What are the Scanner, Pulse and Core-Layer surfaces?
Scanner, Pulse and Core-Layer are three views of the same LiquidityScan signals at increasing selectivity. Scanner is the raw per-engine detection feed. Pulse re-filters those signals to only ones with RSI confluence. Core-Layer folds them into multi-timeframe alignment chains. Only the Scanner detects anything new.
The key idea to hold onto: the scanner pulse core layer stack is not three separate detectors finding three different things. It is one detection layer (the scanners) plus two filters that sit on top of it.
Pulse and Core-Layer add no new pattern recognition of their own — they narrow and stack what the scanners already found. Getting that relationship straight is what makes the whole platform legible.
Why three surfaces exist
Every scan produces far more raw detections than any trader can act on. A single momentum candle on BTCUSDT might trip Super Engulfing on the 4H and CRT on the 1H at the same time, across 400+ pairs.
Breadth is useful for discovery, but it is noise if you are trying to find the two or three setups worth watching today.
The three surfaces answer that with three axes of filtering, each stricter than the last:
- Breadth (Scanner) — every closed-candle detection, unfiltered. You do the filtering.
- Depth (Pulse) — only detections that also pass an RSI momentum check on the same timeframe.
- Height (Core-Layer) — only detections that stack in the same direction across two or more timeframes.
Read left to right, selectivity increases and count drops. Read right to left, you trade a smaller list for more context per item. None of the three is "better" — they are entry points for different jobs, which is why understanding the scanner pulse core layer relationship matters more than memorizing any single engine.
Scanner: breadth, every detection
The Scanner surface is the ground floor: each engine surfaces its own pattern on confirmed, closed candles only, and the live forming bar is always dropped so nothing repaints. LiquidityScan runs a family of these engines, and each is its own feed:
- Super Engulfing (SE) — RUN / REV momentum-engulfing candles.
- CRT (Candle Range Theory) — a wick sweep that closes back inside the prior range.
- CISD — a change in state of delivery / market-structure shift with a resulting Fair Value Gap (FVG).
- ICT Bias — directional bias from the last closed candle breaking the prior range.
- 3-OB — a three-bar order-block break.
- RSI Divergence — price-vs-RSI divergence at an extreme.
- plus the zone/structure engines: OB+ / OB++, nested FVG+ / FVG++, Market Structure (BOS / CHoCH), Liquidity Sweep → Reversal, and Asia Range Sweep.
Each engine is deliberately narrow — it detects one thing and grades it (SE gives RUN/RUN+/REV/REV+; CRT grades STRONG/WEAK/FAILED; OB+ becomes OB++ on displacement ≥ 1.5× ATR). The Scanner does not rank engines against each other or decide which detection deserves your attention.
That is the point: breadth first, judgment yours. Use the Scanner when you want to see everything happening on a pair or across the market, then apply your own read.
One honest limit lives here and propagates upward: most engines surface context, not a trade. Only Super Engulfing outputs a geometric entry/SL/TP ladder, and even those targets are ratios (1:1.5 / 1:2 / 1:3), not predictions. Everything else — CRT, CISD, OB+, FVG, Bias — renders a level or a zone for you to evaluate.
Breadth does not mean noise floor. Two guards keep the raw feed clean before you filter it. Signals are derived only from confirmed, closed candles, so nothing on the Scanner repaints once a bar prints.
And the signal lists hard-filter any crypto pair trading under $20M of 24-hour quote volume, so illiquid names never clutter the feed (TradFi symbols are exempt from that floor). What reaches the Scanner is already liquid, stable detection — the filtering Pulse and Core-Layer add is about relevance and conviction, not cleaning up junk.
Pulse: depth, only the momentum-backed detections
Pulse is a quality filter, not a new pattern. It takes existing base-scanner signals — SE, CISD, CISD+IFVG, CRT, ICT Bias and 3-OB — and passes through only the ones that also carry RSI(14) confluence on the same timeframe.
If a signal has no matching RSI condition, Pulse drops it. Nothing is detected here that the scanners did not already find; Pulse just refuses to show the momentum-empty ones.
It exposes two strictness modes side by side:
- Standard — RSI recently touched the oversold/overbought zone without crossing back through the midline. The momentum condition was met near the signal, not necessarily at the exact close.
- ★ Strict — RSI is inside the zone at the exact signal close. Strict is a subset of Standard: everything Strict shows also qualifies as Standard, but not vice versa.
Two more constraints tighten the feed. Pulse is direction-matched — an oversold RSI reading only qualifies long signals, overbought only shorts — so you never see a bearish setup "confirmed" by bullish momentum. And an optional Kill Zone toggle restricts results further to ICT time-of-day windows, if you only trade specific sessions.
The practical effect: Pulse narrows a broad Scanner feed to the detections where price structure and momentum agree. It is the surface to open when you have too many raw signals and want to keep only the ones with a second, independent reason to care.
Core-Layer: height, alignment across timeframes
Core-Layer is a composite, and like Pulse it detects nothing new. It takes live SE, CRT and ICT Bias signals and folds them into multi-timeframe alignment chains — stacks where the same direction lines up across timeframes.
A Weekly bullish SE, a Daily bullish CRT and a 4H bullish Bias on the same pair become one aligned stack rather than three loose signals.
The mechanics matter for trust:
- It groups live signals by symbol and direction, and keeps one per timeframe.
- It requires at least two aligned timeframes — a lone signal is not a chain.
- It prunes temporally incoherent stacks: a stale weekly signal cannot stack with a fresh daily, because they describe different moments.
- It classifies an anchor — Weekly, Daily or 4H — so you know the highest timeframe the alignment rests on.
This is the ICT idea of top-down analysis made mechanical: higher-timeframe direction should agree before you act on a lower-timeframe trigger. Core-Layer is the surface to open when you already believe in a direction and want to confirm the timeframes are stacked behind it.
Because it operates on the same underlying signals, an alignment chain is only ever as good as the base detections inside it — it concentrates conviction, it does not manufacture it.
The alignment chains build over the W, 1D, 4H and 1H timeframes, extending down to 15m and 5m via the sub-hour engines, so the anchor you read tells you exactly how high the agreement reaches.
A 4H-anchored chain and a Weekly-anchored chain both clear the two-timeframe minimum, but they are not the same claim: the Weekly-anchored stack says the higher-timeframe draw agrees, while the 4H one only says the intraday timeframes do. Reading the anchor is how you avoid mistaking a shallow stack for a deep one.
How the three surfaces nest: a worked example
Because Pulse and Core-Layer both sit on the scanner output, a single signal can appear on all three, gaining context at each step. Walk one through with an illustrative BTCUSDT sequence:
- Scanner. The 4H candle closes as a bullish Super Engulfing REV+ after sweeping a prior low near 60,000 and closing back above it. It appears raw in the SE feed, alongside dozens of other detections marketwide. On its own it is a candidate, nothing more.
- Pulse. On the same 4H, RSI(14) had pushed below 30 into oversold and has not crossed back through the midline. The signal survives into Pulse (Standard). Had RSI still been sitting under 30 at the exact 4H close, it would also flag ★ Strict. Because the setup is long and the RSI reading is oversold, the direction matches and it passes. Dozens of momentum-empty detections did not.
- Core-Layer. The Daily already shows a bullish ICT Bias and the Weekly a bullish CRT, all fresh and all pointing up. Core-Layer folds the 4H SE into that stack: three aligned timeframes, Weekly anchor. The same signal is now the lowest-timeframe trigger inside a coherent HTF alignment chain.
Nothing about the candle changed between step 1 and step 3. What changed is how much agreement surrounds it — raw detection, then momentum confluence, then multi-timeframe alignment. That is the entire scanner pulse core layer logic in one setup: increasing conviction at each surface, from a single filter you apply yourself to two the platform applies for you.
Which surface to use when
Pick the surface by the job, not by which sounds strictest. Discovery wants breadth; conviction wants height.
| Surface | What it does | When to use it |
|---|---|---|
| Scanner | Raw per-engine detection on closed candles. Every pattern, every grade, unfiltered. | Discovery and breadth — you want to see everything and apply your own read. |
| Pulse | Filters base signals to those with same-TF RSI confluence. Standard vs ★ Strict, direction-matched, optional Kill Zone. | Narrowing a busy feed to momentum-backed setups only. |
| Core-Layer | Folds live SE/CRT/Bias into multi-TF alignment chains (≥2 TFs, W/D/4H anchor). | Confirming a direction is stacked across higher timeframes before acting. |
A natural workflow chains them: discover on the Scanner, filter to momentum on Pulse, then check whether the survivor sits inside a Core-Layer chain. But you can also enter at any surface — open Core-Layer first if you already have a directional thesis, or live in Pulse if you only ever trade momentum-confluent setups.
One caveat applies across all three. They do the same honest job at different resolutions: they surface context and candidates, not trade calls. None of them publishes a win rate, and the platform does not compute one — the WIN/LOSS labels some engines carry are internal lifecycle states, not performance statistics.
Pulse passing a signal is not a prediction that it works; a Core-Layer chain aligning is not a guarantee the move follows. They raise the amount of agreement behind a candidate, which is different from raising its probability to a number.
LiquidityScan is analysis tooling, not a signal service or financial advice, and the scanner, Pulse and Core-Layer surfaces are best read as three ways to organize your own judgment rather than replace it.
Frequently Asked Questions
Does Pulse detect patterns the Scanner misses?
No. Pulse detects nothing on its own. It only re-filters signals the base scanners already produced — SE, CISD, CISD+IFVG, CRT, ICT Bias and 3-OB — keeping the ones that also show RSI(14) confluence on the same timeframe. Every Pulse signal exists first as a raw Scanner signal.
What is the difference between Pulse Standard and Strict?
Both require RSI confluence with the signal's direction. Standard means RSI recently touched the oversold/overbought zone without crossing back through the midline. Strict means RSI is inside that zone at the exact signal close. Strict is a stricter subset — every Strict result also qualifies as Standard.
How many timeframes must align for Core-Layer?
At least two. Core-Layer keeps one signal per timeframe, groups by symbol and direction, prunes temporally incoherent stacks, and classifies an anchor of Weekly, Daily or 4H. A single-timeframe signal is not an alignment chain, so it will not appear as a Core-Layer stack.
Can the same signal appear on all three surfaces?
Yes. A closed-candle detection appears raw on its Scanner feed, survives into Pulse if it has matching RSI confluence, and stacks into Core-Layer if the same direction aligns across two or more timeframes. It is one signal gaining more context, not three separate detections.
Related query paths
To go deeper on how detection, filtering and alignment work under the hood, follow these in order.
- What is LiquidityScan? — the platform overview that frames all three surfaces.
- How the LiquidityScan Scanner Works — raw candles to a detected setup, the Scanner layer in detail.
- Order Block Scanner: Real-Time OB & FVG — how the zone engines feed the Scanner surface.
- ICT Top-Down Analysis: Multi-Timeframe Alignment — the theory Core-Layer mechanizes.
- Is LiquidityScan Reliable? — how detection is validated and why signals don't repaint.
- Best ICT Scanner: How to Choose — where a multi-surface scanner fits among the alternatives.
- How to Build a Custom ICT Scanner With No Code — a related angle on custom scanner builder.