Settings

This page is the knob-by-knob reference for Axiom MACD Osc CTX. Every input is documented as a deliberate decision with a consequence, not a feature checklist. If you are looking for "best settings," you will not find them here; there are no best settings for a workbench that has to fit a different operator every time. What you will find is what each control does, what it costs, and what to watch for when you move it.

Read it top to bottom the first time. Come back to individual sections when you are configuring. The default tripwires section near the bottom of the primary-inputs block collects the mistakes the inputs most commonly hand to new users; skim that section even if you think you know what you are doing.

How the input panel is organized

The TradingView inputs dialog groups controls into three kinds of bands:

• Oscillator, Display, Master Smoothing — three short groups that govern the pane as a whole.

• MACD 01 through MACD 10 — ten groups, one per slot, each with the same primary inputs. Slots 1 through 3 are enabled by default; slots 4 through 10 are disabled with weight 0.

• MACD 01 PU through MACD 10 PU — ten Power User groups, one per slot, parallel to the primary slot groups. Each carries the slot's On Bar Close? control and the advanced parameters that apply only when the slot's chosen MA family needs them.

The pattern is consistent. Learn slot 01's inputs once; slots 02 through 10 are the same controls with the same semantics.

The "Oscillator" group — the pane's global dials

ATR Length

Type: integer. Default: 14. Minimum: 1.

The lookback that every slot uses to measure its own ATR. Each slot computes ATR in its own request.security context, which means a 5-minute slot uses 5-minute ATR, a 60-minute slot uses 60-minute ATR, and a slot with an optional ticker uses the optional ticker's ATR on the slot's chosen timeframe. ATR Length sets how many bars of ATR history each slot averages over.

Lower ATR Length (say 7) means the yardstick adapts quickly to recent volatility. A sudden spike in range will shrink normalized values across all slots until the spike is absorbed into the average. Higher ATR Length (say 50) means the yardstick is steadier — a single spike will barely budge it — but also slower to reflect a genuine regime change.

The trap to watch for is using a short ATR Length in an environment where ATR is itself unstable (news-driven ranges, opening minutes, illiquid overnight). The pane will feel jumpy in ways that are not really about MACD; they are about the yardstick moving. If the pane is feeling jumpy and you can trace the jumpiness to ATR rather than to price, lengthen ATR Length before you reach for other knobs.

ATR Sensitivity

Type: float. Default: 1.0. Step: 0.1. Minimum: none enforced, but positive values are what the pane was calibrated on.

The pre-sigmoid scaling dial. Each slot's raw MACD, signal, and histogram are divided by that slot's ATR, then multiplied by ATR Sensitivity before being mapped through the sigmoid into 0 to 100.

• At 1.0, the pane behaves as calibrated. The slot fast line typically lives well inside the interior of 0 to 100 on normal evidence, drifts toward 30 or 70 on developed moves, and only approaches 0 or 100 on extreme stretches.

• Push ATR Sensitivity above 1.0 and the slot lines saturate faster. At 3.0, the slot lines will push against the boundaries during ordinary moves and pin near 0 or 100 during decisive ones. You are not "seeing more"; you are exaggerating the read. Sometimes that is what you want — to make a subtle divergence visible on a quiet instrument, for example — and sometimes it is covering over a genuinely flat pane with apparent action.

• Drop ATR Sensitivity below 1.0 and the slot lines flatten toward 50. At 0.5, even decisive underlying MACD moves may only produce slot values between 40 and 60. The pane looks calm. That calm is a choice you are making; it is not evidence of market calm.

The honest posture is to leave ATR Sensitivity at 1.0 until you can point to a specific reason to move it — and, if you do move it, move it on one slot at a time if the per-slot behavior is what you want to study. (There is no per-slot ATR Sensitivity; the dial is global. A single-slot study means pushing sensitivity up, watching a slot, and putting it back.)

Overbought Level and Oversold Level

Type: float. Defaults: 70 and 30. Step: 0.1.

Reference levels, drawn as dashed gray guides on the pane. Nothing inside this indicator fires on them. They do not gate the slot lines, they do not change the normalization, and no native alert watches them. They exist so the eye has stable brackets in the bounded pane.

Moving them is a preference. Running 80 / 20 instead of 70 / 30 gives you tighter guides that fewer slot movements touch. Running 65 / 35 gives you looser guides that more slot movements touch. Neither setting changes what any slot is doing under the hood; both are purely visual.

The trap to head off, loudly, because the pane looks so familiar: these are not RSI thresholds. The 0-to-100 pane is a stretch read on top of MACD/signal/histogram, not a measure of overbought-ness in the mean-reversion sense. Limitations & Trust Boundaries covers the RSI-misread trap in detail.

The "Display" group — what actually draws on the pane

Plot Blended K/D

Type: boolean. Default: true.

Controls whether the blended fast and blended slow lines draw on the pane. The blend's computation runs regardless — alerts based on the blend still fire whether the lines are visible or not, and the histogram columns still draw. This is a visibility control only.

If you find yourself watching the slot lines more than the blend and want a cleaner pane, turning this off is a legitimate choice. The blend is still there if you want to re-enable the lines later.

Plot Blended Histogram

Type: boolean. Default: true.

Controls whether the blended histogram columns draw. Independent of the blend lines; you can run any of the four combinations of blend-lines-on-or-off and histogram-on-or-off.

Leave this on unless you have a specific reason to hide it. The four-state histogram color code (Visuals & Logic teaches it in detail) is the cheapest place on the pane to see the convergence story accelerate or unwind. Readers who disable the histogram almost always end up reading the blend crossing its signal slightly late because they lost the "rising-above-50, falling-below-50" cue that was sitting right there.

Blended Line Width:

Type: integer. Default: 3. Minimum: 1.

Thickness of the blended fast and blended slow lines. Purely cosmetic. Set it to 4 or 5 on a dark theme if the blend is getting visually lost under the slot lines; set it to 2 on a bright theme if the blend is overwhelming them. The default keeps the blend heavier than the default slot lines (2), which is the intended hierarchy: the blend is the headline, the slots are the evidence.

The "Master Smoothing" group — the optional extra pass on the blend

Enable Master Smoothing

Type: boolean. Default: false.

When true, applies one additional MA pass to the already-computed blended fast, blended slow, and blended histogram. When false, the blend you see is the direct weighted average of the slot values with no extra smoothing on top.

The master pass is a visual filter for the blend. It adds lag in exchange for a calmer blend line. It does not change anything inside the slots — the slot lines, their colors, their per-slot alerts, and the four-state histogram coloring are all untouched. It does change the blend-based alerts (Blended MACD Is Bullish, Blended MACD Is Bearish, and by extension the alignment alerts remain independent, but the blend's own crossover events happen later under master smoothing).

The honest tradeoff:

• OFF (default): the blend is a faithful weighted average of the slot evidence. Noisy during chop; expressive during developed moves.

• ON: the blend reads more regime-like. Calmer through chop; delayed through turning points. Blend-based alerts are lagged. Slot lines are unchanged.

Master smoothing is a regime-read choice, not a correctness improvement. If you enable it because the blend "looks choppy," you are not fixing noise; you are deciding that you value a calmer regime view more than the faster read. That is a legitimate choice. Making it by accident is not.

Master MA Type

Type: enum. Default: EMA.

Which MA family drives the master pass. The full list of families comes from the underlying library (AxiomCharts/AxiomMovingAverageLibraryExtended/1); the input panel dropdown shows exactly what is available to this build, and the MA Family Parameter Table below names the families that unlock additional Power User parameters when selected.

If you are not sure which family to pick, leave it at EMA. EMA is the default for a reason — it is the textbook smoother that matches the MACD-family heritage of the indicator — and any other family's behavior should be a deliberate choice you can describe out loud. "This one looks cleaner" is not a deliberate choice; it is a preference for the visual character of one smoother over another without a defensible reason behind it. If you find the blend's real-time twitch distracting, that is usually a case for shorter per-slot lengths or different weighting, not for adopting an exotic master family.

Master Length

Type: integer. Default: 3. Minimum: 1.

How much smoothing the master pass applies. Length 3 on an EMA is a light touch — enough to reduce bar-to-bar noise without visibly lagging regime changes. Length 10 is moderate lag. Length 20 is heavy lag.

If you are enabling master smoothing, start at 3 and go up only if you can point to specific jitter you want to suppress. Heavy master smoothing makes the blend look confident in ways that are not earned; the underlying slot evidence has not changed, only the rate at which the blend absorbs it.

Master ALMA / KAMA-FRAMA / Jurik / Laguerre / VAMA parameters

Type: various (see table below). Exposed in the Master Smoothing group, but active only when Master MA Type is the corresponding family. If you leave the master family at EMA (or any family that does not use these parameters), these controls are inert.

See the MA Family Parameter Table for which parameter applies to which family.

The per-slot primary inputs — slot 01 as the archetype

Every slot group carries the same primary inputs. Slot 01 is documented here as the archetype; slots 02 through 10 carry identical controls with different defaults where noted.

Enable MACD 01

Type: boolean. Default: true (slots 01–03), false (slots 04–10).

The master switch for the slot. When false, the slot does not compute, does not plot, does not contribute to the blend, and does not fire any of its per-slot alerts or vote in the alignment alerts. An enabled slot is "in the math" in every sense; a disabled slot is not present at all.

This is the lever to reach for when you want a slot fully removed. If you only want it silent from the blend, use Blended Weight = 0 instead. If you only want the line hidden while the slot stays active, use Hide MACD 01 Plot. The three silences are not interchangeable; see Quick Start step 6 for a hands-on walk.

Hide MACD 01 Plot

Type: boolean. Default: false.

Hides the slot's line from the pane without changing anything else. The slot computes, contributes its weight to the blend, fires its per-slot alerts, and counts toward the alignment alerts exactly as before. The only difference is that the line is not drawn.

Useful when a slot's job is to feed the blend or fire its alerts and the line is visual clutter on an already busy pane. Not useful as a way to "remove" the slot.

Source:

Type: series. Default: close.

Which chart series the slot reads. Usually close, occasionally hlc3 or ohlc4 for smoother intra-bar behavior, sometimes high or low to study the specific side of the bar's action. The source is read on the slot's symbol (either the chart symbol or the optional ticker) at the slot's timeframe.

The source choice is most meaningful on slots where you want the slot's behavior to differ in character from the textbook close-based MACD. Running every slot on close and expecting diversification is like running ten copies of the same polygraph and expecting independent testimony.

TimeFrame:

Type: timeframe string. Default: "5" (slot 01), "15" (slot 02), "60" (slot 03), "" (slots 04–10).

The higher timeframe the slot runs on. "" (blank) means "use the chart timeframe." Any non-blank value means "use this timeframe, as long as it is greater than or equal to the chart timeframe."

The slot timeframe must be greater than or equal to the chart timeframe. Setting it lower throws a runtime error that names the offending slot — MACD 01 timeframe cannot be lower than the chart timeframe. MTF & Repainting walks why the rule exists; Troubleshooting reproduces the error text verbatim for quick recognition.

When a slot reads a higher timeframe, the slot's On Bar Close? setting (in the slot's PU group) decides whether the slot reports the previous closed HTF bar's value (stable, slower) or the live forming HTF bar's value (faster, repaints until the bar closes).

Fast Length:

Type: integer. Default: 12. Minimum: 1.

The fast MA length inside the slot's MACD calculation. With the default MA family (EMA), Fast Length = 12 and Slow Length = 26 reproduce the textbook MACD fast leg.

Fast Length must be strictly less than Slow Length. Setting them equal or reversing them throws a runtime error that names the slot — MACD 01 Fast Length must be less than Slow Length. Reproduced in Troubleshooting verbatim.

Slow Length:

Type: integer. Default: 26. Minimum: 1.

The slow MA length inside the slot's MACD calculation. Must be strictly greater than Fast Length.

Pushing Fast Length and Slow Length up together (for example, to 24 and 52) calms the slot — the underlying MACD evidence is smoother, the bounded line moves less — but adds lag. Pushing them down together (for example, to 5 and 13) sharpens the slot; the line reacts to smaller movements, at the cost of firing on noise.

MACD MA Type:

Type: enum. Default: EMA.

The MA family used for both the fast leg and the slow leg of the slot's MACD calculation. Available families are listed in the MA Family Parameter Table below.

EMA is the textbook MACD MA. Choosing anything else is choosing to study what that family's smoother-or-sharper profile does to the MACD read on this instrument. It is not inherently better or worse. If you are picking a non-default family without being able to describe the property of the family that matters to you, you are adding a layer you do not need.

Signal Length:

Type: integer. Default: 9. Minimum: 1.

The smoothing length applied to the slot's MACD to produce the slot's signal line. With the default family (EMA) and the default length (9), the slot produces the textbook MACD signal.

Signal MA Type:

Type: enum. Default: EMA.

The MA family used for the slot's signal line, independent of the MACD family. You can, for example, run an EMA MACD with a Jurik signal, or an ALMA MACD with an EMA signal. The Power User parameters for the MACD family and the signal family are separated in the slot's PU group as K (MACD) and D (signal).

Line Width:

Type: integer. Default: 2. Minimum: 1.

Thickness of the slot's line on the pane. Purely cosmetic. Useful when you want a particular slot to read heavier than the others — for example, setting MACD 03 (the 60-minute slot) to width 3 makes the higher-timeframe slot visually dominant when that is the slot you are leaning on most.

Blended Weight:

Type: float. Default: 33.3 (slots 01–03), 0.0 (slots 04–10). Minimum: 0.

How much this slot steers the blend. The blend is a weighted average across every enabled slot whose weight is greater than zero and whose values are not na. Slots with weight 0 are present in the pane — they plot (unless hidden), they fire their per-slot alerts, they vote in alignment — but they do not steer the blend.

Weights do not need to sum to 100. Equal weights are a starting point, not a recommendation. If you are leaning on a longer-timeframe slot for regime context, weighting it heavier is a legitimate choice; if you are adding a short-timeframe slot as a responsiveness contribution, weighting it lighter is equally legitimate. The weight is the operator's expression of which slot's voice matters more in the summary.

The trap to watch for: weight 0 silences the blend contribution but does not silence the alignment alerts. A weight-zero slot is still "enabled" for alignment. See Limitations & Trust Boundaries for the cost of that interaction.

Optional Ticker:

Type: symbol. Default: "" (blank).

When blank, the slot reads the chart's symbol. When set, the slot reads the given symbol at the slot's timeframe instead. The slot's line still draws on the current chart's pane, but the evidence behind the line is the optional ticker, not the chart symbol. ATR for that slot is computed on the optional ticker's OHLC too.

Use cases: a regime context slot (SPX or ES while trading a constituent), a correlated-asset slot (BTC while trading an alt, DXY while trading a pair), or a study comparison (the same instrument on a different venue).

The repeated trap to head off: a slot with Optional Ticker = "SPY" is information about SPY, not about the chart symbol. It is not confirmation. Limitations & Trust Boundaries and Workflows cover the cross-asset misread in detail.

Default tripwires — mistakes the input panel most commonly hands you

Read this once before you start tuning. It is not the full limits page; it is the set of small mistakes that bite first-day readers.

• The 60-minute slot needs an hourly bar under it. On a fresh load, MACD 03 may not draw for the first few minutes because its 60-minute signal has not formed. Scroll back a few days to seed history. If the line still refuses to draw after a week of backscroll, something else is wrong; start in Troubleshooting.

• The slot-timeframe rule is enforced at every edit. If you set the chart to 15m and one of your slots is still at 5m from a prior session, the pane will clear and display the runtime error. Change the slot to ≥ 15m or change the chart.

• Fast Length < Slow Length, strictly. Equal values trip the runtime error. The default 12/26 satisfies the rule; pushing either length past the other trips it.

• Weight 0 does not remove a slot from alignment alerts. If you want a slot silent from alignment, disable it. If you only want it silent from the blend, weight 0 is correct. If you want it silent from the pane but present everywhere else, hide the plot.

• Hide MACD NN Plot does not exclude the slot from the blend. It hides the line. See Quick Start step 6 or the description of Hide MACD 01 Plot above.

• Defaulting ten slots is the alignment trap. If you enable MACD 04 through MACD 10 with defaults and give them weights, you have ten slots reading the same MACD on slightly different cadences. Alignment will look like consensus and read like one slot copied ten times. Limitations & Trust Boundaries leads with this trap.

The per-slot Power User inputs — what matters when

Every slot has a parallel "PU" group that carries the On Bar Close? switch and the parameters that apply only when the slot's chosen MA family needs them. The PU controls are exposed in every slot regardless of which MA family the slot is using; the family decides which of the controls actually affect the slot's behavior.

On Bar Close?

Type: boolean. Default: true.

The repaint switch, one copy per slot. Always active; not tied to any MA family. When true, the slot reports the previous closed value from its request.security context — slower to react on the chart but stable on confirmed history, and the slot will not revise earlier bars. When false, the slot reports the live forming HTF bar's value — faster to react on the chart, and the slot's current value will be revised as the HTF bar develops.

On Bar Close? is the most consequential per-slot control in the pack. MTF & Repainting is the full treatment. The short version: ON is the default for a reason, OFF is a legitimate choice with a named cost, and mixing postures across slots (for example, ON on the 60m slot and OFF on the 5m slot) is allowed but should be a deliberate decision.

MA Family Parameter Table

Which Power User parameters matter for which MA family. If the slot's MACD MA Type picks family X, the K parameters for family X are active on that slot's MACD calculation. If the slot's Signal MA Type picks family Y, the D parameters for family Y are active on that slot's signal calculation. Parameters not listed for a family are inert for that family — they are exposed in the PU group but do nothing when that family is selected.

The left-column list is not the complete catalog — the dropdown in the input panel is the source of truth for which families your build exposes. What the table guarantees is which PU parameters light up when you pick ALMA, KAMA/FRAMA, Jurik, Laguerre, or VAMA. If you pick a family that is not one of those five, the PU parameters in the slot's PU group are inert for your slot even though the controls are still visible.

For the MACD leg, the controls are prefixed K (K ALMA Offset, K Jurik Phase, and so on). For the signal leg, they are prefixed D. That lets the reader run, for example, an ALMA MACD with a Jurik signal and tune each family's parameters independently inside the same slot's PU group.

Specific parameter ranges and defaults (reported by the inputs dialog):

• K ALMA Offset / D ALMA Offset: default 0.85, step 0.01.

• K ALMA Sigma / D ALMA Sigma: default 6.0, step 0.01.

• K ALMA Floor Offset? / D ALMA Floor Offset?: default false. Rounds the offset down before use.

• K KAMA/FRAMA Fast / D KAMA/FRAMA Fast: default 2, integer.

• K KAMA/FRAMA Slow / D KAMA/FRAMA Slow: default 30, integer.

• K Jurik Phase / D Jurik Phase: default 0, integer.

• K Jurik Power / D Jurik Power: default 2.0, step 0.1.

• K Laguerre Alpha / D Laguerre Alpha: default 0.5, step 0.01.

• K VAMA Vol Length / D VAMA Vol Length: default 20, integer.

These defaults are sensible starting points for each family. They are not "the right" settings; the operator tunes them to the behavior they want to study. If you are running EMA MACD with EMA signal (the out-of-the-box configuration), every parameter in the above table is inert for your slot, and you can ignore the PU group except for On Bar Close?.

Settings tradeoffs the pane asks you to own

Three tradeoffs sit above the individual controls. They are the decisions you are making, often implicitly, when you tune:

1. Comparability vs raw magnitude. The bounded 0-to-100 pane buys cross-timeframe and cross-instrument comparability at the cost of showing absolute MACD magnitude. A slot at 75 is stretched in its context; the slot does not say by how much in price units. If your process needs absolute magnitude, the pane is not the tool; For the Geeks is candid about why.

2. Responsiveness vs stability. On Bar Close? = OFF, higher ATR Sensitivity, shorter Fast Length / Slow Length, shorter ATR Length all push toward a more responsive pane that reacts to smaller movements. ON, lower sensitivity, longer lengths, longer ATR Length push toward a stabler pane that filters more. There is no free lunch. Every dial you pull toward responsive costs you stability; every dial you pull toward stable costs you responsiveness.

3. Summary vs evidence. Enabling master smoothing, disabling the slot plots, and leaning on the blend is a summary-heavy configuration. Disabling master smoothing, enabling more slots, and leaning on the slot lines is an evidence-heavy configuration. Both are legitimate. The trap is doing one while thinking you are doing the other — for example, enabling master smoothing and then reading blend-based alerts as if they were early signals on the unsmoothed evidence. They are not. They are lagged signals on the smoothed evidence.

When a setting choice feels ambiguous, ask which of these three tradeoffs you are touching. Often the answer makes the choice obvious.

Where to go next

• For the mental model of what the pane is plotting — slot lines, blend lines, histogram, reference levels — Visuals & Logic.

• For the full On Bar Close? treatment, MTF & Repainting.

• For how master smoothing interacts with the blend-based alerts, Alerts.

• For the named anti-patterns every operator of this pane should be able to describe, Limitations & Trust Boundaries.