]

tpg

The Pager Geek wrote:Is it the VOCODER or digital signal itself?

It's the vocoder. It's a known fact that running a constant tone or set of tones will make a perceptual codec (IMBE, AMBE, GSM, Speex, etc.) rather unhappy. Todd covered this fairly well below.

Good to see you alive again, TPG.

Digital voice as a technology is fine, but the APCO Project 25 implementation is not for high noise environments, thanks to the IMBE vocoder. Digital radio is great on the battlefield, which poses similar if not more severe auditory noise challenges to communications as firefighting. It's used under gunfire, aircraft noise (such as that on board a helicopter with the doors open), wind and engine noise from small boats and armoured vehicles, etc.

Who's silly enough to say that you have to use 700MHz P25 for absolutely everything that could require a radio, anyway? Take a leaf out of the book of armed forces, who most definitely don't say that any one of their radio systems is a perfect fit for every situation. They use the right tool for the right job, and the same should apply for civilian public safety communications. In this case, you use your fancy 700MHz P25 systems for dispatch, strategic traffic and whatnot, and use plain FM on one of the lower bands on the fireground.

I was a member of the largest fire service in the world. Although they're planning to move a significant portion of their comms FM to P25 within the next few years, they will retain their analog FM on UHF and VHF for fireground communications.

To try and quantify some of the comments on here so far,
the vocoder that is being used in the P25 systems has a
problem with human voice and high background noise.
It doesn't matter what the high background noise is
being generated from, the vocoder now being used
can't handle it.

A number of people and departments have been
complaining to the largest supplier of the P25 equipment
and they have not come up with any solutions so far.

This is one of the reasons that NFPA has always taken
the stand of recommending that fireground communications
be done on simplex channels and in the analog mode.
You don't have a vocoder to distort the audio path in
an analog transmission.

Try and picture it in a slightly different way. In a digital
transmission the vocoder takes the clear analog audio
and converts it into a digital stream. At the distant end,
the radio receiver is using a similar vocoder and trying
to re generate the digital stream and make it back into
clear audio.

If the vocoder that was designed for normal audio and
looking for the normal gaps in the human vocie tries
to function with all this noise, it just doesn't know what
to do. All the vocoder is seeing is this constant high
level of trash. So it encodes this high level of trash into
mostly high levels. At the distant end this distorted
audio has more trash added as the vocoder doesn't
know what to do with all the high level noise. If you
even had a slim chance of getting anything inteligable
out the far end, it too was probably distorted even
more in the conversion to an analog sound.

Take a look at the vocoder system that is used in an
EDACS network. It doesn't have these issues. It seems
to handle the high level noise and is able to translate
the digital stream back into a usable human voice.

Don't start slinging arrows at me for supporting one
system over another. I am just using this as one
type of digital system over another. I am one of the
first people to cry foul over a Pro Voice system as
locking out any competition. You can only get your
radio from one source. There is no bid competition
to keep prices in check.

Jim

A separate communications platform would work but this in reality is impractical. When a municipality / agency / company purchases new equipment they expect that this equipment is going to perform the role that they purchased it for. The whole concept of an additional expense to perform a role that the equipment purchased was to perform is borderline absurd don't you think ?. How would you feel if you bought a car and you went to drive it on Saturday and it wouldn't start. You tow your vehicle to the dealer and they inform you that you need to purchase a weekend capable car in order to drive the vehicle on the weekend. You ask "well why didn't you tell me that this isn't a weekend capable car" ? The dealer says "well you didn't ask". or worse yet "we didn't know when we sold it to you". I am of the opinion that firefighting is seen as a sideshow market by many vendors and that equipment is marketed as being practical without truly taking into consideration the operational environment that thier products are being used in. What is needed in firefighting is a radio that will deliver speech that is understood when the user presses the push to talk in an austere environment. Nothing fancy is needed. Newer systems and capabilities of these systems are geared toward marketing gee whiz technology at a the expense of the original mission of the equipment it is replacing. Technology is great but it should at the least live up to the standard of what is replaced. Yes these problems existed with legacy FM and I'd be a fool if I was to tell you otherwise. I can tell you that in my testing the transmissions were not completely indecipherable but were bad enough that it will require firefighters to repeat transmissions and this further complicates operations. There is a solution and it will be found - but first you have to admit there is a problem. I just wished that the vendor and P25's proponents would have put as much time and expense into researching the facets of its employment as they did in it's marketing.

Nexrad16 wrote:Houston FD just experienced a 'close call'. Short story is that those involved made it because they were able to communicate. I question what may have happened if they were using the P25 Digital standard. Some follow-up this week with HFD.

You're talking about the North Loop highrise fire that happened last Wednesday, right?

HFD has their own set of issues, with bad simulcast audio coming from out-of-synch analog simulcast transmitters. They do, however, have all 8 of their repeatered channels programmed for simplex talkaround and use it frequently for highrise ops (but not in this case, however.)

Fun fact: they also use XTS3500 portables flashed for IMBE in analog mode.

I'm alive... this "woke" me from the dead...

Here's what gets me:
ANY radio using a vocoder processes the voice the same for analog and digital.

When speaking digital:
Analog voice gets converted to digital via DSP, then sent to receiving radio where it is decoded back to analog.

When speaking analog:
Analog voice gets converted to digital via DSP, then converted back to analog, then sent to receiving radio and the process is reversed.

(Motorola radios, I'm not sure on off brand stuff.)

A PR1500 analog radio uses a vocon and DSP to convert the audio to digital, then back to analog, then transmits it. Just because what you hear from your scanner, doesn't mean it's not a DSP processed signal.

The true test would be PR1500 vs HT1000 or something non-vocoder.

tpg

TPG- you're correct. A PR1500 uses a vocoder and runs the signal A/D/A. However, the digital signal inside the radio remains at full bandwidth. As such, a PR1500 (or XTS5000) operating in analog mode doesn't have a too-significant issue with the Vibra Alert (no more significant than a Saber or an MT2000 would).

The issue comes when the IMBE/AMBE/etc. codec is applied. IMBE is a "perceptual" codec... it uses the properties of human speech to "cheat" and compress audio that's fairly large in the beginning down to something that's very small. IMBE works in largely the same way, say, MP3 does... they program the codec with a very clear understanding of what it's expected to compress. When you present the codec with something it's not expecting (like a steady tone or even a composite tone), Bad Things happen. Try sending a 1KHz test tone down an Astro in digital and see what it sounds like on the other end. Even better, something that's slowly changing, like a slow string of DTMF. Blech.

So, yes, I somewhat misspoke earlier. The vocon itself is not to blame. The codec is the issue. Unfortunately, there isn't a codec that can deal with these problems. The solution would be to attempt to clean up the wideband signal in the vocoder before you funnel it off to the codec (try to strip off the continuous tones). However, that's going to require some significant horsepower and smarts as well.

Also, PM me when you get a minute...

tvsjr wrote:However, the digital signal inside the radio remains at full bandwidth. As such, a PR1500 (or XTS5000) operating in analog mode doesn't have a too-significant issue with the Vibra Alert (no more significant than a Saber or an MT2000 would).

The issue comes when the IMBE/AMBE/etc. codec is applied. IMBE is a "perceptual" codec... it uses the properties of human speech to "cheat" and compress audio that's fairly large in the beginning down to something that's very small. IMBE works in largely the same way, say, MP3 does...

Exactly. The purpose of a parametric codec is to reduce the amount of required channel bandwidth. If bandwidth wasn't an issue, it wouldn't make sense to run it through the codec. Digitize it, sure, but why compress it if bandwidth isn't an issue? Trying to fit 10 pounds of stuff in a 5 pound bag requires compression. The mp3 analogy is a good one. In radios that have internal DSP, the "full bandwidth" signal inside the radio is akin to a .wav audio file. However, when the signal is run through the codec, information is lost (just like in an mp3 file) in order to reduce the required bandwidth. The coding algorithms are not lossless, like the familiar .zip file in computing.

Here is a question I'd like to pose to the panel at the FCC's first responder summit this month. I've heard that they'll be taking questions by email, and if they do I'll submit this question, but I'll pose it here to the group also:

The FCC is justifiably concerned with spectral efficiency of communication technologies, and just last month adopted a Third Report and Order urging the transition to 6.25 kHz bandwidth technologies in private land-mobile radio systems. Coincidentally, also last month the International Association of Fire Chiefs released an alert to its members regarding a potential problem with digital voice communications in a fire-ground environment due to background noise interference with the digital vocoder. Since in most cases, adopting spectrally efficient technologies requires migrating from analog systems to digital systems for voice communications, my question is: Should public safety (and life-safety in particular) communications systems be held to the same spectral efficiency standards as commercial systems? In other words, should it be acceptable to use technologies that are spectrally inefficient if they are deemed to be more robust than their spectrally efficient alternatives? If so, how should the FCC support this through regulation?

Andy

I'd add some "Is the FCC going to force users to utilize an under developed technology that may jeopordize life safety of the rescuers in addition to those that required the services to begin with? Will the FCC accept the partial blame if the loss of a life comes as a result of this decision?"

Something to that effect, with more P.C. terms....

tpg

The Pager Geek wrote: [rant]
This retarded test actually made it to the IAFC, which cripples my faith in them. Using the above procedure, can you answer the following?:

Is it the VOCODER or digital signal itself?
Is it simplex or through a crappy coverage system?
Is it DSP revision issue? (makes a huge difference!)
Is it AGC settings? Are they set per manufacturer specifications?

Didn't think you'd be able to answer any of them. All of the above are crucial to figuring out a problem. I'm not saying there ISN'T a problem, only your test proves nothing.

Way to add confusion to an already hot topic....

[/rant]

tpg

This is more of a political test than a rigorous technical evaluation. All they need is some good audio clips to start raising hell with everyone they can think of to blame. It's up to the engineers to do the real technical tests, and offer solutions, if necessary. The results of the IAFC tests will just be a stick for beating with, and will probably be quite effective for that purpose. Hopefully it will cause the standards bodies to do a little bit more homework before finalizing things.

Andy

This is great! Everyone is providing some valuable input. Agree, it's the P25 vocoder and not Motorola. We've tested EF Johnson and soon Tait. We expect the same or nearly the same results but we'll testem.

I would like the group to know that this was not a fire department push to go 700. It was the counties push and we were just lucky or unlucky to get a VERY LARGE grant to purchase the XTS2500 radios. All but 26 of them remain boxed until there is a solution or political pressure knocks us out of being able to make the decision to not convert. Something that is occuring as I type.

Anyway, keep the thoughts coming.

A simple short-term solution would be to use the radios in analog simplex mode for fireground ops...you've stated before that's not an option, but I don't see why. The enroute, dispatch, command etc. can all still be on the new whiz-bang system, but you can still keep fireground ops analog simplex while still using those new radios...they will do analog too.

Andy

aaknitt wrote:A simple short-term solution would be to use the radios in analog simplex mode for fireground ops...you've stated before that's not an option, but I don't see why. The enroute, dispatch, command etc. can all still be on the new whiz-bang system, but you can still keep fireground ops analog simplex while still using those new radios...they will do analog too.

Andy

Exactly. It's a strong recommendation (and will likely be codified soon) that fireground comms take place on an analog, simplex, unencrypted, no-frills channel. This is the safest configuration for fire personnel on scene, and it just plain works better.

The 2500s are dual-band 700/800. Perhaps you could use one of the 800 NPSPAC channels in low-power simplex for the time being as an interoperable fireground channel?

Mr. Syntax had a point. The Military has used digital communications on different bands.30-88 MHz, 150-180 MHz, and 220-400 MHz, with few audio problems in VERY high noise environments. The big difference I see is the audio pick devises, Microphones.
The military uses noise cancellation mics. One element is pointed to the operator, while the other is pointed and wired 180 degrees out of phase from the operator. Noise is cancelled at the microphone. They have been doing this for thirty years.

The microphones used on most land mobile radios do NOT have good noise cancellation characteristics.
I would say solve the problem up front first. Use GOOD noise cancellation mics for your test then work your way up the signal path.

mm

I've always used simplex on any fireground op I was on. The thought of using a TRS (analog or digital) is not very heart warming to me.

tvsjr wrote:Exactly. It's a strong recommendation (and will likely be codified soon) that fireground comms take place on an analog, simplex, unencrypted, no-frills channel. This is the safest configuration for fire personnel on scene, and it just plain works better.

I am intentionally taking this somewhat out of context-- but digital comms really have no place in many, if not most, of the current deployments. The degradation of basic two-way operations like signal quality and transmission continuity more than outweigh the fringe benefits of going digital.

I am sure there will be many that disagree, and maybe this is for another thread... my 2cp nonetheless.

The question was asked why we don't accept the system/design as is and work on simplex as a short term solution. Simple, once we accept the system as is we're stuck. There is little doubt that the vendor/s will no longer work to solve the issue.

As I said earlier, to work on simplex is not an option. We have a vhf system that works fine as is simplex/duplex inter op w/blm & usfs. So, if we are to work on simplex then we gained nothing and are not motivated to spend your tax dollars on a system that offers us less and cost us more.

I know of several agencies that operate 99.9% of the time on the system for everything. They do not go to simplex. Salt Lake area, Dayton. Oh., Mont. Co. Oh., Huber heights Oh. and some 20 other agencies within Montgomery County, just to name a few. They've been operating this way for better than 10 years (personal experience) with great success and have no plan to change procedures that I've been made aware of.

Could one argue that when a police officer encounters a bad guy that they should switch to simplex so that they don't have a communications failure? Essentially that's what we're saying, dangerous environments should communicate on simplex. I doubt that any officer would consider this as an option while pulling their weapon for some jukie criminal.

I really don't care to debate the duplex, simplex, on the system off the system as the decision is made (if we can't be on the system then the system offers us nothing and we don't care to waste the tax payers money on "less for more". If fact it offers us far less, no inter op with feds.).

I do appreciate all the friendly discussion. This is great and educational.

Thanks everyone!

Nexrad16 wrote:
Could one argue that when a police officer encounters a bad guy that they should switch to simplex so that they don't have a communications failure? Essentially that's what we're saying, dangerous environments should communicate on simplex. I doubt that any officer would consider this as an option while pulling their weapon for some jukie criminal.

Maybe the guy with the gun would worry about the bad guy, and not the radio. I know I do

Nexrad16 wrote: I really don't care to debate the duplex, simplex, on the system off the system as the decision is made (if we can't be on the system then the system offers us nothing and we don't care to waste the tax payers money on "less for more". If fact it offers us far less, no inter op with feds.).

If indeed the system offers you less, why use it?

The gun thing, agreed.

"If indeed the system offers you less, why use it?" That's our point & position. 3000.00 vs half that and we have inter-op with the feds.

The charge goes on.

Nexrad - you're missing the point. You *can't* solve this problem. It's not a defect with a radio... it's a problem inherent with the use of a perceptual voice codec.

You are endangering the lives of your firefighters by using P25 on the fireground - no two ways about that. You need to drop back to analog simplex for fireground comms. If you need to interop with the feds, you need to assign someone at a command post to carry a fireground radio and a radio on the system.

Never said it was a defect and don't believe it's a defect. It is poor planning and a lack of real world testing. I clearly recognize that the equip. is operating as designed.

We are not endangering our firefighters as we remain on vhf. We will remain there as long as the system design fails to provide for the safety of our personnel.

Our charge is for change in the design. If that cannot occur then we believe the fire service should know about it so that when they sit at a table and listen to a 'pitch' they will know the right questions to ask. We will continue to rally our troops, fire, police & ems to promote change.

Thanks for your comment. Sorry I can no longer comment on operational fire ground policy & procedure (simplex/duplex/trunked & not trunked).

tvsjr wrote:So, yes, I somewhat misspoke earlier. The vocon itself is not to blame. The codec is the issue. Unfortunately, there isn't a codec that can deal with these problems. The solution would be to attempt to clean up the wideband signal in the vocoder before you funnel it off to the codec (try to strip off the continuous tones). However, that's going to require some significant horsepower and smarts as well.

High rate CVSD with a bit of extra cleaning sounds beautiful from people hanging out the side of helicopters, so it's certainly not an impossible task, though it may well be for IMBE.

Microwave Mike wrote:Mr. Syntax had a point. The Military has used digital communications on different bands.30-88 MHz, 150-180 MHz, and 220-400 MHz, with few audio problems in VERY high noise environments. The big difference I see is the audio pick devises, Microphones.
The military uses noise cancellation mics. One element is pointed to the operator, while the other is pointed and wired 180 degrees out of phase from the operator. Noise is cancelled at the microphone. They have been doing this for thirty years.

The microphones used on most land mobile radios do NOT have good noise cancellation characteristics.
I would say solve the problem up front first. Use GOOD noise cancellation mics for your test then work your way up the signal path.

mm

A soldier's fairly basic headset is $700 for a reason.

About a year ago I used a reasonably cheap eBay throat mic with an XTS 5000R UHF in IMBE, from a floodboat fanging along at top speed. Audio from an NMN6191B "noise cancelling" speaker mic was completely unreadable in IMBE, and marginal in FM. From the throat mic, I was told I actually came through clearer in IMBE than I did in FM, thanks to the vocoder's tendancy to ignore low level background sounds. And that throat mic is a piece of crapola compared to a good military one.

mr.syntrx wrote:High rate CVSD with a bit of extra cleaning sounds beautiful from people hanging out the side of helicopters, so it's certainly not an impossible task, though it may well be for IMBE.

CVSD is merely a lossy codec... it doesn't have the perceptual encoding "smarts" that IMBE does. Any perceptual codec is more than likely going to have the same issue, so you either ignore the background noise at the mic (throat/ear mic) or you clean it up in DSP (adaptive filtering) before you dump it into the codec.

Just don't expect that to happen any time soon with P25.

Nexrad16 wrote:Could one argue that when a police officer encounters a bad guy that they should switch to simplex so that they don't have a communications failure? Essentially that's what we're saying, dangerous environments should communicate on simplex. I doubt that any officer would consider this as an option while pulling their weapon for some jukie criminal.

I realize you can't comment on the simplex/repeater/etc. debate, but I'd like to point out that police and fire ops are two different situations. Police officers operate semi-autonomously, and are often far away from other units and their disptach center, which is why they are on repeaters 99% of the time.

In fire response, however, it's never a single person responding to an incident. Depending on SOPs, in most cases the interior firefighters don't need to talk back to the dispatcher like the police officer would. They only need to be able to talk back to the local Incident Commander, who is on scene. In most cases, the most reliable way to do that is on a local simplex (digital or analog) channel. This is more reliable in this situation because there is one less point of failure (no repeater) in the communications link. I say in most cases because in some situations (high rises, subways, etc.) communications infrastructure may be needed to communicate with interior firefighters. For a police officer, the more reliable communications path is the repeater, since in most cases he is out of simplex range of those he needs to talk to. Yes, the repeater is a possible point of failure, but using it is still more reliable than simplex. It's very possible to do a full mathematical probability analysis to determine what the most reliable communications path is for any given situation.

A good analogy in law enforcement would be a special repsonse team...most of those teams use simplex for their tactical communications.

The goal should be to use the communications system/method that accomplishes the goal and has the least number of potential points of failure. Just because you have a lot of communications infrastructure doesn't mean it needs to be used for every type of communications need. The simplex tactical channels can be considered part of a good overall communications "system", even if they do not use fixed infrastructure.


Andy