Thread Subject: video support

From: Karen Peltz Strauss
Date: Wed, Sep 26 2007 6:20 AM

• Return to this mailing list's archives
• View all messages in this thread
• Next message in thread: None
• Previous message in thread: None
• Messages sorted by: Author | Thread | Date

I am going to take a stab at putting some of this info into a proposed guideline today, together with information I received a while back from Norman Williams of Gallaudet University. I will be forwarding it to Paul and Ellen, who agreed to initially review it.

Karen (I changed the name of this thread to make it more in line with the subject matter)
----- Original Message -----
From: Gunnar Hellström
To: 'TEITAC Telecommunications Subcommittee'
Sent: Tuesday, September 25, 2007 8:38 AM
Subject: Re: [teitac-telecom] Reminder - Telecommunications Sub Committeecalltoday at 2pm


Peter,
Good questions.

1. For X, if you ask today what is the most important interoperability format for multimedia converational services, the answer is :

IETF SIP for call control, ITU-T G.711 for audio, ITU-T T.140 for real-time text transported with IETF RFC 4103 and ITU-T H.264 for video possibly with a fallback possibility to H.263.

Note tat I say "interoperability format". There are environments where it is perfectly reasonable to use other standards. For example, the G.711 audio coder takes too much bandwidth for many mobile situations, and does not deliver any better audio quality than the old telephone, so other standards would be used as well. If the devices in a system do not follow the interoperability standards, there shall be a way to take the communication through an interchange point where the protocol and codecs listed above shall be included, and the call control be as open as possible for communication between users of terminals from different manufacturers and providers.

There are also other environments where video is available. The 3G circuit switched videophones may soon be introduced in USA, and they have some usability for sign language, for seeing and recognizing, but barely for lip reading. My intention with the proposed wording was that providers of such systems would contribute to making them part of a network of interoperable communication systems, aiming at the goal that as many as possible shall be able to call each other and use as many as possible of the three media. The currently most natural common protocols are the ones specified under X above. It is not enough to just use a common interoperability protocol, there must also be agreements to authorize calls between providers.

This is the reasoning about what would be best today for the users and procuring parties. From that knowledge there should be a discussion about how to express it best as a requirement.

2. Video quality. There is an ITU document giving the values you ask for with motivations in human perception and practical experience:

http://www.itu.int/rec/T-REC-H.Sup1/en "Sign language and lip-reading real-time conversation using low bit rate video communication "

For good usability >20 frames per second, CIF resolution (352*288) and a latency of less than 400 ms is needed.

Some possibility to use has 12 frames per second and QCIF (176*144)

Lower quality values than that has very low usability and requires severe language adaptation to make any use of it if ever possible.

Within reasonable limits it is more important with high frame rates than high resolution. Thus in mobile phones, where some compromise may be needed, it would be better to provide 20 fps QCIF than 12 fps CIF.

Please read the ITU document for full understanding.

The values have been verified in practice several times after the document was published, and it is the base for procurement approvals by the Swedish Handicap Institute.



Regards

Gunnar
-------------------------------------------------------------------
Gunnar Hellström
Omnitor
= EMAIL ADDRESS REMOVED =
Tel: +46708204288
www.omnitor.se

-----Original Message-----
From: = EMAIL ADDRESS REMOVED = [mailto: = EMAIL ADDRESS REMOVED = ] On Behalf Of Peter Korn
Sent: Tuesday, September 25, 2007 12:58 AM
To: TEITAC Telecommunications Subcommittee
Subject: Re: [teitac-telecom] Reminder - Telecommunications Sub Committee calltoday at 2pm

Hi Gunnar,

Your suggested update makes sense, and adds additional clarity and focus. However, I still have two concerns with it:

1. We still don't have a definition of what "in X format" means. The subcommittee are supposed to have largely completed their work by now, as we head toward final text. Are there any candidate "X formats" for discussion for this provision (that I perhaps missed)? As this is an important facet of the requirement, I don't think we can continue to leave it undefined.
2. The phrase "sufficient quality and fluency to support video communication for sign language, lip reading", etc. while more descriptive and focused, is still somewhat subjective. Is there any objective standard for this? Number of pixels dedicated to the region of the mouth? Number of pixels overall, color/contrast ratios, frames-per-second? Those sorts of things are measurable, and can be independently verified without needing to round up some folks who speak sign language or who lip read, and asking them their opinion of the usability of any particular video transmission medium.


Regards,

Peter Korn
Accessibility Architect,
Sun Microsystems, Inc.

> Proposal for 5F -Video support
>
> Current text:
> Telecommunications products or systems which have the capacity to
> transmit video, text, and voice communications must support internet
> protocol text and voice communications in X format, and they must have
> sufficient transmission bandwidth capacity to support video
> communication such as video relay and point to point video
> communications.
>
> Problems:
> 1. The experienced quality level should be defined, not the bandwidth.
> 2. It is products supporting video that shall meet these requirements.
> Support for text and audio is a follow-on requirement.
> 3. The purpose for the quality requirement should be stated.
>
> These considerations lead to new proposed text:
>
> Telecommunications products or systems which have the capacity to
> transmit video, real-time text, and voice communications must support
> interoperability with internet protocol text, video and voice
> communications in X format, and they must have sufficient quality and
> fluency to support video communication for sign language. lip reading
> and other personal communication such as video relay and point to
> point video communications.
>
>
> Regards
> -------------------------------------------------------------------
> Gunnar Hellström
> Omnitor
> = EMAIL ADDRESS REMOVED = <mailto: = EMAIL ADDRESS REMOVED = >
> Tel: +46708204288
> www.omnitor.se <http://www.omnitor.se/>
>
>
> ----------------------------------------------------------------------
> --
> *From:* = EMAIL ADDRESS REMOVED =
> [mailto: = EMAIL ADDRESS REMOVED = ] *On Behalf Of *Brenda
> Battat
> *Sent:* Monday, September 24, 2007 3:09 PM
> *To:* TEITAC Telecommunications Subcommittee
> *Subject:* [teitac-telecom] Reminder - Telecommunications Sub
> Committee calltoday at 2pm
>
> Just a reminder the committee meets today at 2pm and the agenda is 5-F
> IVR and 5-F Video support
>
> ----------------------------------------------------------------------
> --
> *From:* = EMAIL ADDRESS REMOVED =
> [mailto: = EMAIL ADDRESS REMOVED = ] *On Behalf Of *Diane
> Golden
> *Sent:* Monday, September 17, 2007 3:25 PM
> *To:* 'TEITAC Telecommunications Subcommittee'
> *Subject:* Re: [teitac-telecom] Today's Telecom Call - volume gain
>
> See if this summarizes what we've been discussing:
>
> Base requirement - E&IT that provides a function allowing voice
> communication via audio transducer typically held to the ear (except
> those exempted below) must provide a gain adjustable from the normal
> unamplified level to at least 15-16 dB above the normal unamplified
> level as measured in accordance with the provisions of the FCC
> regulation §68.317 for volume control. The volume at the normal
> unamplified level setting must also meet the requirement in FCC
> regulation 68.317.
>
> Exception 1 - Analog line-powered E&IT and battery powered cordless
> telephones must comply with FCC regulation §68.317
> for volume control.
>
> Exception 2 - Cellular phones
>
>
> Diane
>
> -----Original Message-----
> *From:* = EMAIL ADDRESS REMOVED =
> [mailto: = EMAIL ADDRESS REMOVED = ]*On Behalf Of *Brooner
> Mary-AMB004
> *Sent:* Monday, September 17, 2007 11:46 AM
> *To:* = EMAIL ADDRESS REMOVED =
> *Subject:* [teitac-telecom] Today's Telecom Call - volume gain
>
>
> Gentlepeople: -
>
>
> I have done a bit of drafting to improve our discussions.
> There is still a lot to work out but take a look at this. The
> underlined text is new wording, the [x] text would be
> deletions from current text.
>
>
> See you at 2 pm.
>
>
>
>
>
> .2-E - Volume (Gain)
>
> For incoming voice signals:
>
> 1 Users must be able to adjust the audio output level.
>
> 2. Analog line-powered telecommunications products, wireline,
> and all _wireline or VoIP_ cordless telephones _[_wireline or
> VoIP_]_ must comply with FCC regulation §68.317 for volume control
>
> 3. All cellular phones - TBD
>
> _ 4._ _ __Digital _ [All other telecommunications]products or
> systems that provide a function allowing voice communication _ via an
> audio transducer typically held to the ear_ with a hand must provide a
> gain adjustable from the normal unamplified level to at least 18 dB
> above the normal unamplified level as measured in accordance with the
> provisions of the FCC regulation §68.317 for volume control. The
> volume at the normal unamplified level setting must also meet the
> requirement in FCC regulation 68.317.
>
> _ 5. Other products or system that provide a function allowing
> voice communications should provide adjustable volume up to 18dB._
>
> _6. Where additional volume is needed, ... _
>
> _*Advisory Note 1* Volume (gain) on cellular and PCS handsets is
> currently the focus of review and study in ATIS Incubator. It is
> recommended that no recommendation at this time be made by the
> TEITAC, but rather, the recommendations from the study can be
> given to the Access Board at a later time._
>
> __
>
> _*Advisory Note 2* There are other digital products and systems
> such as digital telephones and systems that draw from Ethernet, as
> well as laptop telephone emulators which use a headset. None of
> these choices can be measured using FCC regulation 68.317 for
> volume control. In the case of laptop telephone emulators, they
> are typically used with a headset which is an accessory._
>
> *Discussion or Rationale:*
>
> * Language for cellular phones pending outcome of ATIS Study Group
> 11 work
> * Further research needed for 20 dB requirement
> * Review telecoms/communications to be sure requirement is inclusive
>
>
>
> __________ NOD32 2545 (20070923) Information __________
>
> Detta meddelande är genomsökt av NOD32 Antivirus.
> http://www.nod32.com
> ------------------------------------------------------------------------
>
>

• Next message in Thread: None
• Previous message in Thread: None