QuickMagic

QuickMagic's pitch is narrow and honest: turn the video you already shot, or a sentence you already typed, into motion data your 3D pipeline can consume. Strengths. The capture workflow is genuinely low-friction — upload phone or webcam footage to the cloud pipeline and get back editable animation data. Tracking coverage spans full-body, upper-body, hand and finger, and facial motion options, with single- and multi-subject support and both static and moving cameras. Frame rates run 24, 30, 60, and 120 FPS, and you get pose, frame-rate, and motion-cleanup controls plus a customizable first-frame pose and an anti-penetration system aimed at limiting joint interpenetration. Export coverage is unusually wide for this price class: FBX, BVH, BIP, C4D, VMD, Mixamo, UE4, UE5.5, UE5.6, MetaHuman, Character Creator & iClone, Roblox, OnlyFace, and UniRobot, landing in Blender, Unreal Engine, Unity, Maya, 3ds Max, MotionBuilder, Cascadeur, iClone, MikuMikuDance, UEFN, and Cartoon Animator. The text-to-motion generator produces editable drafts for previs and prototyping, and the Unitree G1/H1/H1_2 presets carve out a robotics research niche few consumer tools touch. Weaknesses. Quality tracks source quality — subject visibility, lighting, motion blur, occlusion, and camera movement all move the needle, so chaotic or heavily occluded shots need a cleanup pass. Facial capture is offered as an option, but there's no claim of finished facial animation out of the box. It's cloud-only, so teams that must keep footage on-premise should look at desktop or studio-grade systems instead, and it can't do live on-set capture where you need motion as it happens. Where it fits. Indie game developers prototyping in Unreal or Unity, VTuber/MMD/iClone creators wanting quick dance or performance motion, artists generating previs from text, and robotics researchers producing human-motion references. Charlie Driscoll's showcase — long-running smartphone footage retargeted onto a UE5 MetaHuman for a medieval battle scene — is the clearest picture of the intended workflow, and Defonten's room-filmed tests retargeted onto free Paragon characters show the same path. The Cascadeur team's own walkthrough (video mocap in QuickMagic, then Animation Unbaking and Fulcrum Motion Cleaning, AutoPosing, AutoPhysics) is a realistic look at how much polish happens after export. Where it doesn't. Live on-set or XR capture, projects demanding zero-cleanup facial performance, and heavily occluded or motion-blurred shots. Export availability is plan- and workflow-dependent, so confirm your target format before you commit.