Axiom CVD Osc STR

Axiom CVD Osc STR is a cumulative-volume-delta oscillator with room for five independently configured slots, a weighted blend of the active ones, and four optional reading layers built around that blend — pivot-based divergence against chart price, a Keltner-style stretch envelope, a Bollinger Band Width Percentile column, and a Donchian channel. Everything is mapped into a bounded 0-to-100 pane so slots on different timeframes and different symbols can be compared against each other without a rescale argument in your head.

Where this version sits

Axiom indicator families use a Base -> CTX -> STR progression when the full set exists. Base is the free, focused version: three context slots, chart-symbol only, limited filtering, and one global bar-close posture. CTX expands that same core CVD logic into broader context: up to ten slots, per-slot symbol and timing control, and deeper Power User surface for shaping the read. STR is the structure expansion at the end of the series. It keeps CTX-style per-slot control, but it runs five slots instead of ten because the extra processing budget goes into structure on the blended CVD itself.

That trade is the point of this trim. You give up CTX's ten-slot breadth and gain four structure reads around the blend: divergence against chart price, Keltner stretch, BBWP width regime, and Donchian range. STR sits at the end of the sequence; it is the structure version, not a reduced context version. It is the more advanced structure variant for readers who want to ask what the blended CVD is doing relative to its own recent behavior without leaving the pane.

Those are the questions STR is built to help you ask. It does not answer them. Nothing in this pack will describe a divergence triangle, a Keltner touch, a Donchian breakout, or a BBWP flip as a trade. They are readings of a derived line. They belong inside your process. They do not replace it.

The clearest way to think about the pane: STR is an instrument for measuring pressure, with four extra dials wrapped around the main needle so you can see how that needle is behaving relative to its own recent past. The needle is still the read. The dials are framing.

Why this exists

CVD as a category gets traders into trouble in two stubborn ways, and STR is shaped around being honest about both.

The first is that the underlying bid/ask data most people imagine they are looking at is not actually on the feed. Most retail platforms only expose OHLCV per bar — open, high, low, close, volume. Tools that paper over that by calling "close above open" a buy and "close below open" a sell look reasonable on strong directional bars and quietly misread every bar where the body is small, the close is inside the range, and the real participation was somewhere in the wicks. The participation model that runs inside each STR slot is Axiom's own wick-aware estimator. It reads body, close location, and wick asymmetry as separate inputs, carries state forward across indecisive bars rather than flipping on every neutral print, weights each bar by its own volume, and stays honest about what it is. It is an estimate. The script says so in its preamble, and so does this pack — on every page where a reader might otherwise lean on a number without remembering where it came from. If you want true bid/ask delta, no setting inside STR turns the slot into footprint data, and you should pair it with a true-delta source in your process.

The second trap is subtler and harder to escape. Even a well-built CVD line, on its own, leaves you to guess what the read is doing relative to itself. Is this stretch unusually far from where the blend has been sitting, or is it a routine excursion? Is the width compression unusual by the blend's own history, or normal? Did that lower low in price coincide with a higher low in the blend, or was the geometry ordinary? Most tools leave those second-order questions to your head, which is where they usually get answered after the position is already on and your judgement is no longer free.

STR's four structure layers exist so the second-order questions live in the pane instead of in retrospect. The Keltner envelope frames stretch. The Donchian channel frames range. The BBWP column frames width regime. The divergence module frames pivot geometry. None of them are conclusions. All of them are framing. The pack's job is to keep that posture intact and to refuse, on every page, to let any of the four extras drift into predictive language. If you find a sentence in this pack that does drift, that is a polish miss, and we want to know.

What STR adds after CTX

STR's footprint is easiest to read in the family order. Base is the clean three-slot CVD oscillator. CTX widens that same CVD read to ten slots with optional symbol and per-slot timing control. STR comes after CTX: five CTX-style slots because the remaining budget is spent on structure around the blended CVD itself.

• Five CVD slots instead of three. Three come enabled at chart cadence, one step up, and one regime up. The spare two are wired and disabled, sitting ready for a correlated instrument or a higher regime when the question is worth the extra read.

• A weighted blend with optional master smoothing. The blend is a weighted mean of the enabled slots, not of the visible slots — hiding a slot hides only the line. Master smoothing is an optional pass on the blend lines; defaults leave it off so you can see what the blend is doing before you decide to calm it down.

• Divergence triangles. A strict pivot comparison between chart-price pivots and the blended CVD at the same offset. Evaluated only on confirmed bars. The marker placement mode is honest about timing, and the Plot On Pivot setting — which back-shifts the marker to the original pivot bar — is the most carefully framed feature in the kit.

• Keltner envelope around the blend. A stretch ribbon. Upper and lower lines are the blend's smoothed basis plus and minus a multiple of the blend's own smoothed one-bar change. Source is the blend, not price.

• BBWP columns at the bottom of the pane. A percentile rank of the blend's own Bollinger band width across the blend's own recent history. When the column is low, the blend's width has been compressed by its own history. When the column is high, the blend's width has been expanded. This is never a price-volatility read, and the pack will keep reminding you so.

• Donchian channel on the blend. Stepline highs and lows of the blended CVD over the channel length. A touch of the upper stepline says the blend has just printed a new local high by its own recent history. It does not say anything about where price goes next.

All four structure features pull from the same blended line. That is important. When Keltner says "stretched" and the divergence triangle says "geometry formed" on the same bar, that is one line agreeing with itself. It is not two independent reads arriving at the same answer. Treating the four features as if they could confirm each other is the specific kind of mistake this pack is built to prevent.

Who this is for

STR is the right trim for you if most of these are already true. None of them are skill bars; they are read-the-pane bars.

• You treat volume delta as a pressure read, not a directional oracle. When the blend turns down, you do not call that a short. You call it a question that needs the rest of the chart to answer.

• You have already spent time with a CVD oscillator. The idea of stacking three or five of them on a single pane does not intimidate you, but you do not want the full ten-slot CTX surface every time you load the chart.

• You want scaffolding around the blend — a way to see stretch, new highs by the blend's own standards, width regime, and pivot geometry without building four more overlays — and you are willing to spend an evening learning what each of those four is and is not reading.

• You know, from experience, that some divergences confirm turns and some print directly into the continuation they appeared to argue against. You want the pivot logic strict, the confirmation gated, and the marker placement honest about when the script could have known.

• You read more than one chart in a session and want a pane that holds its meaning when you flip from a 5-minute equity to an hourly future without recalibrating your eyes.

Who should probably skip it

• You want a buy-sell signal. STR fires alerts, but they are state descriptors and geometry confirmations. None of them are entries, and any workflow that tries to promote them into entries is explicitly anti-patterned later in the pack.

• You are new to CVD. Start with the base CVD Osc trim. It will teach you the slot idea and the blend idea on a smaller pane with less to misread. Come back to STR when the simpler version feels too narrow for the questions you want to ask.

• You need true bid/ask delta. STR is an OHLCV-based estimate, and no setting inside the indicator turns it into footprint data. If your process depends on tick-resolution participation, this tool is for the framing, not the primary read.

• You want ten slots and cross-symbol density. The CTX trim is the context-breadth surface. Go there.

• You would look at a divergence triangle, glance at the blend color, and act without looking at what the slots underneath were doing. The pack is built to prevent that read, but a manual cannot prevent it if you are not willing to slow down. STR does not pair well with a hurried hand.

What STR is not

• Not a signal service. No arrow ever appears in the pane promising direction.

• Not a true-delta tool. It is a defensible OHLCV estimate. Nothing you can set makes it tick-resolution. The estimate boundary is named on every page where a reader might otherwise lean on a number without asking where it came from.

• Not repaint-proof by default. The On Bar Close? switch on each slot has honest behavior in both positions; the MTF and Repainting page is where the full tradeoff lives. A reader who cares about the historical record matching the live read needs to internalize that page before tuning.

• Not an overbought or oversold pane. The 20 and 80 guide lines are visible threshold marks, not regime rules. The Keltner envelope is a stretch ribbon, not a ceiling. The blend can sit at either pane edge for long stretches under sustained pressure without that meaning anything is exhausted.

• Not a replacement for reading price. The structure features describe a derived line — the blended CVD. They do not describe the instrument. A reader who treats the pane as the chart will misread both.

• Not a pane that gets easier the more you turn on. Crowding has a real cost; STR has settings that exist precisely so you can take features off when they are not the question of the moment.

Where to go next

• If you want the pane drawing correctly on a live chart in one short session, go to Quick Start.

• If you are building a configuration and want to understand what each setting changes and what the downstream cost is, Settings.

• If you want to read the pane well — slot lines, blend, channels, BBWP, divergence — and know which states to trust and which to treat as ambiguous, Visuals and Logic.

• If you are wiring alerts and want to know what each one commits to and what it does not, Alerts.

• If you have been burned by a higher-timeframe indicator that repainted without saying so, read MTF and Repainting before you tune anything.

• For the estimate boundary, the intrabar fallback, and the honest limits of the divergence and BBWP reads, Limitations and Trust Boundaries.

• For named workflows and their anti-patterns, Workflows.

• When a setting is not doing what you expect, a triangle will not show, or the pane looks wrong, Troubleshooting.

• If you want to understand the participation model and the four structure features well enough to defend the reads out loud — without a code walkthrough — For the Geeks.

The indicator is adaptable and customizable by design. Every slot, every weight, every window mode, every structure-feature toggle, every Power User parameter is a control you are expected to change. The rest of the pack is here so you can change them for reasons you can name out loud — to yourself, to a colleague, to the person you are six months from now reviewing why a position moved against you. That is the bar.