[Asterisk-Users] fax still fails, ideas sought! Re: rxfax/spandsp fails to decode

[Stephen J. Wilcox]

Okay having taken in some suggestions and googled this topic to death I'm still 
stuck - anyone got any ideas?

To recap, the faxes are coming in via a digium E1 card but failing to train 
properly or if they manage it sending a garbled and very truncated fax.

A number of folks have suggested clock sync issues.. my zaptel.conf is set to 
use the PRI as primary clock, i have no evidence of issues altho dont know how 
to check (other than the call quality is fine, no clicks, no pri down/ups).

What can i try?

Steve

On Mon, 12 Jul 2004, Stephen J. Wilcox wrote:

> Hi,
>  I just sent this to Steve Underwood, but then found a bunch of posts on the
> mailing list about similar issues.. does anyone have the fix?
> 
> I'm running asterisk CVS-HEAD-06/28/04-18:13:13, spandsp 0.0.1k, libtif 3.5.7
> 
> one thing i just noticed is that calls come in with format '72' which is
> G711A-law or LinearPCM.. it uses PCM for the call, i assume this is ok
> 
> the results of RxFAX vary, it sometimes saves the file in which case i get 
> errors: 
> Fax3Decode2D: Warning, (FakeInput): Line length mismatch at scanline 0 (got 
> 2383, expected 1728).
> Fax3Decode2D: (FakeInput): Bad code word at scanline 1 (x 137).
> 
> and the resulting tif looks to be only a few rows long
> 
> or more commonly it just fails entirely.. i paste the output below so you can 
> see. is there anything obvious i'm doign wrong here?
> 
> TIA! Steve.
> 
>     -- Executing RxFAX("Zap/1-1", 
> "/var/spool/asterisk/faxes/20040712-183339.tif") in new stack
> Changed from phase 0 to 1
> Start receiving document
> Changed from phase 1 to 4
> Sending ident
> >>> CSI: 40 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20
> DIS:
> Preferred octets: 256
> Can receive fax
> Supported data signalling rates: V.27ter and V.29
> R8x7.7lines/mm and/or 200x200pels/25.4mm OK
> 2D coding OK
> Scan line length: 215mm
> Recording length: A4 (297mm)
> Receiver's minimum scan line time: 0ms at 3.85 l/mm: T7.7 = T3.85
> R8x15.4lines/mm OK
> Minimum scan line time for higher resolutions: T15.4 = T7.7
> >>> DIS: 80 00 ce f0 80 80 01
> HDLC underflow in state 9
> Changed from phase 4 to 3
> Slow carrier up
> <<< TSI: 43 31 37 31 31 36 35 34 35 34 38 30 20 20 20 20 20 20 20 20 20
> TSI without final frame tag
> Remote fax gave TSI as: "xxxx"
> <<< DCS: 83 00 86 90 00
> DCS with final frame tag
> In state 9
> DCS:
> Can receive fax
> Selected data signalling rate: V.29, 9600bps
> 2D coding OK
> Scan line length: 215mm
> Recording length: A4 (297mm)
> Minimum scan line time: 5ms
> Get at 9600
> Changed from phase 3 to 5
> Fast carrier up
> Fast carrier down
> Fast carrier up
> Coarse carrier frequency 1699.90 (64)
> Training error 56.874846
> Training succeeded (constellation mismatch 44.212022)
> Fast carrier trained
> Fast carrier down
> Trainability test failed - longest run of zeros was 14
> >>> FTT: 44
> Fast carrier up
> Training failed (sequence failed)
> Fast carrier training failed
> Fast carrier down
> Fast carrier up
> Coarse carrier frequency 1700.33 (64)
> Training error 51.989152
> Training succeeded (constellation mismatch 37.988826)
> Fast carrier trained
> Fast carrier down
> Trainability test failed - longest run of zeros was 15
> >>> FTT: 44
> Fast carrier up
> Training failed (sequence failed)
> Fast carrier training failed
> Fast carrier down
> Fast carrier up
> Coarse carrier frequency 1700.32 (64)
> Training error 60.898646
> Training succeeded (constellation mismatch 46.138793)
> Fast carrier trained
> Fast carrier down
> Trainability test failed - longest run of zeros was 17
> >>> FTT: 44
> Fast carrier up
> Training failed (sequence failed)
> Fast carrier training failed
> Fast carrier down
> Fast carrier up
> Coarse carrier frequency 1795.61 (4)
> Fast carrier down
> Fast carrier up
> Coarse carrier frequency 1789.60 (4)
> Fast carrier down
>     -- Channel 0/1, span 1 got hangup
>     -- Hungup 'Zap/1-1'
> 
> 
> 
> 
> 
>