Appearance
Jitter: cut, not smoothed over
A take is one recorded motion-capture performance. Jitter is unwanted high-frequency noise on a marker inside a take. A short, sharp spike is a special case of it: a marker mis-tracked for a frame or two. That is not smoothing work. Post cuts those frames out, the peak plus a small pad, and refills the hole with the gapfill model. It is the way you would replace a spike by hand rather than average it into its neighbours. Jitter that is not a spike is still smoothed.
The switch is Cut and refill jitter on the Clean screen, which is the screen where you launch a cleanup. The switch is badged · whole run, meaning it applies to every take in the run, and it is on by default. Its own words:
Impulsive jitter frames are cut out and refilled by the gapfill model before smoothing. Turn it off if the result feels too aggressive, and the take is then cleaned exactly as it was before.
Sensitivity
Jitter sensitivity is the setting that decides how small a jitter has to be before Post leaves it alone. You set it to one of three bands, and each band has an acceleration floor measured in g. The control's own words:
The acceleration floor under which a jitter is left alone. Changing it re-diagnoses the take, which means analysing it again from scratch, because findings from two sensitivities are not comparable.
| Band | Floor | What it does |
|---|---|---|
| Aggressive | 1 g | Flags and repairs the smallest jitter; may touch motion that was fine. |
| Standard | 3 g | Judged in the context of the motion around it. The default, and what a take is diagnosed and cleaned at unless someone chooses otherwise. |
| Impact | 10 g | For scenes with hits, falls and props: only large, isolated spikes are listed. |
Cutting has a higher bar than listing, so a peak can be listed without being cut. The cut floor is 10 g at Aggressive and Standard, and 25 g at Impact. A region card on the Review screen names the floor the peak had to clear, in the words Cut floor 10 g at Standard. A region is one span of frames on one marker that Post worked on, and Review is the screen you read once a run has finished.
Per project and per take
- Per project. Set a default in the project's settings. A project is the folder your takes are uploaded into. There is also a No project default row, which is a real choice and not the same as Standard: with no default, a take that says nothing runs at whatever band the workers apply.
- Per take. On Clean in Assisted, the mode where you choose the options for each take yourself, the control is badged · this take and overrides the project default for that one take.
The one thing it moves
The sensitivity moves the jitter rules and nothing else. Gap fills, swap and merge repair, joins and the fill checks are the same at every band. Within the jitter rules it moves the floor, which is how hard a frame has to be before it counts. It does not change the shape of the test.
Because findings from two bands are not comparable, choosing a band a take was not diagnosed at re-diagnoses it first. The control says so before you launch:
Re-diagnoses this take, which means analysing it again before cleaning: it was diagnosed at Standard, and findings from two sensitivities are not comparable.
You can always see which band produced what you are looking at. The Diagnose screen says Diagnosed at Standard and the Review screen says Cleaned at Standard. A take that does not record the band says an unknown sensitivity rather than guessing.
What Post will not do
There are three rules that do not move with the sensitivity, or with anything else.
It never creates a spike it did not find
A candidate repair that would put an acceleration jump where your raw capture had none is rejected, whatever else recommends it. Your raw capture is the file you uploaded, which Post never modifies. The run reports the count of frames it made hot that the capture had not, and that number is meant to be zero.
A repair never exceeds a body's physical acceleration ceiling
Every frame Post writes is checked against an acceleration ceiling for that marker's class. The classes are core, limb, extremity and finger, because a torso, an elbow and a fingertip cannot accelerate equally hard. A fill that breaks its ceiling is not shipped. The smoothing steps down through gentler settings until one clears, and if nothing clears, the measured frames are handed back and the hole stays open. A region card reading Nothing filled. The rebuilt frames moved the marker faster than a real body part can move, so Kapnetix did not use them. is this rule doing its job.
Smoothing may leave a measured spike that was already over its ceiling over its ceiling. What it may never do is make that frame hotter.
Spikes the diagnosis named far over the limit
Some spikes are so far over their marker's ceiling, three times it or more, that leaving them is not an option. For each one the diagnosis named, Post tries a short bridge across the spike that stays under the ceiling and moves no captured frame more than 10 cm. If a bridge clears, the spike is repaired and the card says so like any other repair.
If no bridge clears without moving captured frames further than that, Post leaves the frames as they were recorded and tells you. The region card reads:
Spike not repaired. The diagnosis found a spike here that moves this marker faster than a real body part can move. Kapnetix could not repair it without moving captured frames further than the cleanup allows, so it left these frames as they were recorded.
Under it the card says The spike is still in the cleaned file. Check these frames by eye before you export. The region counts as work still to do.
Sometimes the named spike sits inside a gap that Post did fill. The fill is in the cleaned file, so that card says the fill first and the spike second, and its header names the region as a gap. It reads, for a 12-frame gap:
Filled 12 of 12 frames in the gap. Inside the gap, the spike the diagnosis named still moves this marker faster than a real body part can move. Kapnetix could not repair it without moving captured frames further than the cleanup allows, so it left the spike as it was recorded. Revert removes the fill and leaves the gap open again. The spike stays as it was recorded either way.
Under it the card says The fill is in the cleaned file, and the spike is still inside it. Check the spike by eye before you export. This card has a Revert button, because there is a fill to undo. A card for a spike that Post left alone and did not fill around has no Revert button, because nothing on it changed. Both count as work still to do, and the error list names each one the way its card does.
The Review screen also says how many of these spikes the take had, under the take name, for example The diagnosis named 3 spikes far over the physical limit for their markers. Kapnetix repaired 1 and left 2 as recorded. When there was one it reads The diagnosis named 1 spike far over the physical limit for its marker. Kapnetix left it as recorded.
Real fast motion is kept
A strike, a landing and a fall are genuine motion, and the take has them in it. Before a peak is cut, Post checks whether the marker was actually travelling at that moment. A marker that is moving fast is not jitter, and its frames are left alone. The card states the reading: Flank speed 1.24× its median, so the marker was not travelling.
There is one exception, and it is reported rather than hidden. A marker that leaves its trajectory and lands straight back on its own predicted path is cut even when the speed gate declined it. The card says both numbers and the decision, in one sentence: The speed gate declined this peak (1.87× its usual step); it landed 0.4 steps off its own path, so it was cut.
Reading a cut on Review
The region card leads with what happened:
Cut and refilled 6 frames · peak 61 g at frame 1204Cut reverted: the original frames were kept. Post measured the cut and then put the original frames back. Nothing shipped, so there is nothing to revert.
Under that line you get the evidence: how many measured frames the cut overwrote, how many times the marker's own acceleration level the peak was, the flank-speed reading, the cut floor for the run's band, and whether the cut came from Pass 1, before the gap fill or Pass 2, over the filled result.
What to do next
If a take comes back over-cleaned, turn Cut and refill jitter off and run it again. That take is then cleaned exactly as it was before the option existed. If a scene is full of real impacts, move it to Impact instead, and expect the take to be re-diagnosed first. If one specific region is wrong, revert it.
Ask us
Something looks wrong on a take? Write to support@kapnetix.ai and send the take id, or the take's name and the project it sits in.