[drats_users] ERROR Correction in slow speed data

[John Davis]

Daniel:
That is incorrect.  There is definitely FEC on the voice.  Check the JARL spec.

R2D2 is normally happens in low signal conditions. In Icom's implementation of the vocoder, once the signal gets so low that the AMBE chip cannot faithfully recreate the voice packets, Icom had the choice to either just mute the voice or allow a little of it through and hope you get some intelligibility before it gets too low.

Forward Error Correction adds extra bits in a complex algorithm when transmitted that can reconstruct a packet when a certain number of bits are missing on the receive end.  That allows a digital signal to still be reconstructed at a lower signal level than if there was no FEC.  FEC is kind of like parity bits on steroids!

Some people using other digital modes criticize D-STAR for having the R2D2.  You will notice that when R2D2 is present, you can sometimes pick out another word or two that you would miss if they chose to mute it at loss of packets.

John WB4QDX

From: drats_users-bounces at intrepid.danplanet.com [mailto:drats_users-bounces at intrepid.danplanet.com] On Behalf Of Daniel Metcalf
Sent: Friday, September 20, 2013 8:21 PM
To: Discussion of D-RATS
Subject: Re: [drats_users] ERROR Correction in slow speed data

I don't think is does. That is why when you radio loses sync you get R2D2 and others remain able to receive. If I read the protocol documents correctly it is all fire and forget much like the VOIP protocols mentioned before.

On Fri, Sep 20, 2013 at 6:23 PM, Joel Black <w4jbb at charter.net<mailto:w4jbb at charter.net>> wrote:
Okay, this goes against what little I have learned about VoIP traffic.

I didn't think there was any FEC with voice.  That's why voice has priority on data networks that use VoIP.  Otherwise, if a data packet has priority, it would bump a voice packet.  When a voice packet is bumped, you lose bits of it resulting in choppy audio.

I have experienced this when some of our first quasi VoIP locations came online.  Voice is live traffic.  Once it is said, it goes.  There's nothing to re-transmit.

Without getting into the engineering (which will leave me drooling and nodding off to sleep), how does D-STAR use FEC on voice?

73,
Joel - W4JBB

On 9/20/13 3:57 PM, John Davis wrote:
That is true. The FEC is on the voice with 2400 bits for voice and 1200 for FEC.

John WB4QDX

From: drats_users-bounces at intrepid.danplanet.com<mailto:drats_users-bounces at intrepid.danplanet.com> [mailto:drats_users-bounces at intrepid.danplanet.com] On Behalf Of Daniel Metcalf
Sent: Friday, September 20, 2013 4:55 PM
To: Discussion of D-RATS
Subject: Re: [drats_users] ERROR Correction in slow speed data


Agreed. From what I have read d-star is a lot like the UDP network protocol where it is transmit and forget there is no handshaking / error correction built in. It is up to the application developer to handle error correction and the d-rats package does perform some level of error correction.

Dan Metcalf - KB3UUN.
On Sep 20, 2013 3:54 PM, "John Davis" <jdavis at gtworks.com<mailto:jdavis at gtworks.com>> wrote:
My understanding is that the low speed data channel does NOT have error correction, however, D-RATS offers Error correction in Message and File Transfer modes where it is most important.

John WB4QDX

From: drats_users-bounces at intrepid.danplanet.com<mailto:drats_users-bounces at intrepid.danplanet.com> [mailto:drats_users-bounces at intrepid.danplanet.com<mailto:drats_users-bounces at intrepid.danplanet.com>] On Behalf Of Dave Welker
Sent: Friday, September 20, 2013 3:19 PM
To: D-Rats Mailing List
Subject: [drats_users] ERROR Correction in slow speed data

Hello all,
Can anyone confirm if the D-RATS slow speed data has FEC (Forward Error Correction) built in now or not?
I understand that D-STAR voice does have error correction built in.  I've checked on the D-RATS home page/documentation
which is a August 2009 reference.

This is what I found in this document Under users manual/files:  Is this still current information?


Unlike the D-STAR DV Voice Channel, which has error correction built into the protocol, the Data Sub-channel has no error correction.



This makes the transfer of data across D-STAR a lot more dependant on a good radio path being established between the two locations wishing to



exchange data.  D-RATS includes the ability to test connectivity between two locations.

Thanks in advance
regards
Dave
W2SRP


Goals are where we want to be;
Objectives are the steps to get there.

There are 2 rules to success in life:
1. Don't tell people everything you know.

_______________________________________________
drats_users mailing list
drats_users at intrepid.danplanet.com<mailto:drats_users at intrepid.danplanet.com>
http://intrepid.danplanet.com/mailman/listinfo/drats_users


_______________________________________________

drats_users mailing list

drats_users at intrepid.danplanet.com<mailto:drats_users at intrepid.danplanet.com>

http://intrepid.danplanet.com/mailman/listinfo/drats_users


_______________________________________________
drats_users mailing list
drats_users at intrepid.danplanet.com<mailto:drats_users at intrepid.danplanet.com>
http://intrepid.danplanet.com/mailman/listinfo/drats_users



--
Dan Metcalf -- KB3UUN
c. 301 769 1111
Amateur Radio has something for everyone. Even if it is just peace of mind that no matter what happens your message will get through.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://intrepid.danplanet.com/pipermail/drats_users/attachments/20130921/b52b234d/attachment.html