Stop Editing

Every tool in your editor is built to keep you editing. The sliders sit where you left them, each one an open invitation; the histogram will tolerate one more nudge forever; and no panel anywhere in the interface lights up to say "done." Doneness is not a feature of the software. It is a decision (yours) and stop-editing is the name of that decision: the moment you take your hands off the file, export it, and decline to touch it again. Not because you ran out of ideas, but because you recognized the point where the next move stops adding and starts subtracting, and stopped there, on purpose.

That makes this the rare lesson about an act rather than a technique. There is no slider recipe for ending a session; there is only the discipline of ending it, and that discipline is real craft, later-arriving and rarer than any of the processing skills it protects.

Two neighboring concepts hold the parts around this one, and keeping the three separate keeps them all sharp. Diminishing returns is the CURVE: the payoff of each successive move falling away, and how to read that slope (/learn/diminishing-returns). The editing ceiling is the CAP: the highest this particular file can ever reach, fixed the instant the shutter closed (/learn/editing-ceiling). Stop-editing is neither the reading nor the wall: it is the ACT (the hand coming off the mouse at the top of the curve, at the base of the cap). FinalFrame surfaces the moment as a flag on your critique (stop_editing, a single boolean that says, in effect, put it down) but the flag is the product performing a judgment this lesson exists to train into you, so that eventually you make the call before the critique does.

What earns stopping its own page is that the failure to stop is invisible from the inside. No photographer ever feels themselves over-working a file. They feel themselves improving it: every one of the moves that ruins the picture is made in good faith, with full attention, right up to the last one. Over-editing is the only common editing failure with no felt symptom at the desk; it is diagnosed afterward, by fresh eyes, on a frame that used to be better. That is how a capture that peaked at a strong B+ (7.5 and up, the band where real skill shows) gets ground back down into the competent B middle by its own maker. The talent was in the file. The frame was in the file. The ending was not.

FinalFrame's stop_editing flag is the product making that call from outside, and it is built from a real composite, four conditions the critique checks together: your Artistic sub-score has cleared 8.8, up near the top of the A- band rather than merely into it; no structural flaw remains; the frame reads consistently for its chosen lane; and the last edits have stopped paying: the diminishing-returns condition, which its own lesson unpacks. When all four hold, the flag fires: you are genuinely at the top, and the product is saying so before you can edit past it.

But the flag has a second, quieter fire that means the opposite thing. It also raises when the file is capped low by the capture itself: when no edit can lift it and the honest instruction is a reshoot, not another pass. Same boolean, two opposite messages, and reading which one you were handed is half the skill of this page. What the flag never is, is a grade. It measures neither the photograph's quality nor your effort; it marks the moment the work should end, and the scores beside it tell you whether that ending is a summit or a wall.

You are reading from two directions at once (the file in front of you, and the flag on your critique) and the skill is catching the stop moment from either one.

The file gives you three signals, in rising order of visibility. The first belongs to diminishing returns and is only borrowed here: the payoff curve going flat, version-to-version jumps shrinking until a whole session moves the score by less than a third of a point. Its lesson teaches the reading; take the tell and keep moving. The second signal is native to this page: the session stops progressing and starts CHASING (each correction creating the problem the next correction exists to solve). You warm the frame, and now it is too warm; you pull saturation, and now it is flat; you add contrast, and now the shadows block up. Nothing in that loop is progress. It is one problem being pushed around the frame by a photographer who should have exported three moves ago. The third signal is the one a stranger sees first: the image begins announcing its edits. The halo along the ridgeline, the over-crunched texture, skin gone waxy: when the processing is the first thing a fresh eye meets, the edit has replaced the picture as the subject. A finished photograph shows the viewer the scene. An over-worked one shows them the work.

Then read the flag: as a route, never as applause or a sentence, and read WHICH route. Fired with the scores high and no structural flaw named: you are at the ceiling and the frame is strong; stop, and be glad. Fired beside a capture note, a missed plane, a dataless highlight, a moment that broke half a second early: the file was capped before you opened it, and the next move is a reshoot, not a slider (/learn/capture-limited). And if you are iterating and the stop signal rises on a version you believed was an improvement, with a warning naming the version to return to: that is the product catching an over-correction you could not feel from inside the session. It is not saying the picture failed. It is saying where the peak actually lived: one version back.

The move is three small physical acts and one routing decision, and everything difficult about it lives in the doing: there is no technique here to hide behind.

Bank the peak before you risk it. The moment the signals say you have arrived, save the state: a version, a snapshot, a virtual copy. Editing degrades a picture one confident slider at a time, and a peak that was never banked is a peak you will spend without noticing. The saved version is what makes stopping survivable: once the best state exists somewhere safe, nothing you do afterward can cost you anything you cannot get back. Diminishing returns calls this same habit banking the peak from the curve's side; this is it from the stopping side.

Then end it, physically. Export the file and close it, because a file left open is a file that gets fiddled with, and "one more look" is the exact mechanism by which peaks get spent. And leave the screen, not as ritual but as method: your eye calibrates to whatever it has stared at longest, which is precisely why the over-cook is invisible while you are inside it. Distance breaks the adaptation; cold eyes see the crunch that warm ones applied. The stop is not complete until you are out of the room the file is in.

Last, route the effort you just declined to spend, because stopping is a fork, not a dead end. Ask why the file stopped. Genuinely strong and simply finished, the flag fired high, then the saved hour belongs to the next capture; that is the worth-it calculus, and its own lesson holds it (/learn/worth-it). Stalled low because the capture set the wall, then you are at the editing ceiling, and the effort belongs in a reshoot (/learn/editing-ceiling). What belongs to THIS page is only the ending itself. The diminishing-returns lesson can read the slope for you; no lesson can lift your hand off the mouse. That part was always yours.

Four ways the concept gets misread, and each one either holds a photographer at the desk too long or sends them away too early.

First: reading stop_editing as a compliment or an insult. It is neither, structurally: the flag that fires on a finished A-tier frame and the flag that fires on a capture-capped one you should reshoot are the same boolean carrying opposite news. The praise or the bad news is written in the scores and the capture note beside the flag, never in the flag itself. A photographer who hears "I did well" or "I failed" has read the one thing the field does not say.

Second: hearing "stop" as "abandon." Ending an edit is not giving up on a frame; it is refusing to make it worse. There is a real ego cost to closing a file with sliders still untouched: a suspicion that a serious photographer would keep working. The truth runs the other way: pushing a file is an early skill, and knowing the moment a picture is done is the later, rarer one. The stop is a mark of command, not the absence of it.

Third: the mirror failure: blowing past the flag because stopping feels like weakness. This is the photographer who watches the plateau arrive, feels the corrections start to fight each other, sees the halos rise, and keeps going anyway, because effort feels like virtue and a closed file feels like quitting. Every move past the peak is the peak being spent. The correction is to rank the signals above the feeling: the file is not asking for more of you. It is asking you to stop.

Fourth: stopping without asking why the file stopped moving. A flat score at a strong number is a finished edit; a flat score at a weak one is a wall the capture built, and it wants a reshoot, not a shrug. Treat the two as one event ("the number quit, so I quit") and you will walk away from salvageable frames and sit content with capped ones. The stop is identical; the reason underneath it is what routes your next hour.