dtmf tones seem oddly long and inconsistent

[kevin gunn]

image#125
just testing in general, i noticed the dtmf tones seemed oddly long
then i compared to my iphone, definitely longer - i would assume there's a spec, but maybe our tones are within that spec ?
I also noticed that regardless of how i would consistently quickly strike the key, the tone length seemed to change arbitrarily

[Bill Filler (bfiller)]

we are not generating the local tones, so currently you only hear what was played back from the other side. This should get fixed when we integrate the tone daemon that Martii is working on

[Martti Piirainen (piiramar)]

As Bill says, there is no locally generated sound at all, that'll be done via bug 1187453.

I tested the behaviour in mako HW with image 138, both participants of the voice calls in Sonera, Finland UMTS network. The tones sent through the network to the other end work fine for me. There is some variance in timing, but I don't think the tones are "oddly long". I have successfully used mako to connect to conference call systems, giving the conference ID/PIN via DTMF.

The technical specification is not very strict: According to [1], the tones and the gaps between tones must be at least 65ms, with no upper limit defined. I think between 100 and 400ms are typical values.

[1] 3GPP TS 24.008 "Mobile radio interface Layer 3 specification;Core network protocols", section 5.5.7.5 "Sequencing of subsequent start DTMF requests by the mobile station"

Kevin, can you be slightly more specific on the length of the tones that you hear in the Ubuntu device and the refence device?

[kevin gunn (kgunn72)]

ok, yeah, the if the spec is that loose, then this makes sense as it fits with about what I was hearing (e.g. some tones very short and some tones maybe 1/2 sec). I was just surprised with that variance hence the bug.
we can mark won't fix