I recommend the 3Dtest Youtube tutorials

iClone Faceware Facial Mocap Tutorial - Intro of iClone Facial MocapPlugin

 https://www.youtube.com/watch?v=6SCvpEMEZu0

Excellent demo of possibilities from a master hand.

iCloneFaceware Facial Mocap Tutorial - Headcams

https://www.youtube.com/watch?v=WNShqaKPU6I

iCloneFaceware Facial Mocap Tutorial - Facial Mocap for Lipsync

https://www.youtube.com/watch?v=ucHFPh5cBnc

iCloneFaceware Facial Mocap Tutorial - Facial Mocap Real Practice

https://www.youtube.com/watch?v=oOCOmTMBT4Q&t=2s

Sean

Thanks for the info, will check out today...

All the tutorials can also be accessed from the Grand Launch page here:

https://mocap.reallusion.com/iClone-faceware-mocap/tutorial.html

Two things to be aware of when watching this. Characters have a HUGE impact on the performance. The closer the Character is to the performer the more natural it will look. With stylized characters like they show in the demo you might have to tweak the influence sliders to ramp up or down certain blend shapes. 

Lip stick can help with lip tracking. also having a good 5-10 second of neutral face before staring face performance has a massive impact on tracking. 

That said, these demos look very good!

Free Range
My IMDB

@schemer. Thanks for that. At time of posting there was no launch page.

I'm curious about this comment.  I realize it's good to get a neutral face capture, and having enough time to do so is important, but since the capture button works more or less instantly why would having 5-10 seconds be any better than, say, 1-2 seconds where you can actually capture it?  This is particularly important to me as I'm using recorded sequences, and I find I can get the neutral face capture with just pausing the video and doing the reference with it paused at a key spot -- am I missing something by not having the video play through?  (I'm using a virtual webcam).

---

Alienware Aurora R12, Win 10, i9-119000KF, 3.5GHz CPU, 128GB RAM, RTX 3090 (24GB), Samsung 960 Pro 4TB M-2 SSD, TB+ Disk space
Mike "ex-genius" Kelley

You method probably works just as well since you are pausing on the neutral face. For live streaming capture though we found in our test to giving it it a bit of neutral face at the beginning before going into the performance had a large impact on performance. A few seconds is probably enough but 5 or so pretty much guarantees face lock on. 

All it needs is a solid neutral reference to get started so with your method and pausing on a neutral face frame you are giving it good solid reference. 

even lighting helps a lot, the camera looking very slightly up at the face vs direct on or down also helps (you see this on all helmet cams), a little lipstick helps if you have lip color that is close to your skin color and if you want to go hog wild, even though it is a marker less system you can use markers. 

Free Range
My IMDB

@Free Range. Lipstick? Oh Jeez, Free range! Thank God for On-line purchase.

Well, YMMV, but I haven't found it to be at all necessary.

(And this old guy's not about to start experimenting -- I swore the last time I was on stage that it was the last time I'd ever put makeup on, and that was over 20 years ago).

---

Alienware Aurora R12, Win 10, i9-119000KF, 3.5GHz CPU, 128GB RAM, RTX 3090 (24GB), Samsung 960 Pro 4TB M-2 SSD, TB+ Disk space
Mike "ex-genius" Kelley

Haha, it really depends on the color of your lips. Some people have lip color that is pretty close to their skin color. It is all about contrast. If you have low contrast lips you will probably need makeup. If it makes you feel better most high-end face tracking uses makeup including lipstick (usually luminous) including ILM, Digital Domain, some Weta.

Free Range
My IMDB