Raw IMBE Voice Bits

Status
Not open for further replies.

jlanfn

California Database Admin
Database Admin
Joined
Dec 31, 2006
Messages
366
Location
San Bernardino County
I have severely limited understanding of the internal components and processes of either the PSR-500 or PSR-600, so this might sound like a nonsensical question.

Could either the PSR-500 or 600 be simply modified (preferably through a firmware update from GRE) to dump the 7200 bps raw IMBE voice bits in real time as it scans a P-25 transmission instead of using the scanner's dsp to decode the audio and route it to the speaker?

Let me emphasize that I would want just the voice bits without any of the data bits.
 

AZScanner

Member
Joined
Dec 19, 2002
Messages
3,352
Location
Somewhere in this room. Right now, you're very col
I'm trying to understand why GRE would want to do that, especially just the voice bits. There's important info hidden in the data bits as well that would be useful - for example, if an encrypted message is encountered, you could direct the scanner to ignore it and resume scanning or searching before it even has a chance to route it to the speaker. If you strip out all but the voice bits, you'd lose that ability.

I'm a fairly advanced scanner user myself and even I don't think I'd have a use for just the voice bits - I'd be far more interested in the whole bitstream.

-AZ
 

AZScanner

Member
Joined
Dec 19, 2002
Messages
3,352
Location
Somewhere in this room. Right now, you're very col
Unless you wanted to use it for a ProVoice provider, I suppose.
Digital Voice Conversion Method - The RadioReference Wiki
Semi Off-Topic note here: I REALLY want to try this method out sometime. Sounds like it costs some serious buckage though... but if I ever get the spare money to do it, I really want to. Has anyone here done it and are they willing to describe their setup? I'd love to learn more.

-AZ
 

jlanfn

California Database Admin
Database Admin
Joined
Dec 31, 2006
Messages
366
Location
San Bernardino County
My (perhaps faulty) understanding is if a person wanted to use some other dsp to convert the voice bits into analog voice for listening, such as a board manufactured by DVSI that actually contains the IMBE vocoder, he would need the 7200 bps voice bits.

The purpose of not using the scanner's dsp would be to obtain significantly better-sounding audio quality from a dsp board that actually contains the IMBE vocoder.

Of course, as I have stated, my understanding of all this is severely limited, so I will defer to the experts.
 

AZScanner

Member
Joined
Dec 19, 2002
Messages
3,352
Location
Somewhere in this room. Right now, you're very col
My (perhaps faulty) understanding is if a person wanted to use some other dsp to convert the voice bits into analog voice for listening, such as a board manufactured by DVSI that actually contains the IMBE vocoder, he would need the 7200 bps voice bits.

The purpose of not using the scanner's dsp would be to obtain significantly better-sounding audio quality from a dsp board that actually contains the IMBE vocoder.

Of course, as I have stated, my understanding of all this is severely limited, so I will defer to the experts.
Um.... who do you think makes the vocoder in ANY digital scanner? Every digital scanner on the market has a DVSI chip in it. You can't get a P25 vocoder from anywhere else. Even the WRG305 winradio receiver uses DVSI software to recover the audio from P25 data in the plugin they sell for their Winradio software (I'd like to tinker with one of those sometime too - sigh, too many wants, not enough dollars).

I'd like the OP to offer a bit more explanation as to what they plan to do with said 7200bps voice data. It's not outside the realm of a guy like our pal Mr. Unitrunker to come up with a way to get at this data, but then what? That's what I'd like to know.

-AZ
 

DonS

Member
Joined
Jun 17, 2003
Messages
4,105
Location
Franktown, CO
Um.... who do you think makes the vocoder in ANY digital scanner? Every digital scanner on the market has a DVSI chip in it.
The latter part is not quite accurate. "DVSI chip" seems to imply hardware from DVSI, and I know that the GRE digital scanners don't contain any hardware manufactured, licensed, or sold by DVSI.

Software, of course, is an entirely different matter.
 

jlanfn

California Database Admin
Database Admin
Joined
Dec 31, 2006
Messages
366
Location
San Bernardino County
Just for kicks, I will try to explain the thought process that led up to my question.

Although it is generally accepted that GRE provides some of the best digital audio quality of digital scanners on the market, I am not particularly satisfied with it. Others on this forum have stated that digital audio quality on the radios professionals use approaches if not surpasses the audio quality of analog. That is certainly not the case with any scanner I have owned or heard. I have verified the disparity between professional radio and scanner performance through monitoring P-25 scanner audio and comparing that to scanner audio of a P-25 patch to analog that a local agency uses.

I wanted to know the technical reasons why this disparity exists, so I asked several questions in the General Scanning forum. None of the responses were quite as specific as what I was looking for, although I got the general idea that scanner audio will not be significantly improving anytime soon.

With many local agencies actively pursuing digital systems, I started thinking of possible solutions. Reading this file from the Wiki gave me the idea that perhaps I could use one of DVSI's products (like the VC-55-PR or the Net-2000-P25, provided there is a way for me to purchase one) connected to a scanner. If the scanner would output the 7200 bps voice bits, perhaps there was a way of feeding the bits into the vocoder to attain the audio. Of course, if scanners already use DVSI's vocoder, then I couldn't expect the resulting audio to be any better than that of a scanner.

I assume that the PSR-500 and 600 at some point in the digital signal processing have to obtain the 7200 bps voice bits before constructing the analog voice, so if I could extract the bits, a connection to a DVSI vocoder might work. I know that the voice bits can be attained through a discriminator tap, but I don't have the capacity to even begin to understand the steps necessary to accomplish this.

With so many uncertainties, I began to see that what I wanted to do was way over my head. Since I can't drop $20,000 for the method described in this Wiki page, I guess I will have to be content with what I have.
 

DonS

Member
Joined
Jun 17, 2003
Messages
4,105
Location
Franktown, CO
If the scanner would output the 7200 bps voice bits, perhaps there was a way of feeding the bits into the vocoder to attain the audio. Of course, if scanners already use DVSI's vocoder, then I couldn't expect the resulting audio to be any better than that of a scanner.
In the case of the GRE-made products, the scanners are already using the DVSI vocoder.

The only improvement would be if you could better reconstruct those "voice bits" from the received RF. That is, use error detection and correction to feed "'better' voice bits" to an IMBE vocoder. Your raw data source for this would likely have to come from a discriminator tap.
 

DonS

Member
Joined
Jun 17, 2003
Messages
4,105
Location
Franktown, CO
Others on this forum have stated that digital audio quality on the radios professionals use approaches if not surpasses the audio quality of analog. That is certainly not the case with any scanner I have owned or heard.
It's also not the case with the local agency (Santa Clara County Sheriff) that uses P25. Whenever I hear an unintelligible transmission on my PRO-96 or PSR-500, it's almost always followed by something like "10-9... you went digital". That is, if I couldn't understand it, neither could the users of the "professional radios".
 

AZScanner

Member
Joined
Dec 19, 2002
Messages
3,352
Location
Somewhere in this room. Right now, you're very col
CML does not publish prices. Call them and they should give you a quote.
OK, cool. I'll have to give that a try tomorrow.

@DonS - Really?! No DVSI chip? Interesting! Learn something new every day!

I still don't think the vocoder's the problem. The software does the same thing a chip would do. I think the problem is the quality of the signal being FED to the vocoder. The old adage of Garbage In, Garbage Out applies here. There's only so much error correction can compensate for, and there's simply no subsitute for a good clean signal to work with. I think the reason the dedicated PS radios sound so good is because you get what you pay for when you spend $5000 on a radio, and that includes the RF components, not just decoder chips and firmware.

It'd be an interesting experiment though. Hell, just the notion of possibly being able to turn any old scanner into a digital one for $115 is pretty tantalizing. I might even be tempted to try it.

-AZ
 

RKG

Member
Joined
May 23, 2005
Messages
1,087
For what (if anything) it may be worth: I have had the occasion on numerous occasions to listen to conventional P25 broadcasts on a PSR-500 and my XTS2500, side-by-side, and they sound exactly the same.
 

davidbond21

Member
Joined
May 18, 2005
Messages
530
Location
New Braunfels, TX
For what (if anything) it may be worth: I have had the occasion on numerous occasions to listen to conventional P25 broadcasts on a PSR-500 and my XTS2500, side-by-side, and they sound exactly the same.
That's been my experience, setting a Racal 25 next to my PSR-500 on some conventional P25 systems. The audio between the two is almost indistinguishable, except the radio will open up the audio quicker and pull in more distant signals(though that's to be expected).
 

Astrak

Member
Joined
Feb 17, 2005
Messages
1,573
Location
Mesa, AZ
That's been my experience, setting a Racal 25 next to my PSR-500 on some conventional P25 systems. The audio between the two is almost indistinguishable, except the radio will open up the audio quicker and pull in more distant signals(though that's to be expected).
I've had the opposite experience with a PRO106 sitting right next to an XTS, the 106 would start talking before the XTS and decode was the same and a little better while on the fringe of the P25 system coverage area.
 

sixtytwo

Member
Joined
Jan 11, 2004
Messages
465
I'm going to jump in here....

I've also been pretty fed up on the lack of decoding quality on P25 simulcast on all of the scanners I have (ummm, that's a lot, including all major brands). Non-simulcast transmissions are great, no issue at all - even at long distances and weak signal strengths.

Do we really know if the VC-55 board is the exact functional equivalent of the on-board decoding? I'm not at all sure that we're dealing with a lack of robust error correction here - that's why I went looking for alternatives today and found the VC-55 (before I found this thread).

I've got wonderful RF coverage where I am, but terrible audio if any at all. Of course with analog you'll hear the simulcast noise and put up with it, but on digital you're going to get higher BER - and you fight BER with FEC.

Or is DVSI the problem?
 
Status
Not open for further replies.
Top