Settings

Settings

This indicator has the widest settings surface of any Axiom oscillator. Ten independent MACD slots, each with its own timeframe, ticker, MA types, lengths, weights, repaint toggle, and full Power User parameters — plus general oscillator settings, display controls, and a master smoothing pass.

This page is organized by decision priority, not by the order settings appear in the panel. Start at the top, work through the decisions that matter most, and you will find that many settings further down can be left at their defaults until your process calls for them.

How to navigate this page

Slot configuration decisions

Before touching individual settings, answer these questions in order. They determine the shape of your configuration.

1. How many slots do you need?

Three is the default and a solid starting point. The default setup — Slot 01 at 5m, Slot 02 at 15m, Slot 03 at 1H — gives you short, medium, and longer-term momentum on a common scale.

Add more slots when:

• You want to monitor a wider range of timeframes (e.g., 1m through daily)

• You want cross-market readings (e.g., chart ticker plus a correlated instrument on separate slots)

• You want to isolate a slot as a standalone monitor (weight 0) while the blend tracks a different set

Do not add slots for the sake of coverage. Four slots on 5m, 10m, 15m, and 20m are all reading similar price action at slightly different lags. That creates the appearance of multi-timeframe depth without actually providing it. Each slot you add should represent a meaningfully different perspective — a different timeframe, a different instrument, or a different MA configuration with a specific reason behind it.

2. Which timeframes?

Every slot's timeframe must be equal to or higher than the chart timeframe. The script enforces this — a lower slot timeframe causes a runtime error.

Choose timeframes that are genuinely distinct. A useful rule of thumb: each slot's timeframe should be at least three to four times the previous slot's timeframe. Slot 01 at 5m, Slot 02 at 15m, Slot 03 at 1H, Slot 04 at 4H creates four levels that are meaningfully separated. Slot 01 at 5m, Slot 02 at 10m, Slot 03 at 15m creates three levels that are too close together to offer independent views.

An empty timeframe field means the slot requests the chart timeframe. This is valid, but the slot still runs through the same slot-request path as every other slot. That means On Bar Close still matters: ON uses the prior confirmed chart bar, OFF uses the current chart bar.

3. Which weights?

Weights control each slot's contribution to the blended output. They are relative — weights of 10/10/10 produce the same blend as 33.3/33.3/33.3 because the tool normalizes them automatically.

Equal weights are the default and give every enabled slot the same voice. This is a reasonable starting point, but it means a three-slot blend is equally influenced by your shortest and longest timeframe.

Unequal weights let you tilt the blend toward the perspective you care about most. If you want the blend to respect the higher-timeframe trend and only flip when the longer view agrees, put more weight on the higher-timeframe slot. If you want the blend to be responsive to short-term momentum shifts, weight the short-timeframe slot higher.

A weight of zero is a deliberate design choice, not a disabled setting. A zero-weight slot still plots, still computes, and still fires alerts — it just does not influence the blend. This lets you monitor a slot independently while keeping the blend clean. See Workflows for concrete patterns using this approach.

4. Do any slots need a different MA type?

The default MA type is EMA, and for most traders that is the right choice. EMA is the standard in MACD and produces the behavior most people expect.

The MA type dropdown exposes the Axiom MA library entries available in the current build — standard types like EMA/SMA/RMA/WMA/VWMA/HMA and specialty/adaptive types like ALMA, KAMA, JMA, FRAMA, T3MA, VAMA, ZLMA, ZLEMA, Laguerre, McGinley, and others. Check the dropdown in the script if you need the exact current list; do not rely on a frozen list in the manual.

This depth exists for traders whose process specifically calls for a different MA characteristic — not as a buffet to experiment with randomly. If you do not already know why you would want KAMA instead of EMA, leave the setting at EMA. The manual will not tell you which MA type is "best" because none is. Each type has different lag, smoothness, and noise characteristics, and the right choice depends on the instrument, the timeframe, and the trader's methodology.

Important: The MACD MA Type and the Signal MA Type are independent. The MACD MA Type controls the fast and slow MAs that produce the MACD line. The Signal MA Type controls the smoothing applied to the MACD line to produce the signal (D) line. You can run EMA on the MACD side and ALMA on the signal side if your process calls for it. Both use the same MA library.

5. Does each slot need its own repaint setting?

Each slot has its own On Bar Close toggle (in the Power User section). The default is ON for all slots, which means each slot uses the prior confirmed bar for its requested timeframe and does not repaint.

Turning On Bar Close OFF on a slot makes that slot update intrabar as the requested-timeframe bar develops. The reading is faster but provisional — it will change when that bar closes. If that slot has nonzero weight, the blended output inherits some unconfirmed data. The blend does not label which portion is confirmed and which is not. If the slot is set to the chart timeframe, ON still keeps the slot one confirmed bar behind and OFF uses the current chart bar.

This is a per-slot decision, not a global switch. See MTF and Repainting for the full explanation and a live verification walkthrough.

Per-slot settings reference

Each of the ten slots has the same set of settings. Slots 01–03 default to enabled with timeframes 5m/15m/60m and equal weights. Slots 04–10 default to disabled with weight 0.

Slot colors

Each slot has a fixed color pair — a brighter shade when the slot is bullish (K > D) and a more transparent version when bearish (K < D):

Colors are not configurable. They exist to make it visually possible to distinguish ten slot lines in a single pane. If you have five or more slots enabled, the color differences become important — learn which color maps to which slot so you can trace individual readings without opening the settings panel every time.

What changing a setting actually does to the blend

This is the part people often miss. The blend is a weighted average across all enabled slots with nonzero weight. Every time you change a slot setting that affects its K/D/H values — timeframe, lengths, MA type, source, ticker — you change that slot's contribution to the blend. If the slot has meaningful weight, the blended output shifts.

Changes that affect the blend: enabling/disabling a slot, changing weight, changing timeframe, changing lengths, changing MA type, changing source, changing ticker, changing On Bar Close, changing ATR Sensitivity.

Changes that do not affect the blend: hiding a plot (the slot still contributes), changing line width, changing slot color (not configurable anyway).

The weight system

Weights are the most important conceptual setting in this tool. They determine what the blended reading actually measures.

How weights work

Each slot has a weight. The blend computes a weighted average: for each of K, D, and Histogram, the tool sums (value x weight) across all contributing slots, then divides by the total weight. Slots with weight 0 or na values are excluded from both the numerator and the denominator.

Weights auto-normalize. They do not need to sum to 100. Three slots at 10/10/10 produce the exact same blend as three slots at 33.3/33.3/33.3 or three slots at 500/500/500. Only the ratios matter.

What different weight distributions feel like

Common weight mistakes

Mistake: Adding a new slot without adjusting existing weights. If you have three slots at 33/33/33 and enable Slot 04 at weight 33, each slot now has 25% effective weight. The existing slots' contribution dropped from 33% to 25% each. This is arithmetic, not a bug, but it catches people off guard — especially when the new slot is on a very different timeframe and suddenly changes the character of the blend. If the blend was tracking the 5m/15m/1H relationship and you add a 4H slot at equal weight, the blend now leans toward the longer view in a way you may not have intended.

Mistake: Treating weight 0 as "off." Weight 0 does not disable the slot. It excludes the slot from the blend. The slot still plots its K line, still computes, and still fires alerts. If you want a slot gone entirely, disable it.

Mistake: Assuming higher weight means more reliable. Giving a slot more weight means the blend listens to it more. Whether that slot's reading is more reliable depends on the timeframe, the instrument, and the market conditions — not the weight you assigned.

ATR Sensitivity

This is the most misunderstood setting in the indicator. It deserves its own section because it affects every slot and changes how you should interpret every reading.

What it does

ATR Sensitivity is a multiplier applied to the volatility-normalized MACD value before the bounding function maps it to the -100/+100 range. Higher sensitivity means smaller raw momentum moves produce larger oscillator values. Lower sensitivity means larger raw momentum moves are needed to reach the same oscillator reading.

What it feels like at different values

Sensitivity 0.3 — compressed. The oscillator stays clustered near zero. Readings rarely reach the overbought (+70) or oversold (-70) thresholds. This gives you maximum resolution in the mid-range but makes the extremes almost unreachable. Useful if you want to distinguish between small and moderate momentum moves with precision, and you do not rely on the OB/OS thresholds for anything.

Sensitivity 1.0 — default. The oscillator uses a moderate portion of the range. During normal conditions, readings stay roughly between -60 and +60, with excursions to the extremes during strong moves. This is a balanced starting point for most instruments and timeframes.

Sensitivity 2.5 — saturated. The oscillator spends most of its time near the ±100 boundaries. Even moderate momentum moves push the reading to +85 or -90. You lose mid-range resolution — everything looks either "strongly bullish" or "strongly bearish" with little gradation in between. The OB/OS thresholds are crossed frequently and lose their attentional value.

The tradeoff in plain terms

Higher sensitivity gives you more range usage at the cost of mid-range resolution. Lower sensitivity gives you more mid-range resolution at the cost of range usage. There is no correct answer. It depends on what you are trying to see.

If you use the OB/OS levels as attention thresholds, you want sensitivity calibrated so that those levels are reached regularly but not constantly. If the oscillator never touches +70, the threshold is not doing any work. If it spends half the session above +70, the threshold is not useful either.

If you use the oscillator mainly for regime direction (K above or below D) rather than absolute level, sensitivity matters less — the regime flip point does not change much with sensitivity, because it is driven by the relationship between the MACD and signal lines, not by the absolute level.

Verification

Change ATR Sensitivity from 1.0 to 2.5 on any chart and watch the oscillator's range usage change. Then set it to 0.3 and watch everything compress toward zero. Return to 1.0. This takes less than a minute and builds more intuition than any description can.

See For the Geeks for the deeper explanation of why sensitivity interacts with the bounding function the way it does.

General oscillator settings

These settings apply to all slots equally.

ATR Length guidance

ATR Length determines how the normalization denominator behaves. At the default of 14, the normalization adapts to the instrument's recent volatility at a moderate pace. If you trade an instrument with stable, predictable volatility, 14 is fine and you may benefit from extending to 20 or 30 for an even smoother baseline. If you trade an instrument with erratic, fast-shifting volatility (some crypto pairs, meme stocks, illiquid instruments), a shorter ATR Length (7–10) makes the normalization track current conditions more closely — at the cost of a noisier denominator.

Changing ATR Length does not change what the MACD is measuring. It changes the ruler the oscillator uses to scale the measurement. A longer ruler is more stable but slower to adapt. A shorter ruler adapts faster but jitters. And when that jitter is large enough — say, ATR drops 40% after a volatility spike rolls off the lookback window — the oscillator reading can jump noticeably even if the raw MACD barely moved. You see the oscillator spike and think momentum changed. What actually changed was the denominator. If you notice readings that seem to react when price is quiet, check whether ATR just shifted.

Display settings

These are visual preferences. They do not affect computation, blending, or alerts.

Master Smoothing

Master Smoothing is an optional post-blend filter. When enabled, it applies a final moving average smoothing pass to the already-blended K, D, and Histogram values. It does not affect individual slot calculations — only the blended output.

When Master MA Type is set to a non-EMA type, the corresponding Power User parameters become active (Master ALMA Offset/Sigma/Floor, Master KAMA/FRAMA Fast/Slow, Master Jurik Phase/Power, Master Laguerre Alpha, Master VAMA Vol Length).

The tradeoff

Master Smoothing reduces noise in the blended reading. It also adds lag. Every bar of smoothing length delays the blended K/D crossover — not by an exact count, but roughly proportional to the Master Length.

A trader who uses the blended regime flip as a context indicator — not a timing signal — may benefit from smoothing because it filters out minor whipsaws. A trader who uses the K/D crossover for timing decisions will find that smoothing makes the crossover later, which may cause missed entries or late exits.

This is not a "turn it on for better results" setting. It is a deliberate tradeoff between noise reduction and responsiveness. If you are not sure whether you need it, you probably do not. The default is off.

See Workflows for a concrete pattern using master smoothing as a regime filter.

Per-slot Power User settings

Each slot has a set of Power User parameters for the MACD side and the D side. A given advanced parameter only matters when its corresponding MA type is selected. If the slot's MACD MA Type is EMA (the default), the MACD-side advanced parameters do nothing. If the Signal MA Type is EMA, the D-side advanced parameters do nothing.

The same set exists twice per slot: once for the MACD MA Type (controlling the fast/slow MAs) and once for the Signal MA Type (controlling the signal line smoothing, prefixed "D" in the settings panel).

When to touch these settings

Short answer: when you have deliberately chosen a non-EMA MA type because your process requires its specific behavior, and you want to fine-tune that behavior.

If you selected KAMA because you want an adaptive MA that smooths more during ranging conditions and responds more during trending conditions, the KAMA Fast and Slow periods control how quickly that adaptation happens. If you selected ALMA because you want the Gaussian weighting with a specific offset, the ALMA parameters shape that weighting.

These are not "advanced settings most traders will not need." They are depth, available when the trader's process calls for it. The default values are reasonable starting points for each MA type. Adjust when you understand the specific behavior you are trying to achieve.

Power User parameters for Signal (D) line

The same parameter set applies independently to the Signal MA Type. These are labeled with a "D" prefix in the settings panel: D ALMA Offset, D ALMA Sigma, D ALMA Floor Offset?, D KAMA/FRAMA Fast, D KAMA/FRAMA Slow, D Jurik Phase, D Jurik Power, D Laguerre Alpha, D VAMA Vol Length.

This means you can run EMA for the MACD MAs and ALMA for the signal line (or any other combination) with independent parameters for each side. The use case: fine-tuning the signal line's responsiveness independently of the MACD line's behavior.