Hi Steve,
Many thanks for your reply. Yes, I am sorry I should have elaborated a
little more in my initial post. We use a VoIP (SIP) connection between
Callweaver and a Cisco 5400 gateway which terminates to the PSTN. I
suspect the problem might be with the Cisco gateway firmware and the DSP
firmware/levels. I thought however that I would check here with the list
first before tackling the Cisco and TAC which is always a daunting prospect!
I am still running SpanDSP 0.0.3 which has been very stable and given
good results (particularly in receive mode). I did attempt an upgrade
to 0.0.4pre1 back in April but the results were worse so I went back to
0.0.3. I will try the latest code again though and let you know how it
goes. Would you suggest I use 0.0.4pre3 and CallWeaver 1.2rc4?
Thanks again for your feedback and keep up the good work!
Cheers,
Ray
Steve Underwood wrote:
Hi Ray,
Ray Jackson wrote:
Hi all,
We are running SpanDSP with Callweaver and have been having issues
sending faxes using the TxFax module. to certain fax machines (Brother
machines in particular). I realise this is more a SpanDSP question,
but we ran a fax check report with our local Telco and the report
states that our fax service is 'OVER equalised (4dBm)'. A copy of the
technical info in the report can be found at:
http://gravsol.com/faxtechnical.png
I also hear that Brother fax machines don't like to be over
equalised. Is there a way of adjusting SpanDSP's equalisation levels
somehow - would I have to go into the source to do this or can I pass
a parameter via Callweaver? Any assistance would be greatly appreciated.
There are a number of things other than spandsp between the signal
generation, and your measurement point. I just made the following
snapshot of the spectrum coming out of the V.17 modem in spandsp -
http://www.soft-switch.org/V.17_spectrum.png It looks much less smooth
than the graph in the telco report, because it is a short term snapshot
of a rapidly varying signal. If I produced a longer term image it would
look smoother. However, the key point is my graph does not show the
sprectral tilt which the telco's graph shows.
I guess you have an analogue path between txfax and the telco's
measurement point, and this is producing the results they see. 4dB tilt
across the spectrum and very little group delay variation is actually a
good path. Most analogue paths will give worse results than that. It is
worth noting that their analyser recorded a 100% error free result.
Are you running the very latest spandsp code? if not, get the latest
snapshot. A number of FAX compatibility issues (annoying little FAX
protocol issues, and not modem ones) have been addressed in the last few
weeks.
Regards,
Steve
_______________________________________________
Callweaver-users mailing list
[email protected]
http://lists.callweaver.org/mailman/listinfo/callweaver-users
_______________________________________________
Callweaver-users mailing list
[email protected]
http://lists.callweaver.org/mailman/listinfo/callweaver-users
