Troubleshooting

Use this page when the chart is behaving in a way that feels wrong, confusing, or harder to trust than it should. The goal is not to memorize edge cases. The goal is to move from symptom to likely cause quickly, then verify the fix before you keep going.

Fast symptom table

Symptom: the indicator errors immediately on load

Likely cause

At least one enabled slot timeframe is below the current chart timeframe. The default stack starts at `5 / 15 / 60`, which is fine on a one-minute or five-minute chart, but not on every chart.

What to do

1. Open settings immediately.
2. Check `TimeFrame:` for each enabled slot.
3. Raise or disable any slot that sits below the chart timeframe.
4. Confirm the chart now loads cleanly before changing anything else.

Verify the fix

You should see the chart draw the intended envelopes without the runtime error returning.

Symptom: I hid a slot, but it still seems to matter

Likely cause

`Hide BB 0X Plot` only removes the drawing. It does not disable the slot.

What to do

• If you want the slot gone from logic, turn off `Enable BB 0X`.
• If you only wanted less clutter, keep it hidden and accept that it still matters internally.

Verify the fix

Toggle `Enable` off and confirm whether the blend or alignment logic changes. That tells you the slot was active, not merely visible.

Symptom: the blend looks meaningless or flat after weight changes

Likely cause

One of 3 things usually happened:

• you zero-weighted the contributor you were really relying on
• you assumed hidden slots stopped contributing
• all active slot weights are now `0`

What to do

1. List the enabled slots.
2. Check which ones still have non-zero `Blended Weight:`.
3. Decide whether the blend still has any meaningful contributors.

Verify the fix

Move one contributor from `0` back to a non-zero weight and confirm the blend becomes informative again.

Symptom: the alternate-ticker slot looks strange

Likely cause

The slot is being translated into the chart's price neighborhood. It is not supposed to look like raw price from the other market.

What to do

• Put that slot at `0` blend weight first.
• Open a second chart with the source market.
• Compare the broad structure, not the literal price level.

Verify the fix

You should be able to say what that outside market is adding to the read without pretending the prices are directly interchangeable.

Symptom: alerts keep firing when I only wanted the moment of change

Likely cause

You probably chose an ongoing state alert such as `Is Above Basis` or `Is Below Basis` when you really wanted a change alert.

What to do

• Use `Basis Change` for slot or blended basis flips.
• Use the blended cross alerts when you want the moment price crosses a blended boundary.

Verify the fix

Once the new alert is set, check whether it now arrives only when the condition changes instead of while the condition simply remains true.

Symptom: alerts wait for the close even though I turned confirmation off

Likely cause

These are 2 different timing layers:

• `On Bar Close?` changes whether the stack reads confirmed or live-forming higher-timeframe values
• alerts still wait for the chart bar to confirm

What to do

Keep those 2 ideas separate. If you need earlier workflow feedback, decide whether live-forming stack behavior actually helps you. Do not expect the alert gate itself to change.

Verify the fix

Watch the same chart live and confirm: the slot values can move earlier with `On Bar Close?` off and the alert condition still waits for the chart bar to close.

Symptom: changing ALMA on one slot affected another slot

Likely cause

The ALMA controls are global in this indicator.

What to do

• If you need 2 different ALMA tuning profiles at the same time, this build will not do that inside one script instance.
• Either simplify the stack or use a different workflow design.

Verify the fix

Set 2 slots to `ALMA`, change `ALMA Offset:` once, and confirm both slots reflect the change.

When simplifying is the real fix

If several symptoms are showing up at once, the best fix is often not another adjustment. It is a simpler stack. Return to: fewer active slots, confirmed mode, same-symbol context only, one basis type across the stack. That version is easier to verify, and it usually reveals whether the extra complexity was helping or just adding fog.