World’s First 10-Person Live MoCap: CHINGMU Elevates Virtual Idol "Come Come"

Source：

2025 Jul. 29

When virtual idols evolve from solo performances to group acts, and digital live streams shift from one-on-one interactions to multi-character spectacles, the industry's expectations for virtual content have moved beyond mere visibility to demand authenticity and excitement. With Gen Z’s growing appetite for virtual experiences, group live streams of virtual idols are becoming the new frontier of digital entertainment. However, technical challenges such as lagging movements, model clipping, and loss of detail during multi-character interactions have long been hurdles for creators.  

Recently, the virtual idol group from the "Come Come" offline music special project broke through the limitations of traditional multi-person virtual broadcasts, successfully staging the world's first 10-person synchronized motion capture live stream. Ten virtual idols performed together on stage, elevating the realism and immersion of virtual performances to unprecedented heights. Behind this breakthrough was CHINGMU's self-developed optical motion capture system—technology that not only brings digital avatars to life but also infuses all ten virtual idols with the dynamism and vitality of a live concert.  

The backstage "stage" for this 10-person motion capture broadcast was CHINGMU's MCP (Motion Capture Production) base, located in Shanghai's Baoshan District at the 1919 Creative Park. As one of Shanghai's largest professional-grade motion capture studios, MCP is equipped with world-class high-precision motion capture devices, stable systems, and streamlined workflows. Staffed by an elite team of industry veterans with extensive project experience, the facility offers cutting-edge motion capture technology and expansive capture spaces, supporting everything from game animation to film and TV production. Here, even the most imaginative ideas can be efficiently and accurately realized.  

Precision Synchronization: 10-Person Live Motion Capture

Transforming a static virtual idol model into a dynamic stage performer with nuanced expressions requires high-precision motion capture technology. In virtual live performances, however, "scale" is not just a numbers game. When ten characters interact closely on stage, carrying props and moving dynamically, traditional motion capture systems often struggle with occlusion, signal interference, delayed movements, tracking loss, or even awkward model clipping.  

To ensure the realism and fluidity of digital avatars during live performances, CHINGMU leveraged its proprietary high-precision, low-latency optical motion capture cameras. These devices deliver real-time, high-fidelity data on body movements, facial expressions, and subtle details. Integrated with AI-powered motion capture software, the system intelligently calculates skeletal joint rotations and displacements, accurately tracking marker trajectories even in complex scenarios with overlapping limbs and prop obstructions. This ensures seamless, lag-free synchronization between performers and their virtual counterparts, meeting the stringent demands of live virtual broadcasts. More importantly, it allows each virtual character to showcase unique micro-expressions and personality quirks, creating a visually cohesive and natural experience for audiences.  

Fingertip "Micro-Expressions": Bringing Every Finger to Life

Beyond the technical complexities of multi-person motion capture, the most demanding aspect of virtual idol performances lies in capturing the subtlety of hand movements: heart gestures, microphone grips, instrument playing..... These details are critical for realism, yet the flexibility of hand structures and the complexity of their motion paths make fine hand capture one of the industry's most challenging tasks.  

To address this, CHINGMU's optical finger-tracking technology achieves ultra-precise hand motion capture. Leveraging breakthrough finger modeling, intelligent skeletal pose estimation, and data optimization, the system delivers sub-millimeter accuracy for each finger joint. It effortlessly handles everything from basic gestures like "OK" signs, crossed fingers, and clenched fists to complex multi-finger interactions such as playing instruments, turning knobs, or prop handling. This fingertip-level precision allows digital avatars to express individuality, making virtual characters feel more lifelike than ever.  

From body posture to facial expressions down to the tiniest finger movements, CHINGMU's motion capture system breathes soul into digital avatars through full-spectrum high-precision tracking. Whether for virtual idol live streams or virtual brand ambassador campaigns, CHINGMU provides end-to-end customized solutions—from digital avatar modeling and rigging to content production and multi-scenario interactions. The system supports offline data recording, real-time synchronization, and live rendering, ensuring digital avatars don't just "move" but truly "come alive."  

To date, CHINGMU has collaborated on over 30 virtual influencers, including TikTok's top virtual streamer Jin Ju, Kuaishou's virtual host Hu Li Li, Vietnam's Vtuber idol group NIJIGEN, and Bilibili's virtual idol group Four Xiwu. It has also powered major live events such as the REFIRE Hip-Hop Awards, the Tingchao Pavilion Virtual Music Show, and Virtual Fes Hong Kong VTuber 3D Concert, continuously energizing the virtual content ecosystem with cutting-edge technology.  

CHINGMU's motion capture breakthroughs not only push the boundaries of multi-person virtual broadcasts but also redefine how virtual idols engage with fans in real time. Soon, perhaps, discussions about virtual idols will no longer fixate on their realism but instead celebrate the emotional connections they inspire—the ultimate testament to technology's power in shaping virtual worlds.