[Asterisk-Users] TxFax: can't get a fax to destination (log inside)

[Steve Underwood]

Lee Howard wrote:

> Marco Parmeggiani wrote:
>
>> Can someone explain me what's going on and why the receiver of this 
>> fax guives up saying communication error?
>>
>> Start tx page 1
>> >>> EOP: 2f
>> <<< RTN: 4c
>> >>> DCN: fb
>> Disconnecting
>
>
>
> The receiver says communication error because txfax's response to the 
> receiver's RTN signal was a DCN (disconnect) signal instead of a usual 
> Phase B signal (i.e. DCS), also known as "retraining".
>
> In non-ECM mode a receiver analyzes the "quality" of the incoming 
> image data and determines whether or not it is of an "acceptable 
> quality".  Then, after the sender transmits the post-page signal (EOP 
> in this case, meaning "last page"), the receiver transmits either MCF 
> (confirmation, meaning the received page quality was acceptable) or 
> RTN (meaning the page quality was unacceptable and that retraining is 
> required before the next page is received).
>
> My understanding of T.30 (and I'm not alone in this thought) was that 
> RTN is meant to be the antithesis of MCF.  So not only does it mean 
> that retraining is required for the next page, but it also means that 
> the previous page was *not* received properly (i.e. not printed).  So, 
> basically the page will need to get resent for the page to get 
> through.  However, many fax senders (especially old fax machines) were 
> incapable of retransmitting a page because they didn't have the memory 
> to store a full page of image data and so they sent data as they 
> scanned it.  So of necessity these senders had to treat RTN as a 
> confirmation and they left it up to the receiver to call back and 
> request a retransmission of whatever pages came through uncleanly.  
> Some senders may have even produced a transmission report indicating 
> to the sender which pages should be retransmitted in a new call.  
> Ultimately what this did for the fax community, though, was that it 
> gave us a very confused understanding of an RTN signal.  And so 
> various things were programmed with this confusing understanding 
> programmed-in, and this programming often persists in "modern" fax 
> machines because manufacturers don't rewrite their fax code very often 
> if ever.
>
> So txfax is behaving as many fax senders out there behave, although my 
> opinion is that it is behaving incorrectly... and apparently this 
> receiver that you are working with shares that opinion.

TxFAX is doing the right thing, according to the spec. If you don't know 
how to handle RTN you should disconect. spandsp has not yet been taught 
how to retrain and retransmit properly. However, this is largely 
irrelevant. On modern lines a retransmit should be rare. Lately lots of 
people are reporting this problem to me, and I have not had time to 
investigate why it happens. Some say it happens with specific machines, 
and some with all machines.

Regards,
Steve