Upload the self-tape sitting on your desktop right now and get coached on it. Timestamped notes on eye contact, the moment before, pacing and breath, and whether the choice is landing. An honest verdict instead of an endless list. One direction you can execute on the next take.
Most online acting coaching is a scheduled hour on a video call. That hour is valuable, and it is also almost never available at the moment you need it. Sides drop Thursday evening, the tape is due Friday morning, and the coach whose notes you trust is booked into next week. So you shoot four takes, watch them until they all sound identical, pick one at random, and send it.
Take One is built for that gap. You upload the take you just shot, or record straight into the browser, and the coaching comes back on that specific performance. It runs at 2am on a Sunday because there is no calendar to negotiate with. The subscription is one flat price, so a fourth pass at the same scene costs you nothing extra, which changes how willing you are to actually try the note.
The difference between useful coaching and general encouragement is location. "Be more grounded" is not something you can act on. "At [0:31] your eyes drift down and left before each line, so memorize the first four lines cold and the top of the tape has zero downward glances" is something you can do in the next sixty seconds.
Every note comes anchored to a moment in your video. You scrub to that second and watch the thing being described. That is also what makes the feedback checkable: you can disagree with a note, but you can see exactly what prompted it, which is more than you get from a coach's memory of an hour that ended twenty minutes ago.
The failure mode of coaching is the endless loop. You apply the direction, send the next take, and get a fresh wall of brand new notes. Nobody ever tells you the tape is ready, so you never feel progress and eventually you stop taping.
Every analysis opens with one of three verdicts, chosen honestly:
It clears the bar. The notes stop, anything left is marked as optional taste rather than a fix, and you send the tape.
Close, and one specific thing is costing you. You get that one direction properly instead of a full audit.
Something fundamental is not landing. Now you get the detailed pass, because it is actually warranted.
The verdict is never inflated to be encouraging and never withheld to seem rigorous. Every take earns its own tier cold, so a revision can drop from Submit This back down if the new take got worse. That matters more than it sounds: an approval you cannot trust is worth nothing.
Most self-tape problems are rehearsal problems wearing a costume. The downward eye drift is not a camera issue, it is not being off book. Scene Partner is included in the same subscription: paste your sides, tag which role is yours, and it reads every other character out loud while listening for your lines, picking the scene back up the moment you deliver. It forgives paraphrasing, so it feels like running lines with a person rather than reciting at a machine.
Three drills get you off book. Cue mode hides your lines and feeds you the first words only if you stall. Fade mode blanks out more of your words on each pass. Speed-through runs the scene fast with no pauses. Each pass ends with an off-book score and the exact lines you dropped or paraphrased, at any hour, with no reader to schedule.
A human coach in the room with you catches physical habits that no recording reveals, knows which rooms you are right for, and can open doors. None of that is what Take One does. What it does is the part that has to happen between those sessions and at the hours when nobody is available: telling you what your take actually looked like, where it went wrong, and whether it is ready to send. If you want the longer argument on where each one is worth your money, we wrote it up in AI coach versus human coach.
Sample Output
A drama pilot self-tape, coached in full. Every note is anchored to a moment in the tape.
The stillness you hold at the top is real stillness, not held breath, and that is a genuinely hard thing to do on camera. Reading this as grounded single camera drama, so I am judging truth over size. My honest read: the performance is there, and one habit is standing between it and a clean submission. What decided it was the first ten seconds, where you are present with the lens and then start hunting for the next line.
You are present with the lens for the first third, then at [0:31] your eyes start drifting down and left before each line. You are reading. It reads on camera as checking out of the moment. The strongest stretch is [0:48] to [1:02], where you stop hunting for the words and just look at your reader; the difference in aliveness is obvious. Keep whatever you were doing there, that is the tape.
At [1:12], on "I already know what you are going to say", you land something real and underplayed, and it is the most castable second in the tape. The apology at [0:22] is the opposite: you are showing me regret rather than having it, mostly in the eyebrows.
Not cold yet. The downward glance before each line in the middle third is the tell, and it is the single thing costing you here. It is also the easiest to fix.
Memorize the first four lines so cold you could say them walking down stairs, then shoot again and do not let your eyes leave your reader before [0:45]. Everything else in this tape already works.
On playback, count the downward glances in the first forty seconds. If it is zero, send it.
VO reads get the same treatment on the audio side: operative words, breath placement, and where the energy sags, all timestamped.
7-day trial, then $20 per month billed automatically unless you cancel. Everything included.
Start 7-Day Trial →FAQ
A video call gives you a scheduled hour and notes you have to remember afterwards. Take One works on the take itself. You upload the self-tape you just shot and the coaching comes back written down, anchored to timestamps in your video, so you can scrub to the exact second a note refers to and watch what the coach watched. It is there the night before a deadline, and it costs the same whether you run one take or twenty.
It watches the video. The notes cover eye contact and where your eyeline goes on the hardest line, whether you walked in already living or the scene started when the slate ended, pacing and breath, script ownership, how the performance is sized for a close-up, and basic tape craft like framing and whether your eyes are lit. Script analysis tells you about the character. This tells you about your take.
No. Every analysis opens with one of three honest verdicts: SUBMIT THIS, ONE MORE TAKE, or NEEDS REWORK. If the tape clears the bar it says so and stops rather than hunting for improvements to look thorough. If one specific thing is costing you, you get that one direction properly instead of a wall of notes. The point is to let you finish and send the tape.
Yes. Scene Partner is included in the same subscription. Paste your sides, tag which role is yours, and it reads every other character out loud while listening for your lines, picking the scene back up the moment you deliver. It forgives paraphrasing, and three drills help you get off book: Cue mode, Fade mode, and Speed-through. Each pass ends with an off-book score and the exact lines you dropped.
Working on the audio side instead? See voice over coaching. Want the detail on tape notes specifically? See self tape feedback. Related reading: audition prep, the self-tape checklist, and why you are not getting callbacks.
Pricing
7-day trial, then $20 per month billed automatically unless you cancel. No API key required. Cancel anytime.
Take One: Full Access
Cancelling takes about ten seconds and you do not need to email anyone. Manage or cancel from the homepage, enter the email you subscribed with, and Stripe ends your billing right there. You keep access through the end of the period you have already paid for. If you are stuck, write to takeonesupport@gmail.com.