Visuals and Logic
This is the page that installs the reading order you will return to on every session. It is the longest page in the pack you can afford to slow down on, because every minute spent on this page saves you from re-learni...
Written By Axiom Admin
Last updated 22 days ago
Visuals and Logic
This is the page that installs the reading order you will return to on every session. It is the longest page in the pack you can afford to slow down on, because every minute spent on this page saves you from re-learning the pane in the middle of a busy session later. The core move is small and mechanical: color first, value second, blend third, chart last. Everything below is an unpacking of that move and the reasons it matters.
What is on the pane
From bottom rail to top rail:
Lower boundary at 0, drawn in green as a rail. Lines that reach it are at the lowest value the frame allows, not at "the strongest oversold reading possible." A pinned 0 is the frame refusing to go lower, not a measurement.
Oversold reference at 30 by default, dashed gray. This is a user-configurable reference zone. No internal logic fires on it.
Mid line at 50, solid gray. RSI's native middle.
Overbought reference at 70 by default, dashed gray. Same design as the oversold line.
Upper boundary at 100, drawn in red as a rail. Same reading convention as the lower rail.
Three slot lines, one per enabled slot, drawn at per-slot line widths. Slot 01 teal, slot 02 aqua, slot 03 blue. Each slot line plots the slot's smoothed RSI.
Slot colors in full-tone when the slot's smoothed RSI sits above the slot's smoothed signal line, faded when it sits below. Full-tone is the slot's "leading" color; faded is the slot's "lagging" color.
A thicker blended RSI line drawn in lime when the blended RSI is above the blended signal, red when it is below. With master smoothing off, this line is the weighted average of the enabled, non-zero-weight slot RSI lines; with master smoothing on, it is that blend after one more MA pass.
A blended signal line drawn in gray at the same thickness as the blended RSI. With master smoothing off, this is the weighted average of the enabled slot signal lines; with master smoothing on, it is also smoothed one more time.
A tinted fill between the blended RSI and the blended signal. Lime-tinted when the blended RSI is above the blended signal; red-tinted when it is below.
What is not on the pane: there is no histogram here, and the per-slot signal lines are not drawn. If you are coming from the MACD Osc family you will notice the absence. RSI does not have a clean histogram analogue β it is already bounded by construction, and there is no fast/slow pair to subtract into a zero-centered bar. Rather than force a shape that would not teach anything, Base leaves the histogram space empty and lets the color-by-relationship on each slot line carry the directional read. For the blend, the RSI-vs-signal gap is visible because both blend lines are plotted. For individual slots, the signal relationship is visible through color, not through a second line you can measure with your eye.
Reading order β the move that matters
Read the colors first. For each enabled slot, look at whether the slot line is full-tone or faded. Then look at the blend's color β lime or red. That scan takes less than a second once it becomes a habit, and it answers the fast question the pane is built to answer: which slots have their smoothed RSI leading their own smoothed signal, and which do not?
Read the values second. Value tells you the intensity behind the color. A slot at 55 full-tone is leaning up; a slot at 75 full-tone is leaning up with more room already traveled; a slot at 25 faded is leaning down from an already-low level. The value by itself does not answer the directional question β the color does.
Read the blend third. The blend is a composite you shaped. It is the headline of the pane in the same way a magazine headline is a composite summary of the reporting underneath. Do not skip the slot scan and read only the blend; the blend can sit near 50 while the slots are pulling apart, and the blend can paint lime for a stretch while one or two slots underneath it have already started losing their full-tone color. See the section below on quiet blends.
Check the chart fourth. The chart is where the decisions get made; the pane is where the reading gets done. A chart read without the pane is the reader operating from instinct only; a pane read without the chart is the reader operating from pure abstraction. Both reads are part of the session.
The color discipline β RSI vs signal, not RSI vs midline
Here is the single most load-bearing correction this pack makes to a textbook RSI habit. On textbook RSI, a lot of readers will scan the line against the 50 midline and take direction from there. On this pane, that scan is not wrong β it is incomplete. The slot color is driven by a different comparison.
Color logic per slot:
Full-tone if the slot's smoothed RSI sits above the slot's smoothed signal line.
Faded if it sits below.
There is no comparison to 50 in that rule. The 50 midline is still on the pane as a zone boundary, but it is not the color trigger.
Consequence: a slot can be above 50 and faded. A slot can be below 50 and full-tone. Those are not bugs. They describe a state where the slot's smoothed RSI is above the midline but has recently fallen below its own smoothing (above 50, faded), or below the midline but recently risen above its own smoothing (below 50, full-tone). Either way, the color has moved before the line will cross the midline. That is part of what the second smoothing pass buys you β an earlier read on directional change, at the cost of a second pass of lag between raw RSI and what the color shows you.
Reading pair β two slots at value 55
To feel the reading order as a contrast rather than an assertion, picture two slots both sitting at 55:
Slot A at 55, full-tone color. The slot's smoothed RSI is above the slot's smoothed signal line. The slot's RSI is leaning up relative to its own recent smoothing. The value being above 50 is a secondary fact; the directional read is encoded in the color.
Slot B at 55, faded color. Same value. Different story. The slot's smoothed RSI is below the slot's smoothed signal. The slot was recently higher; its own smoothing has not yet caught up. Directional read is down even though the value is above 50.
A value-only read collapses those two into "both above the midline, leaning bullish." A color-aware read keeps the directional difference. This is not a subtle distinction β it is the distinction the pane is built around.
Warm-up behavior on slot color
During warm-up, the raw RSI has to exist before the slot can draw a real value. Once raw RSI exists, the code falls back from unavailable MA outputs to earlier pipeline values: the slot RSI can use raw RSI before its smoothing MA is ready, and the slot signal can equal the slot RSI before the signal MA is ready. That means early color can be flat or uninformative rather than a mature RSI-vs-signal read. The code does include a midline fallback for the rare case where RSI exists and signal is unavailable, but the normal fallback path gives the signal the RSI value. Do not treat the first warm-up bars as steady-state behavior.
The blend β what it is and what it is not
The blended RSI and the blended signal are weighted averages. For each enabled slot with a non-zero weight and a non-na RSI, the blender adds the slot's RSI scaled by its weight to a running total and divides by the total weight at the end. Same for the signal. If no slot qualifies β every slot disabled, every weight zero, every slot still warming up β the blended pair is na and the blend does not draw for that bar. That na is expected behavior, not a bug.
The blend's color is driven by the same comparison the slots use: lime when blended RSI sits above blended signal, red when below. The tinted fill carries the same color at reduced opacity. A lime blend with a faded slot underneath it is a state worth noticing β the headline is up while one of the contributors has already started turning.
Quiet blends with loud slots
Three enabled slots at equal weight, one strongly leaning up, one roughly neutral, one strongly leaning down, can produce a blend that sits near 50 and barely moves. The blend is telling you the weighted average is roughly unchanged. That is true. It is also incomplete. The three slots are disagreeing, and the pane's slot lines will tell you so if you look at them β the slot scan was never optional.
A quiet blend with disagreeing slots is evidence about conflict, not evidence about calm. Trading off the blend alone in that configuration gives you neither the conflict signal nor the headline you thought you were trading; you are trading an artifact of averaging.
The opposite case also deserves naming. A blend that moves sharply while one slot is on an extreme and the other two are middling is telling you one slot is pulling the average hard. That is not a breadth read; it is a weight read. Cross-check the slot that is pulling and decide whether you believe that slot's evidence on its own before you let the blend stand in for it.
The reference lines β zones, not triggers
The 0 and 100 lines are the frame's rails. The 30 and 70 lines are reference zones carried over from textbook RSI reading conventions; they are drawn so a reader who already uses those conventions can keep them visible on this pane, and they are user-configurable so the reader can move them. Nothing in the tool's internal logic uses the 30 or 70 line as a trigger. No alert fires when a slot RSI or the blended RSI crosses either line. If a reader wants a TradingView alert that fires on a crossing of either line, that alert is the reader's to wire in TradingView's alert system, not a feature this tool ships.
Why the reference zones are still there, given they trigger nothing: they carry a reading convention that is useful in context. A slot RSI sitting above 70 is in a regime where the raw RSI (before smoothing) tends to be saturated on a trending move. That is an interpretation, not a triggering rule. Holding the reference alongside the color gives you two pieces of information on one line β where the value is in its range, and which direction it is leaning relative to its own smoothing.
What the reference lines are not: sell and buy triggers. The textbook RSI convention treats the 70/30 pair as reversal zones. On a bounded oscillator that smooths the RSI twice and then averages three smoothed RSIs on three timeframes, carrying that convention as a trading rule imports a semantics the tool does not implement. The rule survives as reading context; it does not survive as trigger logic.
The tinted fill and the blend signal β one read
The tinted fill between the blended RSI and the blended signal is the cheapest surface on the pane for seeing the blend's regime. Lime fill means the blended RSI is leading the blended signal; red fill means the blended RSI is lagging. The fill will widen when the blend's RSI and signal diverge and narrow when they converge.
A widening lime fill with the blend above 50 is a blend in full trend posture. A narrowing fill toward a color flip is the blend's own "directional change imminent" surface. As always, cross-check against the slot lines before you treat the blend's narrowing fill as an independent observation β the narrowing is the blend averaging slots that are themselves narrowing, and the slots will usually tell you why before the blend does.
The reading sequence in one screen
If you only take four moves away from this page, take these:
Scan slot colors. Full-tone or faded, one glance per slot.
Scan slot values. Intensity behind the colors.
Read the blend color and the fill. Headline for the pane.
Then look at the chart and make your read.
Read in that order every session. The order is the teaching. Once the habit installs, the pane stops competing with your RSI reflex and starts adding to it.
A note on when the order breaks down. Under time pressure β a fast session, a bar you feel you are already behind on, an alert that arrived inconveniently β your eye will skip to the value first because values are faster to parse than colors. That is the moment the reading order earns its keep. If you catch yourself reading values first, the correct move is not to trust the shortcut; it is to slow the scan down and restart at color. A session where you re-installed the order three or four times under pressure is a session where the order got stronger, not weaker. Treat the catch as the win, not as the failure.
Where to go next
The switch that determines what the slot reports on each chart bar, plus the slot-timeframe rule β MTF and Repainting.
What the ten alerts are gated on and what they do not confirm β Alerts.
The single largest over-trust risk on this trim, explained in order of damage β Limitations and Trust Boundaries.
The conceptual shape of the pipeline β For the Geeks.