G; Speech controls as sliders in detail/ Verb Moods

Every audio is a cue, and, the program, whence activated to record by audio, steps back to the area around the begining of the cue.



B

The Subjunctive in language.
https://web.cn.edu/kwheeler/grammar_subjunctive.html

Here is a subjunctive sentence: If I were a butterfly I would have wings.
The subjunctive is very rare in English. Where it might occur in speech the most is in TTRPGs.

The trickier is parsing when to transit out of the Game into the Table Top scenery, spontaneously.
This need not happen often, in fact, as part of the sequence the jumping out at the right time to make a humorous joke or resonant/relevant shift in imagery....
this can be done when the word "I suggest..."

The Imperative occurs on all three levels, the Imperative, the Interrogatory and the Indicative.
That the subject does not appear in the sentence sometimes...is a problem. The subject may have shifted...the sentence before is meanwhile in the indicative of the moment of the game play, the next sentence the subject disappears...hence, it might be directed towards a player or the GM, and outside of the moment of game play.
So this is marked as so...somekind of tag occurs when the Subject in a sentence is not referred, and of course the actual tone of voice is the imperative...so that the Main Render can include all three contexts, wherein, the motion of the face is applied to all there models...the Face rig in Avatar...the Figurine model and the model in action...as someone is more likely to be saying something in the imperative when they are doing something strenuous.
(This might establish a useful rubric of the simplest sentences spoken, or, when the speech tends to the Oral, indicating other steams, most of the time more than two steams, to be included in the Uber Render...the Main Render can be all three at once as interactive anyway (just not just based on number of cameras to see the scene again...).
The Uber Render is a randomizer; subsequent custom renders and interactive live streaming tweaking are enabled by having a datastream that has embedded multiple streams but even these contain the narrative and are subject to pre-edit, or the program I am here developing.
The Main Render is just what is rendered, in almost all cases which I instance here in the writing its the "First Render". But I just refer to it as the Main Render.
The overall data is pruned by the program.

What I mean by Interactive....as in shifting the camera to see it from another perspective and slowing it down...these kinds of things are enabled by having this massive database...yet, having some of that database...half of it, deemed "garbage" by just being left out and discarded. Garbage collector is a Java idea.

The Indicative is a within world speech. In other words, when the Actor is detailing the world as it is imagined, describing the actions they will later mocap, the indicative mode is here.
Then, moving out, it is not the indicative mode, but one of the other three. Then, it moves back into the Indicative...based on formality or on triggers.
The Indicative is almost always the most in game...one can call it the mode when the Actor uses the voice of the character in the scene. Except, in that Indicative case, the voice is a trigger, but might not then be indicative, but Imperative.
Nontheless, in game voice work is always recorded.

The Interrogative is most often the OTHER speech...that is, the Avatar speaking to the GM in order to fill in details. This is also recorded, but the automated EDL will cut out most of the banter.

The Exclamatory is also included, as previously defined.

The Conditional, with Might, Could and Would, do form a switch to the Avatar and the interplay with the team and GM over the table.

the conditional within voice stays in the scene,
but again, the voice has a lot of cuts away while using voice as the Avatar, and conversely, the Character in game might have the Actors/Avatars voice in some scene.
Each probably best visualized as one half of each...conditional within is 50, suddenly cutting to conditional (and then set within a non voice time frame and other constraints) 50.
While in voice out of the game play, 25% chance of cutting away.
In Avatar voice while in the game play: 12%.

But these are not quotas, just factors which contribute to constraints and not really part of the program, just a way to visualize how the data should resolve.

Setting up a speech training...a buzzer sounds when not in voice mode yet still in the scene. A buzzer sounds when getting into the scene, you are not in voice...and buzzer sounds when you run words together too much, thus, speaking too quickly for the poor old computer...

Repeating phrases, but each time with a different colorization. This is a exercise but also the penalty for having a buzzer..you must execute five different tones and inflections on a single sentence.

There is a time when you can be recording all three AND the one that is in the Main Render, as the first presented render...one speaks in the imperative, causes an interupt, and continuing in the Imperative, switches to two of the three, and then continuing switched to the other.

This is done to the beat of four if it can be.

All things that occur are mirrored in the Character Render, which is most of the Steams...again, as one is concentrated on, the others are set to duplicate as well across the "spectrum of three", and then dissolve along the same path, from three, to two, to one.
Except, the end path for the Actors who are not controlling the Main Render, they might end up for instance, in a jump cut, as the Avatar at the table, in voice, as the Main Render includes them as a cutaway, even though the end point for the Main Character who is in the spotlight of the Main Render, controlling all three streams by maintaining the Imperative (it drops the Table Top Figurine, and switches to the cross cut between In game world and Table Top, at the third sentence).

This can be sustained if the other Actors maintain in game speech as established by the Character taking the Imperative first.

Though, they might be immediately upstaged, this brings in a buzzer.

When a Actor ceases to speak and allowed the one second footer of a pause, while they pause the program, in the later Uber Render, is in search mode, and so there is a camera order frequency that is an insert configuration that is first a reverse of the GM customary order around the table.
As the GM goes to each Actor in turn, as a routine, so the does this reaction to a long pause do the same, as in reviewing reactions.

People should get used to be in character as a reaction...but those who are not are indicated by bad or non existent facial capture. Sometime the motion capture wont work, but in this case its very subtle changes, and the program will only allow two...of course, inclusion of the name and then a pause...(this pause is intuitive on the part of the team) then means that Actor's Avatar as named is the first Reaction cut, and the next is...