Author
|
Message
|
animagic
|
animagic
Posted 3 Years Ago
|
Group: Forum Members
Last Active: 2 hours ago
Posts: 15.6K,
Visits: 30.3K
|
For this WIP thread, I'm posting the finished video first as it has already been completed. After that I will explain some production aspects. Sharon reprimands and warns young Rita (her daughter or perhaps younger sister, we don't know).
EDIT: Per an astute comment from Smit_R, redid the render with the cigarette turned around.
Edited
3 Years Ago by
animagic
|
|
|
animagic
|
animagic
Posted 3 Years Ago
|
Group: Forum Members
Last Active: 2 hours ago
Posts: 15.6K,
Visits: 30.3K
|
The two characters are left-overs from the 10-year Challenge, as is the set. They were very disappointed that they didn't win anything and were eager to try again. Here is their entry image: I imagined that the older woman on the right, Sharon, would lecture the young woman on the left, Rita. Only Sharon has dialog, Rita just listens, so I wanted Sharon to be in view. Initially, Rita wasn't even going to be there, but upon reflection ( ), I would have them both be reflected in a mirror. Like so: As iClone doesn't have a mirror yet, one option is to use a video as I did here. To complicate matters, I wanted slight camera movements (as with a handheld), so the refection would have to follow that. Years ago, Rampa published a camera rig for that, which I used (see https://www.youtube.com/watch?v=_POPVMjRflc). This gives me a setup like this: The challenge is to calibrate the view screen such that it overlaps the actual scene as seen through the Mirror camera. I double checked by duplicating the bar counter and what's on it and it is accurate enough. This has little to do with lip-synching and character development, but I wanted to make the scene itself interesting. And, the mirror is essential for the story. Lunch is served; more later!
|
|
|
animagic
|
animagic
Posted 3 Years Ago
|
Group: Forum Members
Last Active: 2 hours ago
Posts: 15.6K,
Visits: 30.3K
|
This certainly is a popular contest with some tough competition! New faces too, which is nice. I want to show in this post that aging a character is quite feasible with SkinGen. In addition, I used the Face Old morph (setting 33). The head was created with Head Shot. Here is the original version: The older version of Sharon is shown below, including the SkinGen components that have been used. Hair is the Wavy Short Bob. The eyebrows are Female Trimmed.
Edited
3 Years Ago by
animagic
|
|
|
Amper Sand
|
Amper Sand
Posted 3 Years Ago
|
Group: Forum Members
Last Active: Last Year
Posts: 731,
Visits: 5.2K
|
Excellent - gave it five stars, awesome work!
______________________________________________________________________________ Software developer. iClone hobbyist since January 11, 2020. Best work so far: Robots among us: The Lab and The sad end of a legend In progress: Black Mamba AMD Ryzen 3950X 16C/32T 4.2 GHz, 64GB RAM, RTX2080 11GB, 2TB Samsung 970 Pro SSD, 8TB Seagate BarraCuda HDD, Win11 64-bit, iClone 8.12
|
|
|
toystorylab
|
toystorylab
Posted 3 Years Ago
|
Group: Forum Members
Last Active: Yesterday
Posts: 4.4K,
Visits: 28.7K
|
|
|
|
animagic
|
animagic
Posted 3 Years Ago
|
Group: Forum Members
Last Active: 2 hours ago
Posts: 15.6K,
Visits: 30.3K
|
@ampertox: Thank you for the stars! @toystorylab: Thank you for your comment. There is one more thing to discuss. In this case I could generate the whole dialog in Replica in one take, as there is only a single speaker. I did a number of different takes, though, with a slightly altered text, until it sounded right to me. One thing to be aware of is that every take costs usage time (you can't preview in Replica). I did cheat a little by using "where" from the phrase "I know what you did, and where" from the previous take for the final one. EDIT: Notice that in the last take I had to written "Lissen" to get "Listen" pronounced correctly. You would expect the same phrase to be rendered the same in different takes, but for some reason it's not. I have no explanation. The Replica take was sent to iClone to create the lip-synching through AccuLips. I then used the Facial Animation editor for refinement, as well as motion clips for the body movement with some key-framing.
Edited
3 Years Ago by
animagic
|
|
|
wires
|
|
Group: Forum Members
Last Active: 2 days ago
Posts: 5.7K,
Visits: 26.4K
|
animagic (8/7/2021)
There is one more thing to discuss. In this case I could generate the whole dialog in Replica in one take, as there is only a single speaker.
I did a number of different takes, though, with a slightly altered text, until it sounded right to me. One thing to be aware of is that every take costs usage time (you can't preview in Replica).Hi Job, According to the information provided in the RL tutorial below it is possible to preview the spoken text and make changes before committing to a take, starts at about 1:25 in the video.
Gerry
System: Win 10 Pro (21H2), Asus X99-E WS, CPU i7-5930K -3,5 GHz, 32 GB DDR4 2666-16 RAM, NVidia GTX 1080 Ti GPU - 11 GB VRAM(Driver Studio-536.99), Samsung 850 Pro 512 GB SSD, 6 TB HD storage.
|
|
|
toystorylab
|
toystorylab
Posted 3 Years Ago
|
Group: Forum Members
Last Active: Yesterday
Posts: 4.4K,
Visits: 28.7K
|
wires (8/8/2021) Hi Job, According to the information provided in the RL tutorial below it is possible to preview the spoken text and make changes before committing to a take, starts at about 1:25 in the video.If you watch closely that video (top right corner), every time "generate" is used, the total amount of time is reduced, even if you don't change your text... If you delete a word and generate again, time will be reduced. That's not very userfriendly. In Acapela Box you work untill you are satisfied and then your credit is reduced...
Toystorylab on Vimeo : Crassitudes (my "Alter Ego") on Youtube:
|
|
|
animagic
|
animagic
Posted 3 Years Ago
|
Group: Forum Members
Last Active: 2 hours ago
Posts: 15.6K,
Visits: 30.3K
|
Yeah, I should have said that for every preview time is deducted. Thanks for the clarification.
|
|
|
wires
|
|
Group: Forum Members
Last Active: 2 days ago
Posts: 5.7K,
Visits: 26.4K
|
toystorylab (8/8/2021)
If you watch closely that video (top right corner), every time "generate" is used, the total amount of time is reduced, even if you don't change your text... If you delete a word and generate again, time will be reduced. That's not very userfriendly. In Acapela Box you work untill you are satisfied and then your credit is reduced...
I didn't notice the "small print" on the screen when I posted my answer. This looks like another ripoff deal , meaning that I won't be investing any of my money in this scam when the credits run out.
Gerry
System: Win 10 Pro (21H2), Asus X99-E WS, CPU i7-5930K -3,5 GHz, 32 GB DDR4 2666-16 RAM, NVidia GTX 1080 Ti GPU - 11 GB VRAM(Driver Studio-536.99), Samsung 850 Pro 512 GB SSD, 6 TB HD storage.
|
|
|