Monday, August 11, 2014

2014 VIDEO: Voice centered shooting plan, not what you expect. AMP video notes

Voice centered shooting plan, not what you expect.
AMP video notes  
Scroll down for News and Notes (including articles about new TOS)
AMP (Avatar Media Productions) has developed this plan to produce the best audio quality and the best video presentation of avatar lip movements combined with facial expression.  

Shot labels and shot lists

  • Coding for audio and visual files.
  • The coding is used to prepare the shot list.
  • Each speech by a character is planned as a separate set of shots.
  • Usually, three shots are taken for each speech.  
  • Shots may be taken from various angles and distances. 
  • See the coding plan for details of possible shots.
  • Emo director adds notes on emotional expression as needed.
  • Shot lists will be planned by the director team, consulting with others as needed.

Voice capture

Capture: is done by the voice engineer over a Skype computer-to-computer call.  At least three instances are produced with the voice engineer directing to get defect-free samples with a variety of deliveries.  
Labeling: labeled by the voice engineer in this code:
Review: The production team reviews the voice captures and selects one to be used for video capture. 

Video Capture, Voiced session

Voice sessions are set up exclusively for capturing the face of a speaking avatar delivering a line.  What is captured in the shot session are the mouth movements in sync with the voice.  In subsequent editing, the sound clip will be aligned with the mouth movements to produce a final clip that will present that line to the audience.
- Also captured in a  voice shot is any desired facial expression on the avatar. Depending on the shot list instructions, other characters or things may appear in the background or to the side.   Keep in mind that a single speech, as voiced, may extend over several video clips, showing the speaker, the reaction of hearers, and perhaps some event related to the story.   

Although the voice capture from the voice session will not be used in final production, voice is captured to help in identifying the clip and in placing the voice clip in synchronized position 

Planning of the shots is done by the director, voice engineer, camera engineer, and emote engineer.

The final list is copied into a notecard.  From this, the shot manager makes a series of panels, one for each item on the list.  Each panel carries the code (above) for the corresponding item.   These panels are put near the set (but out of camera sight) in the planed order of capture. 

AMP makes the signs with a sticky note script by Maria Korolov (Available on Kitely market.  She says:
Oh, I almost forgot to mention — this is licensed CC0. I hereby affirm that everything is completely original, and that anyone can do anything they want with these scripts, including using them commercially or reselling them. It’s all fine with me.
The voiced session requires only the voice engineer as speaking avateer.  Other avateers, if needed, can usually be managed by any volunteer or, possibly by Non-Playing Characters (NPC's)

Each shot is initiated by the Shot Manager, who puts a sign with the clip label (Coding for audio and visual files) on camera before the shoot starts.  When the Camera Engineer  calls "rolling"  the shot manager calls out the shot code and removes the sign, putting it in sequence in a place reserved for completed clips.  

Voice engineer takes cue from the shot manager' removal of the panel and plays the voiced speech.

When speech is done, shoot manager calls cut and repeats the shot code. Camera director stops the recording and renames the clip with copy-paste of code from the notecard copy of director's list.

The Voice Engineer runs each speaking avatar in the scene, feeding the prerecorded voice through the avatar. Each speech will be captured in close-up, so that lip movements will closely match the voice.  The avatars are positioned by pose balls so that the positions can be replicated. The same prerecorded voice clip will be reused for each of about three captures of the same speech.  

People needed for the voiced sessions:

  • Voice Engineer 
  • Camera Engineer
  • Shot manager (Volunteer, handles clap board to label the clip)
  • Emote director (handles avatar expressions)
  • Director and/or assistant director
  • Technical Expert  (may be on call)
  • Animation expert  (may be on call)
Speeches may be captured at several angles (left, right , and face-on) and at  several levels of close-up (face-only, upper torso and face, full body).  These may be captured by one camera with multiple reads or with one read and multiple cameras.  If multiple cameras cameras are used, each camera should take a shot of the color panel in the set so that the captures can be matched for color balance.  

The experience of AMP is that it is difficult to get several captures (even using the same viewer and version) to produce reliably matching results, probably because they have different video cards.   So we use a single viewer and replicate the speeches as needed.  Since the avatars are on pose balls and the voices are recorded, the replications give an excellent match.  

Standard shooting sequence for voice captures, assuming one camera

  • Select one character
  •    Select one shot angle and distance
  •       Read through each speech that requires that shot angle and distance 
  •    Repeat the above for each shot angle and distance needed for that character
  • Repeat the above for each character.

Settings to be recorded for all sessions

  • Windlight 
  • Time of day 
  • Standard camera settings 

Voiceless sessions 

Voiceless sessions are mainly used for these purposes:
  • establishing shots
  • avatars in conversations
  • the scene with avatars in position
  • action
Voiceless captures are generally planned from the script or specified by the production team after the voice capture is done as needed to support the voice capture.   

Voice is usually be captured for aid in identifying the clips.  

People needed for the voiceless sessions:

  • Avateers who have practiced any actions needed. 
  • Enough avateers to run all the avatars,
  • Camera Engineer
  • Shot manager
  • Director 
  • Emo director (probably)
  • Technical Expert (probably)
  • Animation expert (possibly)

Post capture

Post capture requires upload and should not be started till all captures are done for the day.  Uploads may take several hours and will require substantial bandwidth.  Camera director sends each ckip to the NMA repository. 

After verification of successful storage in the  NMA repository and backup, Camera Director may delete the character folders from the capture computer
Avatar Media Productions is a project of New Media Arts, Inc.a nonprofit organization tax exempt under IRC 501(c)(3).  You may be able to deduct contributions on your tax return, depending on your tax  circumstances.

News and Notes

Events in the virtual worlds  


      Contests and dated calls/invitations

      The Journal of Virtual Studies currently has a call for papers open.  The journal is peer reviewed and focuses on the use of virtual spaces for knowledge immersion.  Do you work or research in Second Life,  OpenSim or any other virtual platform?  Have you been engaged in research, exploration or reflection when using MMPORG’s?  Have you discovered a virtual space with knowledge immersion possibilities or have you engaged in a fascinating new way to explore the virtual through any number of immersive online applications?    
      These and many other questions are explored by the JoVS.  The Fall 2014 issue is expected to publish in September.  For details, see here: Journal of Virtual Studies 
      The theme is 'TRANSCENDING BORDERS.' Artists & filmmakers can interpret the theme as they see fit. Whether transcending borders between time and space, past and present, one nation and another, one language and another, a culture or civilisation or any other of the innumerable borders that present themselves as we navigate our lives. You the artist... You the machinimatographer...You decide.
      Entries Close at Midnight SLT 31st October 2014. Winner will be announced in December 2014.   First entries 

      Newcomer Help in Virtual Worlds. 



      Video-Machinima in virtual worlds

      No comments:

      Post a Comment

      Note: Only a member of this blog may post a comment.