How to Look Like Your Character on Camera: A Guide for Online TTRPG and Live Performance

Desktop AR apps built for live performers let you assemble a character look from a library of digital assets: armor, masks, effects, backgrounds. The app outputs as a virtual camera into Discord, OBS, Zoom, or anything else that accepts a webcam feed. You stay visible and expressive on camera. The character layer sits on top of you, not in place of you.

That's the category. Which tool to actually use is the harder question, because the search results lump together several different products that do very different things. We've been building in this space for two years, and the honest map of what exists is below.

What tools let you appear as your character on webcam for D&D?

Four categories of tool show up when people search for this, and each one solves a different version of the question.

Virtual backgrounds put your character's environment behind you. Useful for setting a scene, but they don't change how you look. You're still you, in a tavern.

VTuber tools replace your face and body with an animated avatar. The avatar maps to your facial expressions and head movement. This works well when you want a fully animated character on screen, but the human performer is no longer visible. For some streamers and performers, that's the goal. For most TTRPG Game Masters running live sessions, it isn't. Table dynamics depend on reading each other's faces.

Webcam filters (Snapchat-style, OBS plugins) give you one-tap effects. Fun for a clip. Not built for a four-hour campaign session where you want the same character across weeks.

AR identity tools are the category that sits between virtual backgrounds and VTubing. They overlay digital assets, like armor, hoods, horns, glowing effects, environmental elements, on top of your live face and body. You stay visible. The character layer is the addition, not the replacement.

If you're searching as a cosplayer or character artist rather than a TTRPG player, the same map applies. The AR identity category is what gives you a wearable fantasy outfit on camera without committing to a full avatar pipeline.

How is this different from a VTuber avatar?

VTubing replaces the human with an animated character. AR identity tools keep the human visible and add the character layer over the top. Both are valid for live performance. They serve different goals.

A GM running a session usually wants their players to read their face. Surprise, suspicion, the moment a player says something the villain didn't expect. That lands when the table can see it. An AR overlay preserves that. A full avatar abstracts it.

There's also a setup difference. VTuber pipelines often involve a custom-commissioned avatar, rigging, and ongoing software costs. AR identity tools work from pre-built asset libraries you assemble yourself.

How do you keep the same character look across multiple sessions?

This is where the question gets practical, because a one-off look is easy. The hard part is reaching for the same character a month later, or running an NPC-heavy session where you switch between three different faces in one scene.

The workflow that holds up across an ongoing campaign looks like this: build a character or NPC from a library of assets, save it as a named preset, organize presets into campaigns, and load any preset when you need it. Switch between saved looks mid-scene to portray different NPCs. Build a library that grows over the lifetime of the campaign.

This is the workflow Faes AR Pro is built around. It's a desktop app for online TTRPG Game Masters and live performers, $50, one license, no subscription. The build, save, load loop is the point: the orc warlord you authored in October loads cleanly in March, the innkeeper preset switches to the goblin lieutenant without rebuilding, and the library you accumulate over a campaign stays with you into the next one.

Get Faes AR

The reason we built it around save-and-load presets specifically: anything generative drifts. AI face replacement regenerates your appearance frame by frame, and that process degrades under lighting changes, different camera angles, and long session runtimes. You can't build a recurring character on shifting ground. Authored presets stay authored. The orc warlord you saved in October looks the same in March.

What do you need to set this up?

A desktop computer that can run the app, a webcam, and one of the platforms people actually run sessions on: Discord, Zoom, Google Meet, or OBS for streamers. The app outputs as a virtual camera, which every major platform accepts as a webcam input. Select it from the camera dropdown the same way you'd select your physical webcam.

What you don't need: a green screen, a custom-commissioned avatar, a rigging artist, or a subscription. The segmentation runs on the live video feed.

The first session is a one-time setup, working through the asset library to put a character look together. Loading the same preset after that takes a click.

How do streamers and pro GMs do this?

The professionals running paid sessions on platforms like StartPlaying, or actual play groups producing weekly content, generally need three things the casual user doesn't: faster preset switching for NPC-heavy scenes, custom assets that match their established characters, and visual continuity that holds up under recording.

The preset-and-library approach scales to all three. Switch between an innkeeper, a goblin lieutenant, and a returning villain in the same scene by loading saved looks. Bring in custom assets by uploading 3D pieces sourced from artists, marketplaces like Fab and ArtStation, or direct commissions. Run the same character across a 60-episode show without rebuilding.

The aesthetic ceiling is set by the assets, not the tool. Better assets, better look. The ARaura asset library is human-made and commissioned from named artists, with no generative art in any form. That matters more in recorded content than in private sessions, but it's the policy across the whole product.

Is there a free option?

A free version is coming end of summer 2026. It will include a starter asset library and access to the marketplace, where individual assets are available at consumer-friendly prices, meaningfully below what comparable assets cost on Sketchfab or Fab. Players who want one character for one campaign rather than a full GM toolkit are the audience there.

For now, Pro is the available option. If you're a player rather than a GM and you want to wait for the free version, it's coming, and it's a real free product rather than a limited trial. We'll link to it here once it launches.