How a parallel limit lands on a delivery date
A per-second rate prices a second of output and says nothing about how many seconds can be in flight. Batch work is where that matters: the limit decides the wall-clock time of a delivery rather than its cost. As of 2026-09-12.
| Stage of the work | Naturally parallel? | What a low limit does |
|---|---|---|
| Character and scene reference images | Yes, dozens at once | Turns an afternoon into two days |
| Segment generation for one episode | Yes, tens at once | Turns a day into a week |
| Regeneration after review | Yes, but in small bursts | Delays every downstream step by the same amount |
| Joining and delivery | No, one pass per episode | Nothing |
Inclusion rule. Stages of a serial pipeline, and whether a limit on simultaneous generations binds at each one. Stages that are inherently sequential are listed so the contrast is visible. Order. In production order.
1Throughput is a schedule cost, not an invoice cost
A parallel limit does not change what a second costs. It changes how many seconds arrive in an afternoon, which is the constraint a schedule is built from. Two channels at the same rate can differ by a factor of five in how long a week of work takes.
That cost appears nowhere in a price comparison and lands in full on anyone with a delivery date. It is the clearest example of a real cost that a register of rates cannot express in its own units.
2Why so few pages publish a figure
A published concurrency number converts a flexible operational choice into something buyers will hold a seller to. Spare capacity is exactly what a seller wants to keep discretionary, so the usual answer is silence, and the next most usual is a comparative with no base.
A ratio without a base is the worst of the available answers. It works on a tier grid and fails the moment somebody has to decide whether a tier is sufficient, which is the only reason anyone reads the line.
3What to do with an unpublished limit
Measure it before the schedule depends on it. Submit a batch of the size a real episode needs and time the whole batch rather than one job; the useful number is finished segments per hour, not latency per segment.
Where a channel prices extra capacity by the month, that figure is also a signal: it tells a buyer roughly what the seller thinks throughput is worth, and whether the free allocation is generous. Most productions never reach the priced part.
Written to explain, not to record. No sentence here is attributed to a seller, and none of it should be quoted as a fact about one. The sourced material is on the ledger. Related: Bands and grids, Recording a rate.
4What the seconds buy
Price aside, what goes into the seconds behind “Limits and deadlines” is described in SceneMixer's guides:
Every segment starts with a short header that sets the stage before the first shot: Location (a scene chip), Time, Weather, the Opening frame (what the first frame shows), Set dressing (what is in the room, and any background people) and Props (the objects the segment uses).
SceneMixer, The Segment Header, Field by Field (read 2026-10-13)
SceneMixer turns a novel into a watchable short-drama series in one pipeline: upload the novel (Chinese or English) → AI extracts characters, scenes, and an episode-by-episode structure → generates consistent character and scene art → writes storyboard scripts → renders video segment by segment (16:9 by default, 9:16 in one setting) → composites full episodes.
SceneMixer, Turn a Novel into a Video Series with AI (read 2026-10-13)