We take the hard questions first.

ResponseThe studio for remote perception — one suite of tools for practicing intuitives, students of the science, and researchers. You work targets you can't peek at: thousands of curated images, pools you build yourself, written questions drawn blind from an envelope, and a live feed of public markets. Sessions run solo, live-monitored, or hosted for a group — including in a room on paper. A blind AI or human analyst scores them, and every result lands in a record that reports what it can honestly support and withholds what it can't. No claim of proven psychic ability comes with any of it — the record is the answer.

ResponseNo. Nobody on the platform currently has enough blind-judged work for any result to reach statistical significance — and rather than hide that behind a flattering headline number, the dashboard says it on the row itself. Below 12 scored viewings a row shows how many more it needs and no verdict at all, because three on target out of four is a huge-looking margin that means nothing. That is the honest state of things today. If it changes, it will change because the numbers moved, not because the copy did.

ResponseThey are the two numbers on every row of your accuracy table. Margin is how much you beat chance by: 0 is level with guessing, +1 is a good stretch, +2 starts to look like something, +3 and up is hard to put down to luck. Luck odds is how often chance alone would match you, written as odds — "1 in 5" means a guesser does this well one run in five; "1 in 200" is a different conversation. We write odds instead of decimals because a number people misread is worse than no number: it gets quoted. Statisticians would call these the z-score and the p-value. See the table →

ResponseBecause judging is the step where someone decides whether your description actually corresponds to the target — and when you do it for your own session you already know what you wrote. A correspondence can be found after the fact with nobody meaning to bend anything. So self-judged rows are shown, with their hit-rate, and never scored: no Margin, no Luck odds, no leaderboard. Counting them would lift almost everyone's numbers, which is exactly why we don't.

ResponseBecause they are different achievements. Against one decoy, chance is on target half the time; against three, one time in four. Pooling them produces a headline built on an averaged chance rate nobody actually worked against. Each panel size is kept in its own row and measured against the odds you actually faced. For the same reason, when several rows are on screen we tighten the bar — comparing five things gives chance five chances to look impressive — using a Holm correction across exactly the rows displayed.

ResponseNo. Remote viewing is contested. The 1970s laboratory experiments that made it famous drew documented critiques of their methodology, and the field never settled the argument. We claim nothing the evidence hasn't earned — and that's the design. rv.studio doesn't sell a conclusion; it enforces the protocol and keeps an honest record. If a signal is there, your numbers bend. If it isn't, they stay flat. We show them either way.

ResponseNone. rv.studio is method-agnostic: work in CRV, ERV, TRV, SRV, ARV — any of the established remote-viewing schools — or your own intuitive approach. What we enforce is the discipline every method shares: targets assigned blind, impressions locked before the reveal, honest scoring after it. The government-program lineage the field descends from is history worth reading, not a house method we require.

ResponseNo. Beginners start blind on day one — guided protocols, thousands of curated targets, and a weekly challenge to learn alongside. The calibration record starts at session one either way.

ResponseNo. "Viewing" is a historical name, and what viewers report is rarely a finished picture. It can arrive as sound, texture, temperature, motion, emotional tone — or a plain knowing with no image at all. Viewers have different dominant channels, so rv.studio records every modality and orders the channels to how you perceive. Forcing a visual frame would only suppress the channel you're strongest in, so we don't. The perception model →

ResponseNo. Alongside image pools there are written taskings — a question instead of a picture. You write the questions yourself, so the set is no secret, but you draw one at random, so you never know which one you're working. The pool is the envelope. Eight curated collections ship with the studio and you can seed your own envelope from them. At the reveal you say whether it was answered, a clear miss, or close-but-incomplete — the last one reseals it and carries your reading forward to the next pass. Written taskings are deliberately not scored: with no decoys there is no chance baseline, so a number there would be invented. How tasking works →

ResponseYes, both. Host a target event for a group: a sealed target drawn from a pool, one shared between everyone or a different one each, your choice of decoys, and one judging method locked for the whole event so the results stay comparable. Choose in person on paper and everyone works the target in the room; you upload the readings afterwards, including for attendees who have no rv.studio account. The target is sealed when the event is created — the host can't re-roll for an easy one — and locking the event is the upload cutoff, so nothing can be written after the target is seen. Offline and guest results count inside the event but stay off the public boards, because nobody outside the room can audit what happened in it.

ResponseIf you ask it to. Pick your days and the hour in your own time zone, then choose what arrives: a plain nudge, a link that opens a session with your pool already chosen, or a coordinate already drawn and sealed before the reminder is sent, so a target is waiting when you open it. The reminder never names the pool — naming it would narrow the field before you start. We'll say what regular practice does for your numbers when we have data saying it does anything; right now we don't, so we don't.

ResponseNeither the viewer nor the monitor in the room knows the target. That's the classic double-blind — the strongest control against cueing, the unintentional hints that leak from anyone who knows the answer. On rv.studio it isn't an honor system: the server assigns the target, seals it, and records who saw what. Blindness lives in the database.

ResponseBlindness is a ladder of roles, and the server keeps every rung blind to anything that could bias it. Double-blind covers the room: neither viewer nor monitor knows the target. The next rung is the analyst, who scores the work without knowing the real-world question or how it turned out — and scores lock before anything is revealed. The top rung is the question itself: written by a separate human tasker, or drawn blind from the live market feed, with the image–outcome pairing sealed server-side so no one can leak what no one knows. When every rung holds, that's triple-blind. The server labels each session with the tier it actually reached — solo, monitored, double-blind, or triple-blind — and your verified accuracy counts only fully double- and triple-blind work. Solo and self-scored runs are reported separately and can never inflate it.

ResponseImpressions lock before the reveal. A blind judge — AI, a separate human analyst, or both — ranks your transcript against decoys without knowing which target is real or how anything turned out. Every score then lands in a row that keeps its own chance baseline, carries a Margin and Luck odds, withholds any verdict below 12 scored viewings, and is Holm-corrected against the other rows on screen. Self-judged work is shown but never scored. The ugly rows ship too. The evidence model →

ResponseARV — associative remote viewing — binds a session to a future event with a fixed set of possible outcomes. Each outcome is secretly paired with an image; you view blind, and the outcome whose image best matches your work becomes the prediction. On rv.studio it resolves against a live market feed and scores your calibration against the market's own. Treat it as what it is: a calibration-bound experiment with honest base rates, not a financial edge. With two outcomes, guessing alone hits 50% — and the board keeps that chance line on every stat.

ResponseYes. Your sessions and calibration record are yours, and they're encrypted where they're stored. For stronger privacy, generate an end-to-end encryption key (Settings → Encryption) — it is created and held only in your browser. Any session you start as Private is encrypted with that key, and we can't read its notes or sketches. Direct messages between two people who've both set up keys are end-to-end encrypted too. (Full terms in our Privacy notice.)

ResponseEnd-to-end, where only you hold the key: Private viewings — their notes and sketches — and direct messages where both people have keys. Not end-to-end: group chats, chats in Spaces (our shared community areas), and standard viewings. Those stay readable to the service on purpose — that's what lets blind AI scoring, a separate analyst, and live monitoring work on them — and they're still encrypted at rest. The trade-off is yours to make per item: maximum privacy, or the features that need the data readable.

ResponseNothing to start. rv.studio is request-to-join — no payment, nothing to buy — and Free is a complete practice, not a demo: blind tasking, live sessions, honest scoring, a full calibration record. Some AI actions run on credits. Pro — the certified-record tier for putting your results in front of a skeptic — is in early access, and its pricing isn't set yet. See Pro →