Clouds Unified
EMA13, EMA21, SMA50 and SMA200. The separate 100-day level remains independently toggleable.
Open saved cloud chart ↗TradingView first.
One home for the work.
Start with a saved chart. Workshop experiments live below, separately marked. Open only the view under discussion; this page keeps the rest without running background charts.
Updated 24 September 2026 · Clouds are closed for this review and remain unchanged on Station. “Source publication” links open documentation, not chart layouts.
Clouds & moving averages
Approved baseline. No new cloud styling or additional Station lines in this pass.
Full moving-average fan
The many-line view stays available for detailed review. It is not being added to every Station chart.
Open saved MA fan ↗Current-distance alternative
A compact distance ladder can show where averages are now without drawing their entire histories. The old “Distance to Moving Average” indicator has not yet been identified with certainty; possible matches are in Research below.
Oscillators · begin on TradingView
RSI timeframe fan + slow cloud
Two price charts with six locked-source RSI lines and a slow context cloud below each. Updated saved view, not an unapplied candidate.
Open MTF RSI view ↗RSI / Williams / Stochastic MTF
Three independent families on the right of price, each with its own slow cloud. Review these before the separate merged experiment.
Open independent MTF comparison ↗Compressed oscillator workbench
The preserved workbench reference for the left/right arrangement. Its current content has not been reverified in this pass.
Open original compressed workbench ↗What the combined preview shows
Indigo: RSI · Pink: Stochastic %K · Teal: Williams %R +100.
The preview starts with RSI and %K on, Williams off. “8H K” means Stochastic %K calculated on eight-hour candles. Overlapping translucent colors can look purple; purple is not a fourth oscillator.
Combined means curves overlaid on one 0–100 scale—not a blended score. %D is calculated but not drawn in this preview.
How the three calculations differ
| Indicator | What it measures | Our review settings |
|---|---|---|
| RSI | Smoothed gains relative to smoothed losses. | 14-period Wilder smoothing; 0–100. Recursive memory extends beyond 14 bars. |
| Williams %R | Close within the rolling highest-high / lowest-low range. | 14 bars; −100 to 0 natively. Add 100 only for the shared display scale. |
| Stochastic %K / %D | The same raw range position, then smoothed. | 14-bar range → SMA3 gives %K → another SMA3 gives %D. |
If the range is 90–110 and the close is 106, raw Stochastic is 80 and Williams is −20. Their displayed curves differ because %K is smoothed. RSI cannot be obtained from those three prices alone.
Fourteen 3H bars are nominally 42 bar-hours; fourteen 4H bars are 56. Session-truncated bars and market closures matter. These are comparison horizons, not universal quality thresholds or equivalent indicator memory.
Sources: TradingView RSI · Williams %R · Stochastic.
Diagonals & parallel channels
Trendoscope · continued diagonals
Upper-green / lower-red boundaries, duration-weighted rails and dotted continuations. Detection still uses the chart timeframe; this is not a locked multi-timeframe registry.
Open saved continued-diagonal chart ↗Source and version details
Read original Trendoscope source publication ↗ — indicator documentation, not our saved chart.
V3 visual and V4 screening candidates are locally compiled, not installed. Existing source licenses remain applicable. Continuations are removed when their parent pattern retires; no claim of unlimited history.
Parallel-channel blend
Two shared-slope families: Auto Parallel Channel confirmed/developing sets, and SCINTILLA APC upper-/lower-primary offsets. Four enabled lanes, not four independent detectors.
Open saved parallel-only chart ↗Independent diagonals and historical limits
Open saved independent-fit comparison ↗. Those methods have not been merged into Trendoscope.
Retention, calculation lookback and object limits are separate. Preserve emitted channels before retirement; do not stretch one present-day channel backward and call it historical detection. Increasing retained patterns may also alter overlap logic, so capture and replay must verify that behavior.
Comparing 1H, 3H and 4H without changing the viewing window
Use the same ticker, session, adjustment settings and common date window. Detect separately on each source timeframe, then display captured rails over the same candles and scale. Compare anchors, duration, slope in a common basis, boundary distance and confirmation time—not screenshot angles.
Fresh MCP check: BTC 1H / 3H / 4H were read from the same saved V2 chart and the chart was restored to 4H. Only 4 / 10 / 12 original rails had fully checked endpoints. Their durations were 19–26 / 54–255 / 92–316 hours. Unresolved anchors and unequal loaded windows prevent a fair comparative statistic; these are observed subsets, not pattern-frequency estimates.
There is no fixed “RSI14-equivalent” timeframe for a variable-length zigzag pattern. Measure actual durations across these three sources, then select useful coverage. Agreement across correlated timeframes is a hypothesis to test, not automatically independent evidence.
Support & resistance
Pivots, levels and diagonals
A reference, not an approved clean support/resistance preset. Its NQ 3H pane was read through the MCP on 24 September with extended hours enabled.
Open existing pivot / level chart ↗One visible zone, preserved evidence
Cluster nearby confirmed levels for display while retaining every original pivot, method, source timeframe and confirmation time. Count distinct market contacts—not multiple methods recognizing the same contact.
Store slope, boundaries, duration and touches explicitly where available. Missing confirmation or identity stays unknown, not inferred into certified statistics.
View templates · installed TradingView review
24 September update: these are saved charts with the studies installed, not empty layout proposals. Start with the independent families; use the merged experiment separately.
Two tickers + MTF RSI
Clouds, 100-day level and relative volume, with a locked-source RSI fan and slow RSI cloud beneath each ticker.
Open two-ticker MTF view ↗Price + three MTF oscillators
Price and relative volume left; indigo RSI, teal Williams and pink Stochastic in separate panes right. Each family has its own slow context cloud.
Open independent MTF review ↗MTF families + MACD / volume delta
Independent MTF oscillators left, price centre, MACD and cumulative volume delta right. MACD and CVD retain their existing native settings; they are not MTF fans.
Open full momentum desk ↗Business metrics around price
Quarterly revenue, net and gross margins left; market capitalization, P/E and quarterly debt/EBITDA right. Price, clouds and relative volume centre. Revenue includes TradingView’s native change annotations, not a separate growth-rate series.
Open fundamentals desk ↗Merged MTF families
All three families on one 0–100 display. Williams is shifted by +100 only; nothing is averaged into a score. RSI context cloud starts on; Williams and Stochastic clouds can be enabled separately.
Open merged MTF experiment ↗Locked sources, clearer clouds
Visible sources: 3H, 4H, 6H, 8H, 12H and daily. Optional 2H starts off. Each independent slow cloud spans the minimum and maximum native oscillator values from 2D, 3D, weekly and two-week sources—not moving averages. RSI fill is now24% opaque; Williams and Stochastic22%. Lines stay thin; native reference boundaries are clearer.
The three independent layouts now allow higher viewing timeframes. Lower-source history is sampled to the last available intrabar per viewing bar; developing live readings remain current. Source calculations stay locked, with extended-session intraday requests. A sampled-view notice explains the reduced historical detail. Provider intrabar-history limits still apply. The separate merged V3 experiment remains unchanged and retains its3H viewing limit.
Saved and visually checked, not a completed numerical parity or statistical validation. RSI signal-cloud expansion and MACD MTF remain proposals. Pattern geometry still follows its existing timeframe rules; no new MTF geometry claim.
Broader view progression
| Detail | Price surface | Additional information |
|---|---|---|
| Multi-ticker overview | Clouds everywhere | Keep price dominant; no automatic pattern overload. |
| Three tickers | Clouds + selected geometry | Compact oscillator pane below each chart. |
| One ticker · compact | Price above | Combined RSI / Williams / Stochastic below. |
| One ticker · detailed | Price left | Oscillator-only charts right, linked symbol/time/viewport. MACD and volume as explicit additional panes. |
| Deep study | One or two price charts | Separate oscillators and selected fundamental panes; shared navigation. |
Workshop experiments · not TradingView approval
For behavior TradingView cannot yet demonstrate. No workshop replaces the saved baseline.
Oscillator visual experiment
Combined curves share a 0–100 scale. Separate/full modes currently stack below price, not on the right. Full mode includes daily MACD and volume below it, which can sit below the fold.
Independent numerical TradingView parity remains open. No new Station integration.
Persistent-diagonal demonstration
Real extracted BTC 4H and daily rails from 24 September, shown across viewing timeframes. Not synthetic—but not a continuously refreshed feed or complete historical registry.
Open captured-geometry demonstration ↗Unresolved anchors are excluded. Pattern identities and confirmation times are incomplete, so this is not a performance backtest.
Cloud Workshop
The approved visual reference stays available. No new line or color changes here.
Open cloud reference ↗Research & evidence
Distance to Moving Average · possible matches, not identified yet
- Price Distance to its MA · DGT — source publication ↗
- Relative Distance to Moving Average — source publication ↗
Neither is confirmed as Alan’s old indicator. No new distance panel is installed.
Pattern candidates · source publications
Existing statistical handoff and prior checkpoint
The earlier Trendoscope packet has already been sent for analysis; no resend required.
Stored geometry / database brief · Markdown · Full prior checkpoint, test counts and provenance.
Volume, standalone pivots and Fibonacci remain separate discussions. Source pages are research, not evidence that a script is installed.