[asterisk-dev] [Code Review] Fax Gateway Implementation T30<->T38

[Kevin P. Fleming]

On 04/14/2011 01:32 PM, Pavel Selivanov wrote:
>
>> No, this is not correct. DAHDI can report detection of CNG tones if an
>> echocan (usually a hardware echocan) reports them, but it does *not*
>> setup a CNG detector.
>>
>> DAHDI *does* setup a CED detector and react to detection of CED when
>> appropriate. In addition, in current DAHDI releases, it does not
>> 'switch the echocan off', instead it tells the echocan to switch into
>> 'FAX' mode; the exact effects that result are dependent on how the
>> echocan implements that operation.
> This is not correct too :-)
> if chan->ec_state->features.NLP_automatic is true => setting mode to
> ECHO_MODE_FAX
> if chan->ec_state->features.NLP_toggle => setting mode to ECHO_MODE_FAX
> & sending DAHDI_EVENT_EC_NLP_DISABLED event to asterisk.
> if no features registered => setting mode to ECHO_MODE_IDLE.
> In __dahdi_ec_chunk, echocan_process is called if
> (ss->ec_state->status.mode != ECHO_MODE_IDLE)
>
> For example, DAHDI will not pass data to OSLEC, because OSLEC do not
> support any "features".

If OSLEC does not support any 'features', then NLP_automatic should not 
be set to true. NLP_automatic indicates to DAHDI that the echocan will 
handle its own NLP based on CED detection and that DAHDI does not need 
to do the detection.

It does appear to be a bug that the audio is not passed to the echocan 
in ECHO_MODE_FAX, though. That should be reported and get fixed.

-- 
Kevin P. Fleming
Digium, Inc. | Director of Software Technologies
Jabber: kfleming at digium.com | SIP: kpfleming at digium.com | Skype: kpfleming
445 Jan Davis Drive NW - Huntsville, AL 35806 - USA
Check us out at www.digium.com & www.asterisk.org