Settings

This page is the knob-by-knob reference for Axiom CVD Osc CTX. It is written for a reader who has the Inputs dialog open on one screen and this page on the other, and who wants to understand what a control actually changes — not just what it is named — before flipping it.

Three ground rules before you touch anything:

• Each of the ten slots owns its own copy of every control below. A setting under CVD 03 changes CVD 03 only. There is no global "set all slots to X" switch, because the point of the tool is exactly that slots can carry different postures on purpose. A 5-minute session slot alongside a 60-minute rolling slot is not a misconfiguration; it is a reader asking two different questions at once and wiring the pane to answer both.

• A default is a starting position, not a recommendation. The first three slots ship on a 5/15/60 stack with equal weights because that combination is a legible baseline — the three timeframes are far enough apart to genuinely disagree, and equal weights make the blend easy to read against the slots. It is not the "best" configuration for your instrument, your timeframe, or your question. It is the configuration with the smallest number of opinions baked into it.

• Power User fields are dormant until the matching MA type is selected. Tuning ALMA Sigma while your MA type is SMA changes nothing. The dialog does not dim these fields; it is your job to know which ones are live. If you change a Power User field and the line does not budge, confirm the MA type before assuming the knob is broken.

Reading this page

Each control appears once, in the slot-level section, with its name as it appears in the dialog, its default, its range where one exists, and a description of what it changes downstream. If a control carries a misuse warning it is called out under Watch for. If a control has a non-obvious tradeoff with another control it is noted under Interacts with.

After the slot-level block, three smaller sections cover the General Oscillator, Display, and Master Smoothing groups. The Power User group is documented once, after the primary slot fields, because every slot has an identical copy of it.

The slot block — one through ten

Each slot (01 through 10) has the fifteen controls below. Behavior is identical across slots; only defaults differ. Slots 01, 02, and 03 ship enabled. Slots 04 through 10 ship disabled with Blended Weight at zero.

Enable CVD NN

• Type: toggle.

• Default: true for slots 01–03, false for slots 04–10.

• What it changes: whether the slot runs at all. A disabled slot draws no line, contributes nothing to the blend, and does not vote in the All CVD Slots Bullish and All CVD Slots Bearish alignment alerts. Its per-slot alerts also do not fire.

Watch for: Enable CVD NN = false is the only way to fully exclude a slot. Hiding the plot or setting the weight to zero does not deactivate a slot; both are softer moves documented below.

Hide CVD NN Plot

• Type: toggle.

• Default: false.

• What it changes: whether the slot's line is drawn on the pane. Nothing else. The slot's CVD is still computed. Its weight still contributes to the blend. Its per-slot alerts still fire. Its bullish/bearish state still counts in the alignment alerts.

Watch for: This is the single most common first-week misread. Hiding is a visibility choice. To exclude the slot from the blend, set Blended Weight to 0. To exclude it from alignment alerts and per-slot alerts, set Enable CVD NN to false.

TimeFrame:

• Type: timeframe.

• Default: 5 for slot 01, 15 for slot 02, 60 for slot 03, blank for slots 04–10.

• What it changes: the timeframe the slot reads its OHLCV from before building its CVD estimate. Blank means "inherit the chart's timeframe."

Watch for: A slot timeframe lower than the chart timeframe is a runtime error, not a soft warning. The pane will go blank and the status row will name the offending slot. Folding a 1-minute CVD read into a 5-minute chart bar has no honest representation at chart resolution; the guardrail refuses rather than quietly averaging.

Interacts with: Lower TF Precision: must be strictly less than TimeFrame:. See the precision field below.

Lower TF Precision:

• Type: timeframe.

• Default: 1 for slots 01, 02, and 04–10; 5 for slot 03.

• What it changes: the lower timeframe the estimator walks through to decompose each slot-TF bar into signed-volume contributions. Finer precision means the participation model sees more internal structure. Coarser precision is faster and has more history available on TradingView.

Watch for: Lower TF Precision set equal to or greater than TimeFrame: is a runtime error. A 1-minute CVD slot cannot use 5-minute precision — there is nothing to decompose. When a slot looks back past the intrabar history TradingView allows for your plan, the estimator falls back silently to a cruder single-bar OHLCV delta. The line does not visually announce the fallback. The MTF & Repainting and Limitations & Trust Boundaries pages describe when this bites.

Window Mode:

• Type: enum — Session or Rolling.

• Default: Session.

• What it changes: the interval over which the slot's CVD accumulates and the interval against which its normalized value is rescaled into the 0-to-100 pane.

• Session — CVD resets to zero at the start of each Window: interval. A dashed vertical line drops in the slot's color at each reset edge. Normalization rescales against the high and low observed inside the current session.

• Rolling — CVD evolves continuously using a sliding lookback equal in length to Window:. Older bars drop out of the window as newer ones enter. There are no reset edges and no dashed markers in this mode.

Interacts with: Window: — described next. Window mode is a per-slot choice. A reader can run slot 01 in Session mode for intraday pressure and slot 03 in Rolling mode for regime context, and the pane will honor both.

Window:

• Type: timeframe.

• Default: D (daily).

• What it changes:

• In Session mode, the anchor timeframe that defines when the CVD accumulator resets. Common choices: D for an intraday session reset, W for a weekly one, 60 for an hourly accounting cadence.

• In Rolling mode, the size of the sliding lookback. A D window in Rolling mode looks back over roughly one day of bars at the slot's timeframe.

Watch for: Window: must be greater than or equal to TimeFrame:. A window finer than the slot TF has no interpretation and the indicator will throw a runtime error. Window: set too short on a slow slot (for example, 60 on a daily slot) compresses the normalization range so hard that every reading clusters near 0, 50, or 100 and the pane loses its ability to distinguish calm from extreme. Extend the window when it does.

CVD Length:

• Type: integer.

• Default: 3.

• Minimum: 1.

• What it changes: the smoothing length applied to the CVD line itself. Lower values track the raw participation estimate closely. Higher values present a calmer line at the cost of reaction latency.

Interacts with: CVD Type: — different MA families respond differently at the same length. An EMA at length 3 and a Jurik at length 3 are not interchangeable.

CVD Type:

• Type: MA type enum — SMA, EMA, RMA, WMA, VWMA, HMA, ALMA, SWMA, DEMA, TEMA, TRIMA, LSMA, KAMA, JMA, FRAMA, T3MA, VAMA, ZLMA, ZLEMA, LAGUERRE, MCGINLEY.

• Default: SMA.

• What it changes: the moving-average family used to smooth the slot's CVD. Each family has a different latency-versus-noise posture. Power User fields for ALMA, KAMA/FRAMA, JMA/Jurik-style, LAGUERRE, and VAMA become live when those families are selected. The Axiom Moving Average Library manual explains each family's behavior and the fields it consumes; this page documents which fields exist on this indicator, not how every MA family behaves.

Watch for: Mixing exotic MA types across slots (KAMA on slot 01, Jurik on slot 02, FRAMA on slot 03) will make the pane harder to read, because each slot now has its own character of responsiveness. Unless you know exactly why you are mixing, keep the MA type consistent across slots for a given workflow.

Signal Length:

• Type: integer.

• Default: 3.

• Minimum: 1.

• What it changes: the smoothing length applied to the signal line the slot CVD is compared against for color and for the per-slot bullish/bearish state. A longer signal length means the slot's crossings become less frequent and the color flips less often.

Signal Type:

• Type: MA type enum — same set as CVD Type:.

• Default: SMA.

• What it changes: the moving-average family of the signal line. Power User fields are shared with CVD Type: — there is one ALMA Offset, one Jurik/JMA-style Power, and so on, per slot, used wherever the chosen family needs it.

Line Width:

• Type: integer.

• Default: 2.

• Minimum: 1.

• What it changes: the thickness of the slot's plotted line on the pane. Purely cosmetic; no effect on computation, alerts, or blend.

Blended Weight:

• Type: float.

• Default: 33.3 for slots 01–03, 0.0 for slots 04–10.

• Minimum: 0.

• What it changes: the slot's contribution to the blended CVD and blended signal. The blend is a weight-weighted mean of all enabled slots whose weight is non-zero. Weights do not need to sum to 100 — only their relative proportions matter.

Watch for:

• Setting a weight to 0 keeps the slot visible but removes it from the blend. This is the "watch without steering" pattern used in cross-ticker studies — see Workflows.

• A slot with weight set very high (say 100) swamps the others even if they are enabled. If the blend looks like it is ignoring one or two slots, check the weights.

• Weights do not auto-rebalance when you enable or disable a slot. Enabling slot 04 at weight 33.3 alongside three active slots at 33.3 each gives you four voices at roughly equal weight; the blend line will shift. If that surprises you, rebalance deliberately.

Pressure Sensitivity:

• Type: float.

• Default: 1.50.

• Range: 0.25 to 4.0, step 0.05.

• What it changes: how quickly the participation model commits to a decisive read and how strongly it carries the prior state through bars that remain ambiguous. Lower values require clearer evidence and carry the prior posture with less force. Higher values accept smaller asymmetries as stronger evidence and carry the prior posture more strongly through still-ambiguous bars.

Watch for: The two ends of the range have very different characters. Pushed near 4.0 on a fast timeframe, the classifier is more permissive and the carried posture can stay loud through ambiguous bars until clear opposite evidence arrives. Pushed near 0.25, the classifier becomes cautious enough that the line can appear muted for long stretches through genuinely active tape, because the knob is refusing to commit to anything short of strong evidence and is carrying the prior posture weakly. The honest interval for most instruments is closer to the middle than to either end. Move in small steps, watch what changes, and when in doubt keep different slots on the same value so you are not comparing apples to oranges across the stack. For the Geeks explains what the knob is actually pressing against at mental-model depth.

Wick Weight:

• Type: float.

• Default: 0.20.

• Range: 0.0 to 0.50, step 0.05.

• What it changes: how much wick asymmetry (upper-wick-heavy vs lower-wick-heavy) contributes to the slot's per-bar intent score. At 0.0, the estimator ignores wicks entirely and reads body-and-close-location only. At 0.50, wick asymmetry has a materially larger influence, but it is still one part of the bar-shape read rather than a standalone signal.

Watch for: On wick-heavy instruments (highly reactive symbols, certain crypto pairs, news-driven bars), raising Wick Weight makes the slot more responsive to rejection behavior — a long upper wick on a near-flat body is read with more bearish intent than it would be at default. On wick-quiet instruments, the knob barely moves the line; if you are not seeing a difference, check whether the bars you are looking at have meaningful wicks.

Optional Ticker:

• Type: symbol string.

• Default: blank (the chart's current symbol).

• What it changes: the symbol the slot reads OHLCV from. Blank inherits the chart symbol. Set to another symbol — QQQ, BTCUSDT, an exchange-qualified ticker — and the slot computes its CVD from that symbol's bars, rescales into the same 0-to-100 pane as everything else, and plots in the slot's assigned color.

Watch for:

• Per-slot normalization rescales the foreign symbol's CVD into its own window's observed high and low. The position of the optional-ticker line relative to your chart's slots is not a raw offset; it is "where this ticker is inside its own window." Do not read the two lines against each other as if they share a zero.

• Session-mode windows anchored to the same daily timeframe can land at different wall-clock edges for symbols traded on different venues (SPY and BTCUSDT do not share a session reset). The dashed reset markers will also not align across slots in that case. Rolling mode avoids this mismatch.

• Pair the optional ticker setup with Blended Weight = 0 when you only want the slot as a study alongside the blend, not as a voice steering it. See Workflows for the documented cross-ticker pattern.

The Power User group — one block per slot

Every slot has a matching Power User group (CVD 01 PU through CVD 10 PU) that carries the repaint switch and the family-specific MA tuning fields.

On Bar Close?

• Type: toggle.

• Default: true.

• What it changes: the repaint posture of the slot.

• true — the slot reports the value of its higher-timeframe bar one step back (the last confirmed value). The line steps and holds. It does not move during the formation of the current slot-TF bar.

• false — the slot reports the live value of its current higher-timeframe bar. The line drifts while the bar forms. It may change direction several times before the bar closes. Once the bar closes, the value is final.

Watch for: ON is the default because it is the honest posture when nothing else overrides that default. OFF is a real option with a real use — the scalping desk that wants the earliest visible read and accepts the drift as the price of that read — but a value taken from an OFF slot before its higher-timeframe bar closes is provisional by construction. The difference is not rhetorical. A slot you flip OFF and then read as if it were ON is the exact failure mode the ON default was put in place to protect you from. If you do flip a slot OFF, keep a note of which slots are OFF and why; future-you will not remember on its own and will read them as if they were still ON. The MTF & Repainting page documents when OFF is a legitimate choice and gives you verification moves to confirm the slot is doing what you expect.

The MA-family fields

The remaining fields in the Power User group are tuning knobs for specific MA families. They are only live when the slot's CVD Type: or Signal Type: selects the matching family. The Axiom Moving Average Library manual covers what each field does to the MA it belongs to — behavior is identical here.

• CVD ALMA Floor Offset? (bool, default false), CVD ALMA Offset (float, default 0.85), CVD ALMA Sigma (float, default 6.0) — consumed when the slot's MA type is ALMA.

• CVD KAMA/FRAMA Fast (int, default 2), CVD KAMA/FRAMA Slow (int, default 30) — consumed by KAMA and FRAMA.

• CVD Jurik Phase (int, default 0), CVD Jurik Power (float, default 2.0) — consumed by JMA/Jurik-style smoothing.

• CVD Laguerre Alpha (float, default 0.5) — consumed by Laguerre.

• CVD VAMA Vol Length (int, default 20) — consumed by VAMA.

Watch for: A slot whose MA type is SMA and whose ALMA Offset is set to 0.10 behaves exactly as a default SMA. The pane is not silently using the field; the field is dormant. Confirm your MA selection before concluding that a Power User field is or is not doing what you want.

General Oscillator group

Overbought Level

• Type: float.

• Default: 80.0.

• Step: 0.1.

• What it changes: the upper horizontal guide on the pane. Cosmetic only — the level does not feed the state logic, the color logic, or any alert. Move it if your eye wants a different reference line.

Oversold Level

• Type: float.

• Default: 20.0.

• Step: 0.1.

• What it changes: the lower horizontal guide on the pane. Same rules as above — cosmetic only.

Watch for: Because these levels do not feed an alert, there is no "crossing oversold" alert inside this indicator. A reader coming from a classical RSI tool sometimes expects one. The Alerts page inventories what is and is not wired up.

Display group

Plot Blended CVD/Signal

• Type: toggle.

• Default: true.

• What it changes: whether the blended CVD, the blended signal, and the translucent fill between them are drawn on the pane.

Watch for: This setting hides the plot, not the math. The blended CVD is still computed. The blended bullish/bearish alerts still fire. If you want a slot-only pane with no blend visual, this is the right control; if you want to prevent the blend from being evaluated at all, set every slot's weight to 0.

Blended Line Width:

• Type: integer.

• Default: 3.

• Minimum: 1.

• What it changes: the thickness of the blended CVD and blended signal lines. Cosmetic; no effect on computation.

Master Smoothing group

The master smoothing block applies a second smoothing pass to the blended CVD and blended signal after they have been computed from the slot weights. It is off by default. Per-slot smoothing is not affected.

Enable Master Smoothing

• Type: toggle.

• Default: false.

• What it changes: whether the master pass runs at all. When off, the remaining master fields have no effect.

Master MA Type

• Type: MA type enum — same set as the slots.

• Default: EMA.

• What it changes: the MA family used for the master pass. The matching Master Power User fields below become live when the family is selected.

Master Length

• Type: integer.

• Default: 3.

• Minimum: 1.

• What it changes: the smoothing length of the master pass. Higher values make the blend line calmer and slower to respond. The same latency-versus-noise tradeoff as the per-slot length applies, but now it compounds with any per-slot smoothing already baked in.

The Master Power User fields

These are family-specific knobs for the master pass. Identical in role to the per-slot Power User fields, but they affect only the blended lines.

• Master ALMA Floor Offset? (default false), Master ALMA Offset (default 0.85), Master ALMA Sigma (default 6.0) — consumed by ALMA.

• Master KAMA/FRAMA Fast (default 2), Master KAMA/FRAMA Slow (default 30) — consumed by KAMA and FRAMA.

• Master Jurik Phase (default 0), Master Jurik Power (default 2.0) — consumed by JMA/Jurik-style smoothing.

• Master Laguerre Alpha (default 0.5) — consumed by Laguerre.

• Master VAMA Vol Length (default 20) — consumed by VAMA.

Watch for: Master smoothing adds lag in exchange for a quieter blend. The tradeoff is real: a blend that is visibly calmer than the slots beneath it is also reacting later than the slots beneath it. Turn it on when your question is "is the stack agreeing over five to ten bars?" Turn it off when your question is "where did the stack just flip?"

A settings cheat-sheet

For a reader who just wants the short version, arranged by the job each control does.

Slot identity

Windowing and timing

Estimation character

Smoothing

Blend steering

Pane furniture

Four settings traps to remember

1. Hidden is not excluded. Hide Plot hides the line. The slot still steers the blend and still votes in alignment alerts. Use Blended Weight = 0 to remove from the blend. Use Enable CVD NN = false to remove from alerts.

2. Weights are relative, not absolute. The blend is a weight-weighted mean. Making all slots 50 is identical to making them all 33.3. The question is not "do they sum to 100?" but "what is each slot's share relative to the others?"

3. Power User fields are dormant unless their family is live. Changing ALMA Sigma on a slot whose MA type is SMA changes nothing. Confirm the MA type before you tune.

4. The OB/OS levels are decoration, not triggers. There is no "line crossed 80" alert. If you want a state-based alert, see the per-slot bullish and bearish alerts documented in Alerts.

Where to go next

• To see what each control actually draws on the pane, and how to read the slots against the blend, Visuals & Logic.

• Before you flip On Bar Close? to OFF on a slot you intend to trust, MTF & Repainting.

• For documented setup recipes that use these controls together, Workflows.

• For the alert surface these settings feed into, Alerts.

• For the honest limits on what Pressure Sensitivity and Wick Weight can and cannot tell you, For the Geeks and Limitations & Trust Boundaries.