How to run virtual casting calls

Define the casting brief

Everything downstream depends on the brief, because a brief is what turns a pile of individual auditions into a comparable field. Before a single invite goes out, write down five things: the roles you are casting, the specific looks or types each role needs, the usage the final work carries (a national campaign, a regional spot, an editorial), the deadline the client is holding you to, and — most important for the session itself — exactly what talent must show. Spell out the reads, the wardrobe, the angles, and the slate you want, so nobody improvises a different audition.

The brief also sets the technical floor. Decide the framing you will hold for every candidate (a slate plus a set of clean digitals, say, or a specific two-hander read) and whether you need stills, video, or both. When you know the deliverable up front, you direct toward it instead of collecting variety you will never use. A tight brief is also what keeps a client aligned: agree it with the producer before the call, and a disputed shortlist later becomes a question of who best fit an agreed standard rather than a matter of taste. This is the same discipline that a professional virtual casting session runs on — the brief is the spec, and the session is you applying it identically to everyone.

Invite talent to a virtual room

With the brief locked, open a room and send one join link. The split that makes a virtual call work is that only the talent runs the app — the phone in their hand is the camera — while everyone watching joins from a browser with nothing to install. That means agents, producers, and brand clients click the same kind of link and are simply in the room, no account juggling, no download.

Batch the invitations rather than opening the floodgates. Group candidates into short, back-to-back slots so an open call moves at a steady clip and nobody sits in a digital waiting room. Send talent a two-line prep note with the link: find a quiet space, prop the phone at eye level on a stand, face a window for light, and have the CLOS app installed before the slot. Because the rooms are access-controlled and encrypted, an unreleased campaign brief and the faces attached to it stay private, which matters when a client hasn’t announced the project. The scheduling mechanics scale the same way for a large virtual open call: one link, one brief, one operator moving down the list.

Direct each session live

This is the step that separates a casting call from an inbox of clips. When the candidate joins, you see their live feed and take control of the camera on their phone before the read even begins — you set the frame, pull focus, and lock exposure yourself, so the slate is composed the way your brief specifies rather than however their handset defaulted. Then you direct like you would in a suite: give the read, push a second take, adjust a pose, steer a line, ask for a profile turn. The talent never touches a setting; they act, and you run the camera.

Direct with your voice more than you would in person, since you have no body language to lean on. Give one instruction at a time, call the moment you capture so they hold, and keep the energy up between takes to get a natural expression. Because you hold the same controls for everyone, the twentieth candidate is framed and lit to the identical recipe as the first — which is the whole point of directing live rather than collecting tapes. On Pro you can also switch to the telephoto or ultra-wide lens and hold white balance steady, so a close read and a full-length both come out clean. You are coaching a performance and guaranteeing the capture at the same time.

Capture stills and video

A live call should produce finished reference, not a screen recording of a video call. CLOS triggers the phone’s real sensor, so what you capture is a genuine full-resolution photo or a real video clip — sharp enough to judge a face closely and clean enough to drop straight into a client shortlist. Fire stills for digitals and slates; roll video the instant a read is working. Decide in the moment which the role needs, guided by the brief.

The capture lands in CLOS Cloud as it happens, not after the session. That real-time flow changes how you work: you can glance at what you have mid-slot, notice a soft frame or a blown-out window, and re-take it while the talent is still in position — instead of discovering the problem when you review tapes that night and having to re-request the whole audition. When retouch-grade reference matters, Pro captures RAW and Apple ProRaw off the sensor. Video runs H.264 or H.265, with ProRes available on Enterprise for productions that need it. Either way, the file you keep is the real thing, captured to your standard the first time.

Review and share selects

Because every take is already in the cloud, review is not a separate export-and-email phase — it happens inside the session. Favorite the keepers live as you go, so the shortlist assembles itself while you cast rather than after. Clients and producers watching in the room react in the moment: an agent confirms their talent is comfortable, a brand flags a face it likes, a producer weighs in before the slot ends.

When the block wraps, you share straight from CLOS Cloud — a single select, a curated selection, or the whole block packaged as a ZIP — as a link the client opens in a browser with no account. On Pro the albums don’t expire, so a board built for one brief is still there when a second brief comes in and you want to re-pitch the same faces. The practical result is same-day selects: a casting keeps moving at the speed a production needs, instead of stalling for a day while tapes are collected, transcoded, and chased down an email thread.

Self-tape vs live direction

Neither format wins outright; they solve different problems, and good agencies use both. Self-tapes scale. When you need to filter a huge field down to a manageable list, asking talent to record against the brief on their own time lets you cover more people than any live schedule could. The cost is that self-tapes are undirected and uneven — no two are framed, lit, or performed under the same conditions, so you are comparing submissions as much as talent, and you cannot push a second take.

Live virtual direction is the opposite trade. It costs a scheduled slot per person, but it gives you direction and identical capture for the candidates who matter — the shortlist and the callback — so the decision weighs the performance and the look rather than the quirks of someone’s phone. The common pattern is to use self-tapes to filter and live sessions to decide: cast the wide net asynchronously, then bring the finalists into directed rooms. Worth knowing is that CLOS covers the self-tape-style end too — AI CAM gives a solo candidate real-time framing guidance so they can direct their own capture and still produce a real, full-resolution file to your spec, rather than a shaky handheld clip. For a fuller breakdown of when to reach for each, see self-tape vs live virtual casting.

Common casting-call mistakes

• No brief. Running a call without a written brief means every candidate auditions slightly differently, and the shortlist becomes a matter of taste no one can defend to a client. Decide roles, looks, usage, and the exact reads first.
• No framing standard. If each candidate is framed and lit however their phone lands, you are judging cameras, not people. Hold the same crop, focus, and exposure for everyone by controlling the camera yourself.
• Judging from compressed video. A screenshot from a video call is far below the quality you need to read a face or hand a client a serious shortlist. Capture from the real sensor at full resolution so the reference actually holds up.
• Losing files in email threads. Collecting tapes as attachments and forwarding them around guarantees version confusion and lost keepers. Keep every take in one cloud album and share selects by link, so there is a single source of truth.