SINE
GUIDE

How to read SINE

No maths background needed. This page explains what SINE measures, what each number and chart means, and how that information can sharpen the way you read a market — plus where it stops being useful.

The idea in one minute

Play a chord on a piano and you hear one sound. But that sound is really a few notes played together, each with its own pitch and loudness. A Fourier transform is the maths that takes the combined sound and tells you exactly which notes are in it.

SINE does the same thing to a price chart. A token’s price is one wiggly line, but underneath there can be several repeating swings stacked on top of each other: a slow one over days, a faster one over hours, and a lot of random noise. SINE separates them and measures each one.

For every rhythm it finds, you get four things: how often it repeats, how big the swing is, where in the swing price is now, and how much of the price’s movement it explains.

the chart one line separated ↓ slow swing faster fastest
A real decomposition of a 7-day test series, computed in your browser.

What the numbers mean

The analyzer opens in Simple view, with a written summary at the top. Switch to Detailed for the full measurements. Any ? button explains a term on the spot.

RHYTHM STRENGTH
How regular the price has been, from No clear rhythm to Very strong. It’s based on how much of the movement the main rhythms explain, and is marked down if the rhythm has only repeated a couple of times. This is the first thing to check: if it’s weak, the rest is background detail.
MAIN RHYTHM · PERIOD
How long one full swing takes, high to high: “every 36 hours”. In Detailed view it’s also shown as a frequency in µHz, a unit that makes rhythms comparable across any timeframe.
AMPLITUDE
How big the swing is. ±4% means price has typically risen about 4% above its trend and fallen about 4% below it each cycle.
SHARE · VARIANCE EXPLAINED
How much of the price’s up-and-down movement a rhythm accounts for. 60% means most of the movement follows that rhythm; 10% means it’s barely there and news, trades and noise dominate.
RHYTHM VS. RANDOM WALK
How often 200 simulated random walks produced a rhythm this distinct. p < 0.05 means the rhythm is very unlikely to be chance; p above 0.3 means it probably is.
MEAN REVERSION
Whether price tends to pull back toward its trend (mean-reverting) or wander like a random walk. Pull-back is what makes cycles usable.
STRESS CHECK
How much of recent price action is slow, one-directional pressure rather than normal back-and-forth. Calm, Elevated or High. It flags pressure, not direction.
SIGNAL-TO-NOISE
Rhythmic movement compared with everything else. Above 1, the rhythms outweigh randomness. Below 1, randomness wins.
CYCLE POSITION
Where price sits in its main rhythm right now, moving forward in real time. Rising half or falling half, and how long until the point where past highs or lows formed — if the rhythm holds.
CORRELATION
From −1 to 1: how much a token moves with the market index once trends are removed. 0.7+ means it largely follows the market; near 0 means it does its own thing.
LEADS / LAGS
Whether a token’s swings tend to peak before the market’s (leads) or after (lags), and by how long.
BREADTH
How many of the top tokens rose vs. fell over 24 hours. 16 of 20 up is broad strength; 5 of 20 up is broad weakness.
TREND
The steady drift under the swings, per day. Rhythms ride on top of the trend; the trend is often the bigger force.

Reading each chart

Price swings vs. rhythm

Grey is the real price with its long-run drift and any slow regime shift removed, so only the swings are left. Green is the same swings rebuilt from just the three strongest rhythms. When green hugs grey, the price is behaving rhythmically. When grey wanders far from green, something else — news, a big buyer or seller, a new trend — is in charge. The pulsing dot is the live price.

Where we are in the main rhythm

One full cycle, trough to crest to trough, with a dot that moves in real time at the measured speed. It shows where past highs and lows formed in this rhythm. It is not a guarantee of where the next ones will be.

Rhythm heatmap (spectrogram)

Switch the main chart from Waves to Heatmap to see the same analysis the way an audio frequency analyzer shows sound. SINE runs the Fourier transform again and again on a window sliding through time. Time runs left to right. Cycle length runs up the side, slow at the bottom and fast at the top. The colour runs from dark purple for weak through red and orange to pale yellow for strong.

This shows when each rhythm was there, not just whether it was. How to read it:

  • A steady bright band is a rhythm that has held the whole time. It’s the most trustworthy kind.
  • A band that brightens toward the right is a rhythm that is emerging.
  • A band that fades toward the right is a rhythm that is dying out.
  • A band that drifts up or down is a rhythm speeding up or slowing down.
  • Scattered patches with no band are noise.

The dashed line marks the main rhythm. The written summary reports how consistently it showed up, and whether it has been getting stronger or fading lately.

If the pattern holds (projection)

Below the main chart, SINE takes the trend and rhythms it found and extends them forward, usually one full cycle ahead, starting from the last close. The dashed line shows where price would go if the recent pattern simply continued. The readout names the projected high and low and roughly when they would come.

Anyone can draw a line forward. What makes this one useful is the track record beside it. SINE goes back to up to 12 earlier points in the same chart. At each one it re-runs the whole method using only the data that existed then, projects forward, and checks what actually happened. The result tells you three things:

  • Direction right %: how often the projection pointed the right way.
  • Accuracy vs. “no change”: whether it beat the simplest possible guess, that the price just stays where it is. If it can’t beat that, it has no edge.
  • The shaded band: where price landed 8 times out of 10 relative to those past projections. It widens further out, because uncertainty grows with time.

The verdict under the chart sums it up. Worked reasonably well means a real, measured edge on this chart so far. No better than assuming the price stays put means ignore the line. The projection turns grey in that case. Even a good track record only describes the past: the moment the pattern breaks, and the heatmap will often show it fading first, the projection stops being useful.

Spectrum (Detailed view)

This is the raw output of the Fourier transform: every rhythm in the data, slow on the left and fast on the right. One tall bar standing out means a clear dominant rhythm. Many similar small bars mean noise with no clear rhythm.

Ecosystem pulse

The same analysis applied to the whole market: the top 20 Solana tokens combined into one index. The table shows how each token relates to it — its own rhythm, how closely it follows the market, and whether it tends to move first or later.

Is it just a random walk?

The efficient-market view says prices move like a random walk: each step is independent of the last, so past prices tell you nothing. The catch is that random walks naturally produce big, smooth swings that look like cycles. Any tool that finds rhythms has to prove it isn’t just seeing those chance swings.

SINE uses methods from academic research on this question. A 2025 study of 25 major cryptocurrencies (Öztürk, Istanbul Business Research) found that simple tests say most coins behave like random walks. But once the tests allow for slow regime shifts and non-linear behaviour, most coins, including Bitcoin and Solana, turn out to pull back toward their trend rather than wander randomly. SINE builds three of those ideas in:

1. Remove slow regime shifts first

Instead of a straight trend line, SINE fits a line plus one slow wave (the Fourier method of Enders & Lee), then removes it. A slow boom-and-bust across the window can no longer pose as a rhythm. The wave is limited to under 1.5 cycles per window, so it can never swallow a real, repeating rhythm.

2. Test rhythms against random walks

SINE runs 200 simulated random walks of the same length through the exact same analysis. A rhythm only counts as significant if it stands out from its neighbouring frequencies more sharply than chance produced in 95% of those simulations. Rhythms that could easily be chance are rated down or shown as “No clear rhythm”, even if they look convincing on the chart.

3. Check for pull-back (mean reversion)

The Fourier KSS test asks whether price tends to return toward its trend, which is what makes cycles usable, or drifts off like a random walk. You’ll find the result, and the other two, in the Detailed view and the “Random or not?” part of the summary.

Sources: Öztürk, C. (2025). Exploring market efficiency in cryptocurrencies: Fourier analysis of non-linear dynamics and breaks. Istanbul Business Research, 54(3), 374–389. doi:10.26650/ibr.2025.54.1666799 · Enders & Lee (2012), Oxford Bulletin of Economics and Statistics · Kapetanios, Shin & Snell (2003), Journal of Econometrics · Christopoulos & León-Ledesma (2010), Journal of International Money and Finance.

Stress check: when slow moves take over

Normally, price moves back and forth in no particular order: up a bit, down a bit, like random churn. Before big breaks, that can change. Moves start lining up in the same direction, and slow, persistent pressure takes over from quick noise.

A study of US, UK and German stock indices (Jun, Ahn, Kim & Kim, Physica A, 2019) measured exactly this with Fourier analysis. They slid a short window along the index’s returns and tracked each frequency over time. As global financial crises approached, the low-frequency components rose much more sharply than the high-frequency ones.

SINE’s Stress check applies the same idea to crypto candles. It measures what share of recent return “energy” sits in the slowest moves. Normal random churn puts about 25% there. To decide whether a reading is unusual, SINE compares it with the token’s own returns shuffled into random order. That keeps crypto’s large, fat-tailed moves but removes any real pattern in their timing, so a few huge candles alone won’t trigger it. The result is:

  • Calm: normal churn.
  • Elevated: above normal for this token.
  • High: unusual against shuffled returns and near the top of its own recent history.

What it can’t do: it flags pressure building, not which way it breaks, how far, or when. The original study used monthly stock-index data around a handful of crises. SINE’s version is an adaptation to much faster crypto data, and on purely random data it raises a false “Elevated” or “High” roughly 5–10% of the time. Treat it as a reason to look closer and size carefully, not as a signal on its own.

Source: Jun, D., Ahn, C., Kim, J. & Kim, G. (2019). Signal analysis of global financial crises using Fourier series. Physica A: Statistical Mechanics and its Applications, 526, 121015. doi:10.1016/j.physa.2019.04.251

Where the informational edge comes from

Charts are easy to over-read. SINE’s value is measuring things most people only eyeball, so you can tell a real pattern from an imagined one.

  1. 01 · SIGNAL OR NOISE

    Know when a pattern is real

    Everyone sees “cycles” on charts; most are noise. Rhythm strength and signal-to-noise tell you whether a repeating swing is statistically there — before you trade on it.

    LOOK FORStrong or very strong, repeated 6+ times in the window, SNR above 1.
  2. 02 · TIMING CONTEXT

    Don’t buy the top of a swing

    In a token with a strong rhythm, cycle position tells you whether price is early in a rise or near where past highs formed. It’s context for entries and exits, not a trigger.

    LOOK FORLate in the rising half with a strong rhythm = elevated risk of buying near a local high. The projection panel shows where the next high and low would fall — and whether that has worked here before.
  3. 03 · RIGHT TIMEFRAME

    Match your horizon to the market

    The main period is the market’s natural rhythm. Holding for a small fraction of it mostly captures noise; holding across several cycles means the trend matters more than the swings.

    LOOK FORA 36-hour rhythm suggests trades shorter than ~18 hours are fighting noise.
  4. 04 · WHO MOVES FIRST

    Leaders and followers

    Tokens that track the market closely but peak earlier can hint at where it’s heading. Closely correlated tokens that peak later may be slow to react to a move that has already started.

    LOOK FORCorrelation 0.5+ with a lead or lag of several hours, holding up across refreshes.
  5. 05 · REGIME CHANGE

    Notice when the pattern breaks

    When a rhythm that was strong suddenly weakens, or the main period jumps, the market’s behaviour has changed — often before it’s obvious on the price chart. The heatmap makes this visible at a glance. That’s a reason to reassess, not to trade the old pattern.

    LOOK FORA bright band fading or bending near the right edge of the heatmap, or the summary saying the rhythm is “fading lately”.
  6. 06 · MARKET OR TOKEN?

    Separate market mood from the token’s story

    High correlation means a token mostly rides the market; low correlation means its own news and holders drive it. Breadth tells you whether a market move is broad-based or carried by a few names.

    LOOK FORA token rising while breadth is weak and correlation is low is moving on its own story.
What SINE can’t do
  • Predict prices. Projections show what happens if a pattern continues; markets are under no obligation to repeat it, and a good track record can end at any time.
  • See news, listings, unlocks, hacks or whales. Any of these can override a rhythm instantly.
  • Be reliable on short windows. A rhythm seen only once or twice could easily be coincidence.
  • Handle thin markets well. In low-liquidity tokens a single trade can create a “pattern”.
  • Stay valid forever. Once enough traders act on a pattern, it tends to fade.
  • Replace your judgement. Nothing here is financial advice; always size positions for being wrong.

Try it on a real chart.

Open the analyzer in Simple view — the summary at the top explains everything it finds.

Read the Solana market Analyze a token