<p>Hello,</p>
<p> </p>
<p>I'm back on this.</p>
<p> </p>
<p>I could take samples of what is sent by my carrier as CLIP and what is sent by my FXS port as CLIP, and this is very different.</p>
<p>The modulation to be sent out of FXS seem to be correctly built, but this seem to never get out.</p>
<p> </p>
<p>By further checking what could happen, I could see that for FSK/DTMF/... the data is written to some dedicated buffer (fsk_buffer/...) using the zap_buffer_write function.</p>
<p> </p>
<p>I continued to see what was done with this fsk_buffer, and surprisingly, it seems that nothing is done. I could find the beginning of something in ozmod_analog.c.</p>
<p>these lines commented out :</p>
<p> if ((zchan->dtmf_buffer && zap_buffer_inuse(zchan->dtmf_buffer)) || (zchan->fsk_buffer && zap_buffer_inuse(zchan->fsk_buffer))) {<br /> //rlen = len;<br /> //memset(frame, 0, len);<br /> //zap_channel_write(zchan, frame, sizeof(frame), &rlen);<br /> continue;<br /> }</p>
<p>so that seem to be related to fsk_buffer (by the way sending dtmf tones out of FXS does not seem to work either, tried with send_dtmf application in diaplan), but I don't really get what would be done here (if uncommented) except sending null values to that function, which would (I found this function in zap_io.c) lead to doing nothing as this function verifies the condition zchan->fsk_buffer && zap_buffer_inuse(zchan->fsk_buffer) and returns if this is true.</p>
<p> </p>
<p>That lead me to think that maybe this is a non implemented feature and maybe the data in fsk_buffer should be sent somewhere to the card.</p>
<p> </p>
<p>Could anybody shed me some light on this.</p>
<p> </p>
<p>Thanks</p>
<p> </p>
<p>François</p>
<p> </p>
<p>On Wed, 27 Jan 2010 11:22:13 -0500, Michael Jerris <mike@jerris.com> wrote:</p>
<blockquote style="border-left: 2px solid #1010ff; padding-left: 5px; margin-left: 5px; width: 100%;"><!-- html ignored --><!-- head ignored --><!-- meta ignored --><br />
<div>
<div>On Jan 27, 2010, at 3:48 AM, François Legal wrote:</div>
<br class="Apple-interchange-newline" />
<blockquote>
<p>So I did modify mod_openzap to provide the od trace command to the cli. I started trying to parse the incoming clip, but I have to do it again as the incoming signal is saturated so I can't see anything interesting.</p>
<div> <br class="webkit-block-placeholder" /></div>
<p>I'll post my results whenever I get something.</p>
<div> <br class="webkit-block-placeholder" /></div>
<p>Shall I send the modified mod_openzap.c to the list ?</p>
<div></div>
</blockquote>
We should get a bug opened on this issue on <a href="http://jira.freeswitch.org">http://jira.freeswitch.org</a><br />
<blockquote>
<p>About the difference on the modulation settings for CLIP, could anybody suggest some way to have it configurable by files ?</p>
<div></div>
</blockquote>
We should add some config option based on country or CLIP type to specify this. Lets get it working and we can think how to configure once we have that.</div>
<div></div>
<div>Mike</div>
<div></div>
</blockquote>
<p> </p>