D-Star is being promoted as a new digital voice transmission mode "For the second century of Amateur Radio" - but does it really fit into the spirit of the hobby?

   D-Star is a voice over data protocol being targeted for Amateur radio service - the idea being to digitize the voice, vocode it to a low data rate signal, then transmit that over 2 meter FM to a second radio, where it is decoded and played back. The supposed advantages of this are:

  1. Longer range - the voice signal does not degrade as it does with narrowband FM as the signal gets weaker.
  2. Narrower bandwidth - the D-Star modulation fits within a 12.5 kHz channel spacing.
  3. Simultaneous voice and data - the system has a 1200 bps data channel in addition to the voice signal.
  4. High data rate (128kBit/sec) on the 1.2GHz band.
  5. No distortion of the voice as it goes over a repeater - you don't get the "cascaded c-weight filters" you can get in some repeater systems.

   These sound like great advantages - in fact, that is why more public service radio systems are going to digital protocols like Tetra and APCO-25. We hams are *supposed* to be pushing the boundaries of radio - isn't D-Star a good thing? Well, overall, I would say yes, it is.

   There's just one thing about D-Star that bothers me - the vocoder. The D-Star protocol is an "open standard" that is available to anybody to implement - but then again, so is APCO-25. However, if you want to build an APCO-25 system, you have to pay Digital Voice Systems, Inc. a license fee for the IMBE vocoder that is specified in the APCO standard - otherwise, you cannot convert the data to voice. Want to add APCO-25 to Gnu Radio? Tough - DVSI does NOT license their vocoder for operation on a PC - period. And even if they did, *somebody* would have to pay US$100K for the license, and it wouldn't be under a Free license.

   "Fine, but what does this have to do with D-Star?" Simple - D-Star uses the AMBE vocoder from DVSI, which is *also* not Free - so if you want your GnuRadio to do D-Star you are SOL.

   Now, when it comes to a commercial standard like Tetra or APCO-25 I don't have a (huge) problem with the vocoders being Not-Free. But when it comes to Amateur radio, if I cannot get my mitts on the code, then how am I supposed to:

  1. Build my own radio
  2. or
  3. Experiment with the the protocol

   Yes, I am NOT going to be able to build a modern HT-style radio in my workshop at home - but I *can* build a base rig, and tie it into my computer for decoding - except that I cannot get the AMBE vocoder to do so.

   My concern is that, should something like D-Star become commonplace, then we Hams will be little more than users of a commercial system, just like PMR users. Is that really what is best for our hobby? Is that what is best for our *duty* to society?

   Now, I could understand if there were no vocoders that could perform as well as AMBE - but I've run tests with Speex, and it can encode the audio just as intelligibly as IMBE at IMBE data rates (I know - I design P25 equipment for a living). I've even run the data rate down below the rates quoted for AMBE (2150 bps) and found the audio to be quite intelligible. So why did Icom choose AMBE over Speex? I have no idea, but I would guess it is because AMBE is used in P25 Phase 2 - so the Icom engineers already have experience with AMBE, and none with Speex.

   I'd really rather see things like D-Star based on vocoders we can get, rather than on vocoders we cannot. But that's just my opinion - not my employers.

73 de N0YKG



D-STAR compatible transceiver line-up

  • ID-1
  • IC-2820H and IC-2820
  • ID-800H (discontinued)
  • IC-91AD (discontinued) and IC-91A
  • IC-92AD
  • IC-V82 and IC-U82
  • IC-2200H
  • IC-80AD*

                                                                                                          D-Star Radio Club International