Settings

# Settings

Every input the indicator exposes is documented here, grouped by function. For each setting, you will find the default, what it changes in practice, the tradeoff, and where it can hurt you if pushed too far.

The settings are organized in the order you are most likely to need them, not the order they appear in the panel.

Slot settings

The indicator has three independent RSI slots: RSI 01, RSI 02, and RSI 03. Each slot has an identical set of controls. The defaults differ only in timeframe (5-minute, 15-minute, 60-minute) and line color. Everything described below applies to all three slots unless noted otherwise.

Enable RSI 01 / 02 / 03

Default: on (all three enabled)

Toggles the entire slot. When a slot is disabled, it does not calculate, does not plot, does not contribute to the blend, and does not fire alerts. Disabling is clean — it removes the slot from everything.

Hide RSI 01 / 02 / 03 Plot

Default: off

Hides the slot's line from the chart, but the slot still calculates, still feeds into the blend, and still fires alerts. This is useful when you want a slot's data in the composite without the visual clutter of its individual line.

Watch out: a hidden slot that is pulling the blend in a direction you do not expect will be invisible on the chart. If the blended line is behaving in a way that does not match what the visible slot lines suggest, check whether a hidden slot is the reason.

Source

Default: close

The price input for the raw RSI calculation. Standard options: open, high, low, close, hl2, hlc3, ohlc4, hlcc4. Close is the textbook RSI source. Changing it is rarely necessary, but hl2 or hlc3 can slightly alter how the RSI responds to volatility.

If you are using Optional Ticker (see below), the source is evaluated on the overridden symbol, not the chart symbol.

TimeFrame

Defaults: 5-minute (Slot 01), 15-minute (Slot 02), 60-minute (Slot 03)

The timeframe on which the slot runs its RSI calculation. This is the core of the multi-timeframe architecture — each slot looks at a different window of price data.

Hard constraint: the slot timeframe must be equal to or higher than the chart timeframe. If it is lower, the script stops with a runtime error. There is no workaround. You must either lower the chart timeframe or raise the slot's.

What to think about when changing this:

• The defaults give a short/medium/long spread for intraday trading on 1-to-5-minute charts. If you trade on a 15-minute chart, you would need to raise Slot 01's timeframe to at least 15 minutes.

• Setting a slot to a very high timeframe relative to the chart (e.g., Daily on a 1-minute chart) creates a slot that updates extremely slowly. With On Bar Close on, the line will hold flat for an entire trading day before jumping. This is not broken — it is the reality of what daily RSI looks like on an intraday chart.

• Setting all three slots to the same timeframe defeats the purpose. You get three copies of the same calculation and a blend that adds nothing.

• Leaving the timeframe empty inherits the chart timeframe. Two slots at chart timeframe and one at a higher timeframe gives a 2:1 weight to the chart-level RSI in the blend, which may or may not be what you want.

RSI Length

Default: 14

The lookback period for the raw RSI calculation. This is the classic RSI parameter. Shorter lengths (7, 9) react faster to recent price moves. Longer lengths (21, 28) are smoother and slower.

Tradeoff: a shorter RSI Length makes the raw RSI more volatile. If you then apply heavy RSI Smoothing to tame it, you are partially undoing the shorter length. The two settings interact — shortening RSI Length to 7 while raising RSI Smoothing to 9 can leave you with a line that reacts about as slowly as the default 14/3 combination, but with different smoothing characteristics and a different lag profile. If you change one, check whether the other still makes sense.

RSI Smoothing

Default: 3

The length of the moving average that smooths the raw RSI into the slot's displayed RSI line. This is the first smoothing pass in the indicator's chain.

A value of 1 with SMA effectively passes the raw RSI through unchanged. Higher values calm the line and reduce whipsaws, but add lag — the line responds more slowly to new price data.

This is where smoothing stacking starts to matter. RSI Smoothing is the first MA pass, and it changes the plotted slot RSI line. Signal Length is the second MA pass, but it acts on the internal Signal path rather than moving the slot plot itself. Master Smoothing (if enabled) adds another MA pass to blended RSI and blended Signal after weighting. Stack all three carelessly and the composite picture gets calmer by getting later. See For the Geeks for a concrete breakdown of how the layers stack.

RSI Type

Default: SMA

Selects the moving average used for RSI Smoothing. Eight options are available: SMA, EMA, RMA, WMA, VWMA, HMA, ALMA, and SWMA.

The name is misleading. RSI Type does not change the RSI calculation itself. It selects the post-RSI smoothing MA. The raw RSI is always calculated the same way regardless of this setting. If you set RSI Type to EMA, you get an EMA-smoothed RSI — not a different kind of RSI.

Each MA type has different lag and responsiveness characteristics:

If you do not have a strong reason to change this, SMA or EMA are solid defaults.

Signal Length

Default: 3

The length of the moving average that smooths the RSI line into the Signal line. This is the second smoothing pass. The Signal line determines regime — when RSI is above Signal, the slot is bullish; when below, bearish.

Longer Signal Length creates a slower Signal that does not track RSI as closely, which means the regime holds longer and flips less often. Shorter Signal Length makes the Signal hug RSI tightly, which means more frequent regime changes and more whipsaw in choppy conditions.

This interacts directly with RSI Smoothing. Both smooth the same data in sequence. If RSI Smoothing already made the line calm, a long Signal Length on top of that pushes the regime detection further behind current conditions.

Signal Type

Default: SMA

Same options and same considerations as RSI Type. Selects the MA used for the Signal smoothing pass.

Blended Weight

Default: 33.3

Controls how much influence this slot has in the blended oscillator. Weights are relative, not absolute — they do not need to sum to 100. The indicator normalizes automatically. Weights of 10, 20, and 30 produce the same ratios as 33.3, 66.6, and 100.

Weight 0 is a special case. A slot with weight 0 still calculates, still plots, and still fires alerts — it just does not contribute to the blend. This is different from disabling the slot entirely. Weight 0 is useful when you want to see a slot's data and get its alerts without letting it influence the composite.

Extreme asymmetry has a cost. If you set one slot to 100 and the others to 1, the blend is dominated by a single slot. The blended line will nearly mirror that slot, which means you are not getting meaningful multi-timeframe synthesis — just a dressed-up single-slot view. The blend will look like a composite but behave like a copy. The danger is that it still feels like a multi-timeframe reading, which gives unearned confidence.

Line Width

Default: 2

Visual thickness of the slot's RSI plot line. Cosmetic only — it changes nothing about the calculation or behavior.

Optional Ticker

Default: empty (inherits the chart symbol)

Overrides the symbol for this slot. When set, the RSI calculation runs on the specified symbol instead of the chart symbol. The output is still on the -100 to +100 bipolar scale, so it blends with other slots regardless of the symbols' price differences.

This is a power-user feature. Blending RSI from different symbols produces a composite number that is mathematically valid but not automatically meaningful. The blend does not measure correlation between the symbols, does not account for regime differences, and does not imply that the symbols are moving for the same reason. See Multi-Ticker Mixing for a full treatment of when this is useful and where it misleads.

General oscillator settings

Overbought Level

Default: 70

Sets the upper dashed reference line and the threshold for the "Blended RSI Overbought" alert.

This value is on the bipolar scale. +70 here corresponds to roughly RSI 85 in standard terms. That is substantially more extreme than the traditional RSI 70 overbought line. If you set this expecting the same frequency of overbought events you are used to from standard RSI, you will see far fewer. To approximate the standard RSI 70 level on this oscillator, you would set the overbought level to +40.

Oversold Level

Default: -70

The lower threshold. Same scale logic as above. -70 on the bipolar scale corresponds to roughly RSI 15 — again, much more extreme than the traditional RSI 30 oversold level.

Display settings

Plot Blended RSI/Signal

Default: on

Toggles the blended RSI line, blended Signal line, and the fill between them. When off, only the individual slot plots are visible.

Turning this off is useful when you want to focus entirely on individual slot behavior without the composite overlay. The blended lines can visually dominate the pane, especially with the default line width of 3.

Blended Line Width

Default: 3

Visual thickness of the blended RSI and Signal lines. Cosmetic only.

Master smoothing settings

Enable Master Smoothing

Default: off

When enabled, adds a final MA pass to the blended RSI and Signal lines after blending. It does not touch the individual slot plots. What it does is slow the composite outputs even further, which can make the display look cleaner while also making it later.

Use this with eyes open. The indicator is already downstream of the RSI lookback and the per-slot smoothing chain. Adding Master Smoothing produces a very calm, very stable composite - and a composite that is significantly delayed. It is useful when the blended outputs are noisy from diverging slots, but the cost is always more lag.

Master MA Type

Default: EMA

Selects the moving average for the master smoothing pass. Same eight options as RSI Type and Signal Type.

Master Length

Default: 3

The lookback period for the master smoothing MA. Higher values produce more smoothing and more lag.

Power user settings

On Bar Close?

Default: on

This is the most consequential setting in the indicator. It controls whether higher-timeframe slot data comes from the last confirmed bar or the current building bar. It applies globally — all slots use the same mode.

• On (default): each slot uses the previous fully confirmed higher-timeframe bar's values. The display is stable and will not change until the next HTF bar closes. What you see live matches what the chart will show in history. This is the non-repainting mode.

• Off: each slot uses the current building higher-timeframe bar's values. The slot updates on every chart-timeframe tick. The display is more responsive but the reading is provisional — it can change with every new tick and may look different once the HTF bar closes and the history is finalized.

If you are not sure which to use, leave it on. If you are considering turning it off, read MTF and Repainting first. The risk is specific: with this off, the display updates on every tick, the reading looks current and definitive, and you make a decision based on what you see. Then the higher-timeframe bar closes, the final value is different from the interim one, and the chart rewrites the history. The decision you made and the chart you are now looking at no longer agree. That disconnect is invisible unless you know to expect it, and it compounds over time into a kind of confidence erosion that is hard to trace back to a settings toggle.

ALMA Offset

Default: 0.85

Controls the offset parameter for the ALMA algorithm. Only relevant when at least one MA type selection (RSI Type, Signal Type, or Master MA Type) is set to ALMA.

Critical: this setting is global. The same offset applies to every ALMA instance across all three slots and the master smoother. If you change it to tune one slot's ALMA behavior, every other slot using ALMA changes too. There is no per-slot ALMA configuration.

ALMA Sigma

Default: 6.0

Controls the sigma (width) parameter for ALMA. Same global scope as ALMA Offset.

ALMA Floor Offset?

Default: off

When enabled, floors the ALMA offset to an integer value. Same global scope as the other ALMA settings.

The smoothing stack at a glance

Understanding how the smoothing layers interact is one of the most important things in this manual. Here is the full chain:

One correction that matters here: this is not one straight conveyor belt. RSI Smoothing shapes the plotted slot RSI line. Signal Smoothing shapes the internal Signal path. Those two paths are blended separately and only then optionally master-smoothed.

Raw RSI (ta.rsi)  -> RSI Smoothing (first MA pass)  -> slot RSI value  -> bipolar conversion  -> displayed slot RSI lineSmoothed slot RSI value (still on the standard RSI scale)  -> Signal Smoothing (second MA pass)  -> slot Signal value  -> bipolar conversion  -> internal slot SignalEnabled slot RSI values  -> weighted blend  -> blended RSI raw  -> optional Master Smoothing  -> final Blended RSIEnabled slot Signal values  -> weighted blend  -> blended Signal raw  -> optional Master Smoothing  -> final Blended Signal

Each layer adds lag, but not all on the same output. The slot RSI plot, the internal Signal path, and the blended outputs each carry their own lag profile. A setting of 3 at each layer does not mean "3 bars of lag" - it means multiple separate smoothing operations whose delays stack in different places.

Rule of thumb: if the blended line looks very smooth and very stable, ask yourself how much of that stability is real momentum and how much is manufactured delay. Toggle Master Smoothing off. Reduce Signal Length to 1. See what the less-smoothed picture looks like. If the less-smoothed version is telling a meaningfully different story — choppier, or already turning while the smoothed version held steady — the smoothing was not cleaning the data. It was hiding a change. You can always add smoothing back. You cannot unsee what the raw reading tells you.