Aught! I knew there would be a hiccup or two.
So, a modern smartPhone might make a sound (polyphonic as it is) that is similar to the good 'ol DTMF signals but isn't a match enough for the DTMF decoder.
Perhaps an actual old-style POTS phone should be used to produce the signal to drive this.
In the links I saw with using the DTMF to control a robot they had a cellphone strapped to the body of the robot. I didn't look to close at what they were doing, but I surmised that they were using a second phone to call the one strapped in and manually pressed the answer button to connect the call, and then used DTMF signals phone-to-phone to control it.
The truth is, I don't know how the OP is planning on controling this. hfbouri said, "give a signal from a phone to a circuit (not through cable)" and I'm not sure what that means except that "not a cable" means "not hard wired."
I assumed an audio connected signal at first, then audible sound from a smartPhone, now I don't know. So that is something
we need to know. (The source of the driving signal or the method of control.)
About "capture time," according to the
DTMF datasheet the "Min Tone Duration Accept" is 40ms. That seems pretty quick to me, but if the melody had to be slowed down to accomidate that, I don't think that's a huge hurdle to overcome.
-t
I just invision myself walking up to the door to my house and wistling a tune to diable my security alarm. That'd be so cool. (But then I don't have an alarm, and I can't particularly wistle. Oh well.)
edit:"I see now that the OP said,"i want to play the sound on my htc phone" so
that is the driving signal. If the htc can't make an acceptable DTMF then that is another hiccup!!:endEdit