How to send a private screener to a client.
One link you wrote yourself, one code you can kill.

Public is the fifth, and it is the only one that has nothing to do with a screener. The choice that matters is Password against Unlisted, and it is a question about the recipient list rather than about the footage. A password is worth the extra sentence in the email when the film will be forwarded inside an organisation you do not control. It is friction with no payoff when the film is going to one producer who will open it on a phone in a taxi.

One useful wrinkle: if the access code you set contains an @ character, the gate switches to an email-allowlist mode instead of a shared code. Same setting, different shape of question at the door.

The address a client receives is part of the pitch. A link of the form uncompressed.io/yourstudio/the-film reads like a studio; a string of random characters reads like a file drop. Both work. Only one of them survives being pasted into a broadcaster’s internal ticket without somebody asking what it is.

There is a practical reason too. The owner-only analytics live on the clean handle route, so sending that link is what gets you the data described further down. Slugs are unique per account and editable at any time, so a film can be renamed between rounds without a new page.

When cut two is ready, upload a new version onto the same film instead of creating a second one. The stored file, its size, duration, frame rate and dimensions are replaced on the same record, and the old file is deleted. The link, the slug, the visibility, the password, the embed options and the view count all survive. The client clicks the link they already have and sees the new cut, which removes the most reliable source of confusion in a review round: two links in one thread.

The player itself carries no platform logo on any plan, free included. The only logo it can show is your own, and it is off by default. And because nothing is transcoded on the way in, the screener a client watches is the exact export you graded, not a re-encode of it.

The YouTube row worth reading twice is the sign-in one. Its unlisted mode is a link anyone can forward, and its private mode requires each viewer to be signed in to the exact account the film was shared with. For a festival programmer working through a shared submissions inbox, or a broadcaster whose staff live in a corporate identity system, that is not a small ask. It is the reason screeners sent to YouTube tend to end up unlisted, which is to say not gated at all. YouTube has also changed the meaning of unlisted retroactively before: older unlisted videos were switched to private in July 2021 unless the owner opted out.

Anyone holding the link and the code can forward both, in one message, to anyone. No browser prevents a screen recording. No setting on any platform, ours included, stops a phone pointed at a monitor. The gate is a real control and it is enforced in the right place, on the server, before the page renders, with a rate limit and a code that never reaches the browser. It is still a control over who can open the door, not over what happens once someone is inside the room.

What actually raises the bar is attribution, and that is what Vault is for. Vault masters play only in the native app rather than a browser tab, screen capture is blocked, every session carries a watermark, and a per-person forensic mark is shipped, so a copy that escapes points at the account it was played from. Files are encrypted with AES-256 at rest and playback runs on short-lived links. That is the toolset for footage under embargo. For a colour round on a corporate film, a password and a habit of rotating it is the proportionate answer, and pretending otherwise wastes your client’s afternoon.

A screener that produced no notes produced one of two situations, and they call for opposite emails. The owner-only analytics on the clean handle link separate them: opens, plays and the rate between them, unique visitors, views per person, a per-day series, average view time, total view time, average percent watched, finishes, a retention curve, a watch-time histogram, top countries and top referrer sources, over 7, 30 or 90 days or all time.

The retention curve is the one that changes a conversation. A note that arrives with a curve showing the viewer stopped at 00:40 is not a note about the ending. Your own visits never bump the view counter, so the numbers describe the client rather than your rewatching. Download and bandwidth reporting are not shipped; they show as coming soon in the dashboard, and this page will not claim them.

What is left, once those three are set aside, is the common case: a working studio sending a cut to a client who should not have to make an account, on a link that looks like the studio, behind a code that can be retired the day the job closes, playing the file that was actually exported. That case is free here, and it is the reason this page exists.