Settings

This page is the single reference for every high-impact knob on the Base trim. The inputs dialog groups them into BB 01, BB 02, BB 03, BB Blended, and PU Settings. This page re-orders them the way you actually use them: by slot lifecycle, then the blended inputs, then the one global toggle. The GUI order is a catalog; the order below is a working sequence.

Before the tables, one distinction that deserves its own paragraph because confusing the two is the single most expensive settings mistake on this indicator.

Length: and Basis Trend Length: are not the same knob.

• Length: is the Bollinger lookback. It drives the standard-deviation window and, for the Lite MA types that accept a length, the basis moving-average window. SWMA is the exception in the imported Lite library; its basis uses Pine's fixed SWMA calculation while Length: still drives the band deviation window. Raising Length: usually slows the default SMA basis, but it does not mechanically widen the envelope on every chart.

• Basis Trend Length: is a lookback used only to decide whether the basis is rising or falling. The indicator compares the current basis value to the basis value Basis Trend Length bars back, and reports uptrend or downtrend on that comparison. Basis Trend Length: does not change the band, the basis curve, or the envelope at all. It only changes how quickly "rising" flips to "falling" and vice versa.

If you catch yourself reasoning about "a longer length makes the trend less twitchy," stop. That is two knobs you are collapsing into one word. The twitchiness of the trend read is Basis Trend Length:. The smoothness of the basis itself is Length:. Two knobs. Two jobs.

One worked example to lock this in. Put Slot 01 at 5 minutes with Length: 20 and Basis Trend Length: 3. Flip a chart with a visibly flat higher-timeframe basis. The per-slot basis-trend alert for Slot 01 will still flicker as the basis drifts past its own value three bars ago — because Basis Trend Length: 3 is a very short memory. Now set Basis Trend Length: to 20 without touching Length:. The band is unchanged. The basis curve is unchanged. The flicker calms down because you widened the memory the trend test uses, not the lookback the band uses.

With that out of the way, here is the full knob surface.

1. Per-slot primary inputs (three slots)

Every slot exposes the same inputs. The table below is written for Slot 01; Slot 02 and Slot 03 behave identically with their own defaults.

Moving-average types available on Type: are the Lite library's set. Drilling into moving-average semantics is out of scope for this page; cross-link at Axiom MA Library (Lite). The Pro library exposes additional types and per-moving-average parameters (ALMA offset, Jurik phase, KAMA/FRAMA lengths, Laguerre alpha, VAMA vol length). Those are not available on this trim. If you need them, you are looking for the CTX trim.

A deeper look at Source:, TimeFrame:, and the higher-timeframe context

Source: is evaluated inside the slot's chosen timeframe. When you set Slot 01 to TimeFrame: 5 and leave Source: at close, you are asking for the 5-minute close, not the 1-minute close resampled to 5 minutes. That is a real distinction — if you change Source: to hl2, the indicator reads hl2 on the 5-minute bar, not on the chart bar and not on some averaged construction.

A practical consequence: if you want to cross-check a slot against a stock Bollinger indicator, set the slot's TimeFrame: equal to your chart timeframe. The slot's bands should match a native Bollinger with the same inputs exactly. If they do not, something is different between the two configurations — most commonly the moving-average type, because the Axiom moving-average library's SMA / EMA implementations match platform defaults while other types will not because they are different constructions, not different implementations of the same construction.

That cross-check is worth doing at least once before you act on any slot value in a live context. It is how you turn "this tool claims to compute a Bollinger Band on a higher timeframe" into "I watched a slot agree with a native Bollinger when I removed the MTF context, and now I have a reason to extend that trust to the MTF case." Confidence in a tool is not the same thing as trusting its documentation; confidence is what remains after a reproducible verification you ran yourself. This indicator is built so that verification takes minutes, not an afternoon.

A deeper look at Blended Weight:

The blended band divides each accumulated band by the total weight of the slots contributing, but it only writes a composite when the total weight is greater than zero. Three slots at 33.3 each produce an equal-weight average. Slots at 50 / 30 / 20 produce a weighted average, where the 50-weight slot has exactly half the voice.

A weight of zero is a specific tool, not just a way to turn a slot off without turning it off. Weight zero keeps the slot's lines on the chart — you still see the context — and removes the slot's contribution to the blend entirely. Use it when a slot is worth looking at as a reference but you do not want it voting in the composite. Do not use weight zero as a stand-in for "I'm not sure about this slot." Either the slot is steering the blend or it is not.

The script does not declare a minimum on Blended Weight:. That means it also does not protect you from negative weights or from setting every contributing weight to zero. For a normal read, keep weights at zero or positive and make sure the enabled, contributing slots add up to a positive total. If the total is zero or negative, the blend falls back to zero values instead of producing a meaningful band.

Think of the weights as a sentence you are writing to future-you. 33.3 / 33.3 / 33.3 says "I find the three timeframes equally informative for the question I am asking." 60 / 30 / 10 says "Slot 01's timeframe is carrying most of my attention; Slot 03 is here for context more than for voice." 50 / 50 / 0 says "two slots are speaking; the third is visible only." Each of those is a defensible configuration. None of them is an accident. If you cannot read the weights back to yourself as a sentence you believe, they are the wrong weights.

2. Blended band inputs

If you want the blended band gone entirely — no math, no alerts — turn Enable Blended BB off rather than hiding the plot.

3. The one global PU toggle

This one switch is the single highest-stakes decision on the indicator. The tradeoff is real in both directions and the scope is every slot at once. Before you flip it, read MTF and Repainting end to end.

4. Dependency and interaction map

A few relationships between knobs are worth naming explicitly, because they are invisible in the input dialog:

• Slot TimeFrame: must be at or above the chart timeframe. Violating that raises a runtime error that names the offending slot.

• Hide BB 0N Plot is a cosmetic toggle only. The slot is still computed, still contributes to the blend at its assigned weight, and still counts in the alignment alerts.

• Blended Weight: 0 keeps the slot visible and removes it from the blend. It does not disable the slot. Alignment alerts still include it.

• Enable BB 0N: false is the only input that removes a slot from the alignment alerts.

• Enable Blended BB: false suppresses the blend math entirely. The per-slot plots are unaffected.

• The blended basis-trend read resolves by weight-majority across enabled, non-zero-weight slots when the contributing weights sum above zero. Ties count as uptrend. If the total weight is zero or negative, the blend state falls back to the script's zero-value/default-down edge instead of a useful vote. See For the Geeks for the reasoning.

• On Bar Close? governs every slot together. It is a single switch; there is no per-slot override on this trim.

5. Defaults as defaults

The shipped defaults exist so the indicator loads cleanly on a fresh chart and so you can see what three slots at three timeframes looks like before you commit any configuration decisions. They are not a ranking. They are not a starting point someone tested against a benchmark. They are a shape that shows you the surface.

Treat them as scaffolding. When you tune, tune on purpose. When you leave a default untouched, do it because you know what it is doing, not because you did not have a reason to change it.