Boo-yah! I've been following the facial mocap scene for the past two years and it's so exciting there's finally a truly free workflow from beginning to end. This is going to open the door for so many people 🥳
This video should be part of YT recommendations for any artist or developer. Great work you're doing with animation and mocap. And a big thank you to all devs who made these two addons available for free.
@@CGDive Indeed. Would you show the proper way to bake and export the facial blend shape animation to FBX? I wanted to import the csv in Blender and send it to other DCCs. Thanks. Keep it up.
This is so excellent! I have been searching for months for a way to get mo-cap from pre-recorded video. The Audio2Face workflow was too complicated for me and I definitely prefer something integrated into blender. This was perfect!! I already have FaceIt and had the model pretty much set-up. I followed this tutorial tonight and within no time I had a result. I was surprised how well it captured the head tilts. In the video, the person was wearing glasses and had a lot of glare on the lenses, and surprisingly the software caught most of that movement. I have a lot of clean-up to do but this works great as a starting point. Thank you so much for making this tutorial!
Hi there, first try I did it, was successful, then tried again, faceit and live link face and I got this error multiple times ! Python: Traceback (most recent call last): File "C:\Users\myname\AppData\Roaming\Blender Foundation\Blender\3.6\scripts\addons\faceit\mocap\mocap_base.py", line 842, in execute for x in futils.get_action_frame_range( File "C:\Users\myname\AppData\Roaming\Blender Foundation\Blender\3.6\scripts\addons\faceit\core\faceit_utils.py", line 12, in get_action_frame_range frame_range = action.curve_frame_range AttributeError: 'NoneType' object has no attribute 'curve_frame_range'
Thanks for your video! I've been following the channel for a long time and this is one of my favorites. IMHO This is the best channel about creating animations. You do a great work!
Thank you so much for this video ,You have solved my biggest problem as well as doubt or issue ,from past two years i was wondering about the facial expressions in animation. Thank you once again god bless you
Hi, thank you. I am facing a problem. When I try FaceIT, it reacts in the opposite way when I turn my head. When I turn the head left/right in the video, the character's head turns up/down in Blender. So it reacts in the opposite way - in the video I shoot the head up/down - in Blender it turns left/right. I am using auto rig pro with faceit.
Is there a way to do live capturing on the app without an iphone tho? bc I have only seen video do cvs and then you import but live capture only has 4 apps and all of them are iphone exclusives
This is great! I think a great addition to the suite of tools would be a way to scale up the responsiveness of the captured rig points, i.e. selecting the brows, lips, jaw and setting a multiplier for scale for more intense/detailed performance capture. Is the capture framerate based on the set framerate within blender or does it capture at the camera framerate and interpolate frames? Either way, I'll definitely be grabbing the add-ons to play with! I'm working on a character in unreal engine and so far I've had to just keep the face static.
This is so really good! Both as info and tech... but it's still seems like it could use some manual tweaking after because it didn't seem to capture all your acting as well as it could have
Yeah, its not great. Google mediapipe released a lightweight face and landmark recognition model. That is what I used in the app. They have better heavy models not accessible to the public. Some commercial products possibly train their own models or purchase usage of the better google models. But hey its free (the app creator here)
Thank you so much >> you still the best and you save the world , I am looking now for hand capture with finger movement .. I could not find it.. I hope you can help us (all the world) in this matter after you solve the problem by a nice video ( the above).
Hi! When I transfer a model with a faceit rig, the faceit addon stops seeing the rig. The ability to animate with LiveLinkFace is lost. How can I connect the addon to the ring if it has lost it.
hey guys just curious how does this quality compare to iphone mocaps? if i use a webcam or android phone (without depth sensor), is it still comparable quality to iphone mocaps? or slightly worse?
There's an error showing up while importing the CSV file. It says, ValueError: could not convert string to float: 'V'. I put a mov file on the load audio section and a csv file on the csv import section. Kept the frame similar to the mov file. Then it showed error when I clicked import csv.
Thanks! I have looked at it, and do not see why it can't run on Linux or most other systems as it made in Python with numpy and Tk, but I'm not much of a programmer either and can't get the libraries to work together. Anyone here up for the task?
Hi, this is amazing but I am wondering if this also works without faceit? It's quit expensive for starters. I'm using daz3d models that comes with a face rig as well or I can manully ad a face rig with the metarig from blender. Will that also works?
thanks for your quick response! I'm sorry if i missed the answer within the tutorial. I did watched it and the full tutorial is applied on a faceit-rig but yes I skipped some parts so again sorry if the answer is in there@@CGDive
It didn't ask me to import a video file. It just launches my webcam and starts to track live footage. I don't know where to find the csv file when I stop the tracking.
Good day guys , am a new subscriber ..i howsoever downlaoded face landmark but it shows two options saying (mild filtering ....) and the second which says (aggressive filtering ) when i try to install it , and neither of this options seems to work out ....please assist .
I can't say why. The links in the description box seem to be correct. The mocap extraction app github.com/Qaanaaq/Face_Landmark_Link/ The blender addon: nickfisher.gumroad.com/l/tvzndw?layout=profile
Thanks so much for the tutorial. I used the free addon. The csv file loads with the keyframes inserted but has no effect on the character. What could be the possible solution
@@animbenard9747 I don't use CC so I can't say for sure, sorry. I believe CC character do have ARKit shapes or at least something similar so you have to make sure that you are bringing them into Blender. And also that the shape names are correct. Good luck!
For Auto-Rig Pro face rig? No, it only works for a specific set of shapekeys, the so called "ARKit" shapes. Some info here ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-KQ32KRYq6RA.html
Dear Sir/mam, I am facing ' The specified file is not valid' error while importing Mocap. Please guide me as the addon without this features is worthless for me and i just purchased it after watching this video. Thanks
Hello. You didn't specify which addon you mean. But even if you did, these are not my add-ons. I am just make tutorials about them. Please contact the developer from whom you purchased the product that you are having trouble with. Thank you!
there's a video about using Nvidia Omnverse Audio2Face that will generate the ARKit blendshapes. But I'll say -- it's not easy. I got the blendshapes out to Blender, now trying LiveLinkFace -- 1st time, animations do not play
I can't say. It could be too many things from literally using an invalid file, to blender versions, addon versions, using the wrong feature, etc. etc. Try again. If the problem persists, you can try to share a bug report with the related tool developers.
@@CGDive thank you for the reply.. I emailed the faceit developer.. and together with the csv file.. for now I use hallway cubes osc stream function, just worry the load on my pc will be too greats since I use brekel body with 3 kinects for body motion capture and hallway cube for facial motion capture together.. by the way thank you so much for providing such detailed great quality tutorial