a model for remote conferencing
|telephane/telephone messaging, is user-time efficient...|
we summarize the development of a video conferencing model
based on remote, live, scientific conversational scenarios
[and subsequently adapted on several such test scenarios]
Beyond the conventional teaching model in which students are present
with the teacher, and thoughts are dissected and discussed impromptu
(meaning, the teacher is prepared, and the discussion spontaneously
follows a course guided by the teacher) the works of
classical authors have always inspired readers and listeners to feel
their presence, across Earth-radii of space
and Earth-millennia of time ... thus it is considerably feasible to
design and implement a remote live video conferencing model with
similar classical feel well-adapted to modern scientific technical
conversational modes of thought.
Ca 1984 and 1990, I outlined and 30-page sketched a science fiction
story called, 2100: The Grand Odyssey, in which by-then-contemporary
university research technology will have pushed out beyond lunar space
and its 2.6 second delay loop communications with Earth,
to sun-Earth libration LS#1 (orbiting once each 365.25 days) --
figuring that as heat and light come from the sun, so also shall information.
Thus a 14-second delay loop necessitated a conferencing model
accommodating emergent live but remote distant audiences of research scientists
and government agencies: so developing the multi-way distance-conference.
The essential conference views what each member contributes impromptu,
being prepared but spontaneous in conversation,
without losing the participant audience of tens or hundreds:
Thus both, voices with face visuals, and diagrams with computer graphics,
displaying before large present audiences, and before their several remote audiences,
must be temporally and spatially interlocked:
Typical vidlinking is comprised in a double [or double-width]
display with stereo-audio speakers, microphones, and face-cameras in-close.
[see also stereo-eyes-ed]
Each audience must have interactual control of its video+audio link
while viewing responses from others temporally intersticially,
and by spatially subconferencing as may be needed in situations
for aside conversations in sub-groups of the main audience.
If placed between audience-participant seats,
each main half display contains the face visuals and main controls for
temporal scheduling and spatial sub-grouping, and alternately each second
half display contains the extended computer graphics and other call-up
file imagery (live process or stored record files) and controls.
HDDV (2:1) or wider aspect can include both halves in one display serving
pairs of participants -- or one-per-participant allows full sub-audiencing.
Additionally, much of the vidlink control can be handled by audio
call-to-the-front voice-response technology.
This vidlink model has been tested on several screenwritten scenarios, ranging among:
A live local immediate four-way tele-phane conversation with vox-pause
[see The StarTrails Game]
in which 2x2 cross-talk develops too small to advantage sub-conferencing,
but is cleaned-up by selective partial archiving;
A live remote Earth-lunar 2.6 sec. delay loop two-way conversation
with vox-pause and single-side fast-remind
[see Professors' Spring Fling]
where delayed meanings become critical of asymmetric rank protocols;
A live remote Earth-moon-LS#1 14 sec. delay loop three-way audience conference,
also with vox-pause, and full-duplex fast-remind
[see AD2100 outline]
in which key personnel speak or lecture with auxiliary computer graphics
to large audiences permitted prompt albeit delayed
return questioning and spontaneous (relevant) injected commentary;
A remote correspondence Earth-deep-space multi-hour delay two-way communicator conversation,
again with vox-pause but two sub-level tinkle-annunciated clip-remind
[see Professors' Spring Returns]
operating like indented message forum posting;
And other examples of days delayed video correspondence.
[Asymmetric split-delay was not scenarioed, though it did appear once as a
single-side-broadcast that went live near-equal-delay]
Across all scenarios multi-respondents must engage their temporal
expectancies to larger audiences, simpler, less often speaking, and
listening, thinking, reasoning and responding more topically focused:
jumping freely among train-of-thought concepts and images,
as among multiple, fast-computer internet-web information display "task" panes:
The educational doctrine of brain-residency memory-processing latencies
and staged temporal retentions evokes fast-remind rules and criteria
for the various delays at intra- and inter-planetary communication distances,
for switching-in longer review fractions
after greater delays, locally or remotely cached.
Straight telephone-mail, or with video for simple tele-phane vidlink,
operates much like remote delayed e-mail,
utilizing the telephone or tele-phane for both command and composition,
speak-and-send messaging and scheduling,
directly improved to [voice-prompt] vox-paused listening and interstitial reply,
with partial phoneme or word recovery at continuation interphases --
also preferably with voice-response, verbal-commanded interactivity
(essentially contemporary state-of-the-art technology)
for hands-free telephone/tele-phane utility:
The utility-providing service can further afford to adapt scheduling
to its customers' own recipient-information base, and meet the daily
agenda needs of both, calling intended recipients when and where they are available,
with established regular priorities (successive choices).
As response-delay decreases to live scenarios,
message-oriented telephone/telephane-mail adapts programmably by maintaining
live-connectivity while video imaging+voice retains its speak-and-send format more fragment-oriented,
melding to interlock with the vox-pause controls,
and abbreviating the original echo from two sub-level tinkle-annunciated clip-remind,
to 2-3x speed fast-remind ending in 2 seconds normal-speed,
until the respondent's own loop-delay is brief enough (seconds)
and multiple respondents overlapped enough that fast-remind may be
omitted unto word-boundary recovery, selectively at the audience,
as the natural conversation phrase-time also shortens automatically by the vox-pause.
E-mail-like capability remains then as distinct
functionality in archiving and forwarding copies
of conferences to participating and non-participating recipients.
The intent of design is to model a scalable and extensible voice-video
system, that works similarly at all system and theater sizes: from
person-to-person direct, to multiple large audiences in conference:
from local to global to geo-space to deep-space:
from instantaneous (local) to prompt fraction of a second (global)
to quick seconds (geo-space) to hours (deep-space) to days (certain situations):
Operation is facilitated best by user-consistent protocols.
A live correspondent connects to a conversation by calling-in,
or by answering, or by calling-out and waiting for others to answer -- typical telephony:
If certain correspondents are not available, the [recorded] vidlink session
can be forwarded to them at their receiver address (much like e-mail)
on a reprogrammable schedule: as joining late may be unwise.
With the advent of session recording,
the later recipient can reply to their video record as though live
(though responses will come later still),
and return similarly a larger record to any of the earlier recipients:
The larger interstitial response delay and enlarging
correspondence necessitates gleening and culling,
which can be done both manually and automatically
to typically two sub-levels of clip-remind, denoted by tinklings.
Long-term cache is maintained for the users at both ends,
and the correspondence fragments piecewise reconstructed:
Given that correspondent responses are reciprocal and expectably
regular to two-sub-levels (more so than on prior-era message forums)
this significantly improves conference transmission efficiencies.
When conversing at liberty before internet-video imaged communicants
it is preferable to be able to speak in response spontaneously --
"get a word in edgewise" -- though this is novel in direct speech.
In the local frame, this is synchronized by (auto)vox-pausing the incoming
display while speaking, which continues to fill the receiver cache.
Then upon voice-checking, the vidlink resumes displaying from cache,
having also backed-up to the nearest whole word for verbal continuity
(or words, until a whole phrase is needed).
In its simplicity this provides e-mail-like interstitial responses,
directly addressing each concept thought, as spontaneously deemed needed.
Both correspondents have this verbal interlocking assistance,
and in concurrent speech maintain their perspective by watching
each other's vidlink frame:
FACE IMAGE FRAME COLORS PHASING
|[BLUE] Connected alert|
|[YELLOW] Queued speech|
"Not if we include a…"
"...and factors of..."
Compounded colors are alternating or averaged.
Connected alert, standing-by has no intrinsic operation but connected: status queriable.
Receiving, may include subconferencing.
Concluding, implies not listening, cache running-out to Stand-by.
Listening, implies more than 5 seconds.
Finishing, implies less than 5 seconds to end of speech.
Queued speech, initiates or replies to any speech or segment.
Speaking, is the instantaneously -lone- Speaker.
Interactive, although best to listen, speaking-back vox-pauses the playout of the Speaker.
Each Speaking or Queued is identified for immediate correspondent and thread.
Each Speaking or Queued Frame includes words translated to text --first words, for topical identification--.
Responses to preceding segments are clip- and fast-reminded
(partially tail-excerpted and earlier portions fast-replayed) --
but to only one sub-level tinkle-annunciated.
Individual listeners may command-request two sub-level fast-remind
on any segment -- by voice-command or key/button.
Very late responses to previous segments are cached aside
for discussion when mediator-command-requested during overall silence:
But, individual late speakers intending to catch-up and be heard promptly,
may command-request re-prompted delivery.
While cached on the speaker's side,
segments may be re-edited or deferred before delivery.
If several [3 or more] conversants speak simultaneously
[to the same heard phrase whenever they heard it] a warning is raised,
and the several may be heard together at equalized loudness,
or separately, or replayed, per the listener's command-request
and the per-station choice settings;
Individual speakers intending to be heard above such a multiple response,
may by command-request or per-station choice settings,
elect to be separately re-heard (in case any listeners had not).
Speaking-voice exceeds a vox-actuation setting,
and sustains above a squelch [hysteresis] threshhold--
a color/bar-meter monitors voice intensity:
Vox-actuation may be initiated by direct speech imposition
or by voice-sub-command addressing (eg. prompting "Sir/Sirs/Miss/...")
or by-name introductorily emphasized, or by key/button.
A speaker appears paused within frame notification-colors (green-yellow-gold-red)
until played-out or dequeued.
Squelched intervals are flagged as unnecessary, and free-scanned (which can appear
near continuous on high data-rate connections).
A separate voice-command or key/button (e.g. 'bozo' pairs) dehances auto-response to
private-room-level adjacent conversations:
and actuates the blue-green alternation to other correspondents.
A lag-monitor indicates temporal length of unheard cache.
Sub-conferences are specified by voice-command or by keying for persons (or group),
and the recipients notified of pending conversation, by color, flags, and lag.
Individual listeners may choose layered vs outlined
(threaded) vs linear vs grafted (pushed); momentary overrides; catch-up;
(e.g. compound interactivity speaker and instantaneous-direct-responder, both interactive,
vs. linear and followup-with-reminders)
Individual listeners may prioritize threads (sub-conferences by permission).
Heard segments are so flagged (cf web-forum visited-link indications).
Individual listeners may deselect threads not directly including themselves.
Because both video and speech can be time-compressed successfully,
fast-remind condenses the conversation interval without hampering
the listener: A factor of 2x to 3x "scanning" compression
expands a given response by less than 50%, or as little as 33% --
which by the end of the second sub-level tinkling typically
averages 33-48% of the total correspondent conversation.
A more advanced programmed fast-remind can enunciate key words.
Long correspondence delays require full clip-remind to two sub-levels:
The first sub-level lets the listener hear what is to be answered,
and what prompted that which is being heard (the listeners' own prior words: a self-reminder) --
the first sub-level being demarked by the first single tinkle applied both ways --
the second sub-level (tinkle-tinkle) supplies an additional temporal
precedence of conversation recovered or continued.
The primary user will have full featured accessibility:
an installed personal phone [equivalent to direct automatic log-in: hardwired passcoding]similarly any identity-encrypted cellphone [firmwired passcoding]
any public telephone to portal [touch-toned, dial-clicked, voice-commanded: key-in passcoded]similarly any net-to-telephone application, WAP [key-in passcoded]similarly any dial-in server acting as a relay account [separate service account passcoded]
any web-browser to portal [standard capability, preferably https-secured key-in passworded]
any telephane/vidlink [advanced capabilities: key-in passworded]
direct input speech
voice in video attachments
e-mail [text-to-voice and voice-to-text]
t-mail [standard telephone voice]
v-mail (if video-capable) [telephane/vidlink]
a correspondences tree
an archive tree
drafted compositions save [distinguished in the main mailbox]
auto-forwarding [preventing account overflow]
RECIPIENT SIDE [telephony model base]
Telephony standards have developed spanning a century of technologies:
The relatively special cases must be accommodated: AGC reaction circuits;
AGC squelching threshold; Eb/N0 SNR and signal data rates:
Answering a call, the telephone recipient hears first
a swelling flurry-rush of full frequency range tones (in minor key)
resolving (into major key) under a quickly rising chord formant (the chord, steady):
This is a quick non-vocal introductory cue that
pre-balances the circuit frequency equalizers,
pulls-in the AGC without triggering crow-bar clamping,
and gets itself out of the way of the ensuing first vocal tones
(as low frequencies are slowest to clear the audio channel).
The clip-remind tinklings are distinctive non-vocal 2-tones.
Parties are also tinkle-out notified when others disconnect.
The call recipient non-user without phone-mail may have the [to-be usual]
anytime-programmable "call-me-later" function setting,
to delay incoming calls for an additional [10-60] minutes to,
or to a specified time for reception -- this is distinct from busy or no-answer,
in that it acknowledges to the caller [process] an estimate or revision of
the recipient's current schedule:
It can be set in steps of 3.6-10 minutes. [1 hectahesit = 3.6 minutes]
Recipients (especially non-users) and messages [may] also have
identifying or temporary pass-codes to retrieve and continue prior messages
on the caller's server: as though they had the full service whenever they speak
with any user. This is called, Active-Demonstrator.
Recipients (especially non-users) may have multi-level group passwords:
although this lends to commercial, it permits selectable recipients
access to reply to the larger group: and longer replies.
Recipients' extensioning is handled by interactive pause-dialing.
Recipients must be able to hear text and all audio attachments.
Recipients may fast-scan through recorded conferences.
Recipients (as well as speakers) are given time alerts: especially the last 5-15 sec.
Recipients select translations: languages; educational downfiltering. [*]
Recipients can research: auto-glossary; concordancy; header indexing; up-down-stack response-threading (cactus and back-noted connections); diction correlations.
Recipients (as well as speakers) aids: auto-dictionary-thesaurus-wordfinder. [**]
Auto-Transcription: item-identification (graphics);
Prompted Header insertion: Time-Date; Occasion; Purpose; Auspices; Audience; Topic;
Auto-Header: Response-threading; Response-stacking (ting-tings); Time-Date;
Protocol Header: Server; Location; Recipients; Speakers; Addressee; Subject;
Cueword side-noting: Keywording; Associating; late-/sub-insertion of Header elements;
Moderation: single, multiple, primary group, whole audience; relevancy ranking;
[* translation may involve
keeping all nuances]
[** a much improved utility dictionary-thesaurus-wordfinder with correlation-extensions]
When initiating or replying to a telephone or telephane message, and when forwarding
a recorded vidlink conference session to other non-attendant parties,
the transmission schedule convenience is afforded by-recipient,
as designated in an address book or correspondents' itineraries book:
The session recording is cached either on the sender's side,
like a web-page for any qualified reader to accession,
or as "tv-mail" multiply spooled on the
receivers' sides for delivery to the recipients.
Each recipient is sent a notification, a call or page.
The preferred times for notifications, and re-notifications,
other than at the recipients' individual check-in for received e-mail,
are set to the sender's private information on the recipient,
or by the recipients' services more often [or spontaneously] updated --
inter-schedule-handshaking is accomplished by automatic protocol.
Typical notification times are the next available successively among:
- Immediately upon hang-up [direct immediate is the live-conference mode]
- Next day same time as originally received, same place
- Next work-day same time as originally received, same place
- early morning, initially 6:30AM, at home
- morning work, initially 9:00AM, at work
- lunch time, initially 11:30AM, at work or cell
- after lunch, initially 1:30PM, at work
- afternoon, initially 4:00PM, at work
- evening, initially 6:30PM, at home
- late evening, initially 9:00PM, at home
- any specified hour-and-minute [or hectahesit = 3.6 minutes]
- any re-specified hour-and-minute [or hectahesit = 3.6 minutes]
- any specified selection of such hours or hours-and-minutes
ECONOMICS [system implementation]
By introducing the telephony mail model first as a software system
operating fully cached from either side of a conversation,
product interest spreads most rapidly as both users find it facile:
Individual calls-placed have the full-cache advantage,
operating transparently to both users -- the non-user is even invited
to call-(back)-again for a replay,
or further development of the messaged conversation:
Thus the system acts as a per-message or per-connection installed service to the non-user:
This is nominally equivalent to effectual word-of-mouth advertising
implemented by automated computer and facile operational demonstration.
Such a single-side system can also handle the local telephone conferencing mode,
but distance conferencing responsivity deteriorates at longer delays
and depending on the full-duplex capability of the [typically satellite] telephony circuits.
Longer lunar delays are not likely to involve mere telephony.
Telephany and vidlink will have the full capabilities early-on;
And by time more expensive full-duplex video hardware installation begins,
the telephony clientele base will have extended to reach and involve
most telephone correspondents and potential conferencees in the simple
mechanics, methods, and utilities of the tele-phane/vid-link model.
Product market saturation can be pre-estimated in the venn-diagram
of potentially modern appliance (costed) utilization interest,
times the system usage facility persistency adamancy fraction:
Being, the percentage of telephony users who can affordably appreciate
and demand its availability for regular usage: eg.
Business users may quickly decide to acquire (guarantee) the service
for themselves, while residential users may more gradually find its
community empowerment a necessity, as its luxury adjusts to pragmatic ... well-above 50%.
As the system entails essentially initial software development,
costs are already comparable to contemporary telephony-e-mail services.
Capacities include HTML-like interactive language constructs
facilitating computer-human text-and-voice interfacing: To illustrate:
"<onConnect in:catchstart.wav onQuiet> Hello! You have a special
offer from <out:salesrep.ra> What is your name?
<in:guest.wav inlength=2 sec. wait(2 sec.)=quickexit
<play:muse.mid> Hello, <out:guest.wav>
We have a great deal for you from the makers of <wishwash.ra> description
<quickexit=excuseus.ra> <exit=> <goodbye.ra disconnect>
DOMAIN INTEGRATION [in situ hardware]
The current personnel communications domain is a convenience-compactified trellis structure of Equipments,
Stationary (home, office, wall, desk)
Mobile (walkabout, vehicular)
Handsets (earpiece, mouthpiece, keys)
Advanced Handsets (earpiece, mouthpiece, keys, graphics)
Headsets (monoaural/stereo earphones, mouthpiece, keys)
Facesets (speakers/earphones, mouthpiece, keyboard/keys, graphics/video)
Wristsets (speaker/earpiece, mouthpiece, actuator buttons)
Vidcard (speaker/ear-mouthpiece, keyboard/keys, graphics/video, additional filing, scanning, GPS, etc.)
Vehicles [cars, boats, planes, remote set-ups]
Public [payphone, guest phones]
Linking Via (remote or built-in)
Telephone Base (signal in-out, ring)
Television Component/Base (signal in-out, ring)
Transceiver (signal in-out, ring, radio transceiver)
LAN Switcher (multiple signals in-out, PA, trunks in-out)
PC Terminal (including potentially all of the above)
Local RF/UHF/IR (LAN, Cellular)
Cache (temporary message storage)
Archive (permanent message storage)
A major improvement is the capability to include video on otherwise slower data lines: Eg. Home PC Terminals can have their own Cache and Archive, thus allowing off-time downloading and reply-to of higher quality video messages. Similarly walkabout sets. With the vast majority of telephone calls being brief minutes and dependent on availability of parties, this off-time caching greatly increases throughput and visual talkability.
And since either side supplies the features, this is a self-advertizing addition to the
state of the art telephony.
The Tele-Phane/Vid-Link distance conferencing system is an application
and global superstructure intermeshing extant telephony and Internet,
and interfacing and extending to point-of-connection video-telephony,
PC-powered-telephany, and full vid-link conferencing (built on the telephane model),
accommodating low-bandwidth audio and hi-compression digital video
[see movergence], and higher bandwidths,
in an evermore facile service to communicant, correspondent mankind.
A premise discovery under the title,
© 2001-2005 GrandAdmiralPetry@Lanthus.net
Lanthus Corporation is a N E M O directive project