Re: [rtcweb] Fwd: Last Call: <draft-ietf-rtcweb-audio-10.txt> (WebRTC Audio Codec and Processing Requirements) to Proposed Standard

[Roman Shpount <[e-mail]>]

On Fri, Feb 26, 2016 at 2:38 AM, Harald Alvestrand <[e-mail]>
wrote:

> Roman, you're the expert here - googling for the text unearthed this
> message from you in 2013:
>
> https://lists.w3.org/Archives/Public/public-webrtc/2013Apr/0030.html
>
> Reviewiing that archive might save you the effort of repeating the
> arguments again; at the moment, I have no idea where the upper limit came
> from.
>
>
Harald,

All I am trying to achieve is getting both WebRTC TR and
draft-ietf-rtcweb-audio say the same thing. The current tone duration
limitats are present in the W3C document. I do not necessarily agree with
these limitations, but I do know where they come from. I do think these
limitations should be reviewed by this group since they clearly define the
network related aspects of DTMF tone generation. Whatever is decided by
this group needs to be then replicated to W3C WebRTC document.

The minimum duration for tone and the gap in the W3C document comes from
ITU Q.24 specification.

The maximum duration of 6 seconds is something Cullen put in when he wrote
the document. Maximum duration of RFC 2833 tone or single segment RFC 4733
tone with 8 KHz RTP timestamps is about 8 seconds. If 48 KHz clock used,
max single segment tone duration would be about 1.3 sec. I have not seen
any implementations of RFC 4733 long duration events (
https://tools.ietf.org/html/rfc4733#section-2.5.1.3) in the wild since DTMF
tones are mostly used with 8 KHz clocks and no one really cares about DTMF
events longer then 8 sec. Support for multi-segment long duration DTMF
events would be necessary if someone will decide to use DTMF with Opus
since DTMF digits with duration longer then 1 sec are used by some IVR.

My personal preference (as expressed in the quoted message) not to specify
min and max duration for DTMF tones or gaps. I think having reasonable
default values is enough. If developer cares enough to adjust the duration
values, I assume he or she would know what they are doing. If they set
something silly, they will have to deal with the consequences.

Other people on the list wanted to limit the number of foot guns in the API
and decided to put reasonable min and max limits on tone and gap duration.
This is not my first preference, but I do not think these limits are
unreasonable, so I did not object too much.

Regards,
_____________
Roman Shpount