Hi Anthony,
 
Maybe I'm a little dense, but I don't see the connection between codec2 and
cost. At around $20 for a complete, working, proven hardware solution using
an apparently quite efficient codec algorythm, it is hard to see where the
cost savings would come from.
 
Let's assume you could replace the DVSI codec chip with a $5 generic
microprocessor and free software. You would reduce the manufacturing cost by
a maximum of $15, assuming all the interface required was free. That might
reduce the cost of an ID-880 from about $495 to maybe $475. Hardly
market-shattering news.
 
We don't hear anyone trying to build 'low cost', 'open design' RF amplifier
chips even though many are priced way above $20. Perhaps more comparable, we
don't see hams trying to invent an 'open' microprocessor instruction set to
escape the evil monopoly of folks like Intel, AMD, and Sun. 
 
In truth, the DVSI chip is a bargain considering the relatively small market
over which to amortise engineering and chip fab costs.
 
It seems to me that the uproar over the proprietary codec in DStar is really
greatly overblown. Someone found a way to piggyback on a successful codec
manufacturers existing technology to get a really low cost chip that hams
can use. Let's say thanks and then go to work on more and better ways to use
the chip we have - like the DVDongle and other boards. There is plenty of
space for experimentation and improvement in digital voice without needing
another codec.
 
Finally, if we had another codec, it would take about a month for a Codec3
crowd to organize to combat both DVSI and Codec2.
 
73
Ted
W1GRI

  _____  

From: [email protected] [mailto:[email protected]]
On Behalf Of a cutler22
Sent: Monday, June 07, 2010 09:28
To: [email protected]
Subject: Re: [DSTAR_DIGITAL] Codec2 development - open source vocoder


  


>From the sounds of it - this "Open Source DSTAR codec" is probably runner up
to "DSTAR 2.0", which would indeed be a fork situation of the protocol.
However, it would probably be called something different to avoid confusion.

ICOM's "flavor" of DSTAR is the *only* flavor of DSTAR out there- no other
major manufacturer has touched DSTAR. Even the Kenwood D-STAR radio is a
rebranded ICOM radio....

Of course, market share for DSTAR is ever growing, not declining and it
won't go away anytime soon. However, Codec2 sounds like the better
market-driven bottom-up protocol for next-generation Digital Voice that
could have a bigger market share than ICOM's DSTAR.

-73 de Anthony, KE7HQY

  _____  

From: Tony Langdon <[email protected]>
To: [email protected]
Sent: Mon, June 7, 2010 2:20:04 AM
Subject: Re: [DSTAR_DIGITAL] Codec2 development - open source vocoder

  

At 05:34 PM 6/7/2010, you wrote:

>And that D-STAR protocol definition is defined by the JARL...
>
>How much of a prospect is it the JARL would be willing to update the 
>DSTAR protocol with an open source vocoder if it were competitive to AMBE?

Assuming the JARL were open (pun not intended!) to the idea. The 
rest of this message deals with the case where the JARL to decide to 
amend the D-STAR protocol (if they don't, then we would be dealing 
with a fork instead).

The biggest challenge will be not breaking gear that's in the 
field. The DV Dongle is the easiest case, because the new (open) 
codec can simply be added to DVTool, and run entirely in software 
when needed, with the dongle only being used to encode/decode an AMBE
stream.

The Icom radios would require a series of replacement boards for 
their existing DV boards, to use the new codec, otherwise you're 
going to have a bunch of users who will find D-STAR suddenly becomes 
hit and miss. Also, the spec would need to be updated to include 
notifying the remote end which codec is in use, somewhere in the 
protocol, since this can no longer be assumed. There could be an 
opportunity in the new add-on board to have field programmable 
components on board, so any further enhancements to the codec/DSP can 
be downloaded and installed.

The above is the minimum that would have to be done. There might be 
other unintended consequences that need to be dealt with (I can see 
the possibility for all sorts of "corner cases").
>---------------
>Codec2 is in at least partial code form so far: 
><http://www.rowetel. <http://www.rowetel.com/ucasterisk/codec2.html>
com/ucasterisk/codec2.html>http://www.rowetel.
<http://www.rowetel.com/ucasterisk/codec2.html> com/ucasterisk/codec2.html

Looks promising. Where I see this gaining traction first is in HF 
digital voice experimentation, which hit a major speed bump when 
patent issues became apparent. However, codec2 will need to get into 
the 1200bps and below range to be really effective in this application.

>It seems to be competitive in the current market, a third-party 
>daughter board similar to UT-118 would have to be developed that had 
>dual AMBE/Codec2 capability, and sold at-cost to undercut the 
>current UT-118 cost. Assuming (warning - monster assumption!) the 
>SMT design was donated by hams and it was sold at cost, and it had 
>automatic codec detection and switching capability, it would be a 
>competitive way to bring an open source codec to "DSTAR".

And be able to be fitted without too much hassle/expense to existing 
equipment in the field (like my 91AD).

73 de VK3JED / VK3IRL
http://vkradio. <http://vkradio.com> com





Reply via email to