[Freeswitch-users] Fax processing not successful - result (48) Disconnected after permitted retries

[Steve Underwood]

On 07/18/2012 07:48 PM, afshin afzali wrote:
> Hi Steve,
>
> I don't think that my problem relates straightly to the time that a 
> fax takes to be received. In some cases I could get better result to 
> decrease fax speed (normally causes longer time). Let me tell you my 
> today case. There is a tabular document that I could receive from a 
> Panasonic fax machine and the same time unable to receive that from a 
> Canon machine. Although I'm receiving other (simpler than) documents 
> from the Canon one !
A huge number of Canon machines have a broken V.29 modem, which produces 
such a poor signal you are lucky if other machines can decode it. The 
only real difference a complex image makes is the amount of data. This 
has only two effects:
      - Obviously it takes longer.
      - In ECM mode, if the image exceeds 64k it will be sent in 64k 
blocks, with error correction between each block.
If your Cisco is running software more than a couple of years old, its 
T.38 engine may be horribly broken. It seems that had so much trouble 
with it, they tossed it out and put a complete replacement in there. The 
new one usually works well.
>
> I've tested various configurations on my 5350 device of modifying fax 
> rate, ls_redundancy and hs_redundancy but could not managed to fix this.
>
> Afshin
>
> On Wed, Jul 18, 2012 at 4:00 PM, Steve Underwood <steveu at coppice.org 
> <mailto:steveu at coppice.org>> wrote:
>
>     On 07/18/2012 06:54 PM, afshin afzali wrote:
>     > Hi,
>     >
>     > There are some inbound fax sessions which I got error code 48
>     (and in
>     > some cases 49).
>     >
>     > FreeSWITCH Version 1.0.head (git-54ddef0 2011-12-06 21-53-45 -0600)
>     > AS5350XM Voice / Fax ( T.38 ) Gateway
>     >
>     > I've noticed that this error appears in cases which there are
>     complex
>     > documents (graphics or gray areas).
>     > appreciate all comments !
>     If you have graphics, the pages generally take a lot longer to
>     send. It
>     is not uncommon for T.38 gateways to have stupidly short timeouts,
>     which
>     will cause a call to drop after a page has been in progress for
>     just one
>     minute, or one and a half minutes. Could that be your problem?
>
>     Steve
>

Steve