The Baltic's night alphabet
A light rhythm is a record in time. What identifies a signal is not only the flash but also the length of the darkness between flashes. Four teaching models on a single axis of seconds.
- light
- gap (darkness)
- start of the next period
- background lines: every 1 s
How to read it
The period is the whole repeating pattern, darkness included, not just the length of the flash. When it ends, the same pattern starts again from the beginning. A signal is distinguished by the proportion of light to darkness and by the number of flashes in a group.
| Notation | Sequence within one period | Period | Light / darkness | Share of light | What distinguishes it |
|---|---|---|---|---|---|
| Fl | light 1 s, darkness 5 s | 6 s | 1 s / 5 s | 17% | a short flash, clearly longer darkness |
| Oc | light 5 s, darkness 1 s | 6 s | 5 s / 1 s | 83% | more light than darkness; the signal is the short gap |
| Iso | light 3 s, darkness 3 s | 6 s | 3 s / 3 s | 50% | light and darkness equal |
| Fl(2) | light 1 s, darkness 2 s, light 1 s, darkness 6 s | 10 s | 2 s / 8 s | 20% | two flashes in a group; the longest darkness closes the group |
The descriptions of the Fl, Oc and Iso families and the Fl(2) group are summarised from the V2 material and the links in the Sources section. They were not re-checked against the originals in this version, so before quoting them they must be compared with the IALA definitions.
Overlay a record on a pattern
The observation record (cobalt) lies on the same time axis as the pattern (black). An observation usually starts at a random moment in the period, which is why you shift the start phase. The third track shows where the states diverge even after tolerance is taken into account.
Without scripts: four samples in text
- A. light 1.1 s, darkness 1.9 s, light 0.9 s, darkness 6.1 s (10.0 s in total). Deviations from Fl(2) 10 s of up to 0.1 s; the record covers a full period.
- B. light 1.0 s, darkness 3.2 s (4.2 s in total, shorter than the 6 s period). This is an incomplete excerpt: it fits the start of Fl 6 s, but also a stretch of Fl(2) after the second flash of the group, so it is inconclusive.
- C. light 1.2 s, darkness 4.8 s (6.0 s). This is not Oc (where the light lasts 5 s and the darkness 1 s); it is much closer to Fl.
- D. darkness 5.0 s, then light 1.0 s (6.0 s). This is Fl 6 s started 1.0 s after the start of the period: the same definition, a shifted phase.
The interactive overlay requires JavaScript to be enabled.
- pattern light
- observation light
- state differing from the pattern
- no data (not darkness)
- start of the pattern period
Loading
- Divergence outside tolerance
- —
- Coverage of the pattern period
- —
- Start phase
- —
The same record against all four models
Advanced: interval editor and manual recorder
A record can be at most 60 s long; a single interval 0.1–30 s. Interval by interval, you can correct your own measurement or a sample, and the overlay updates immediately.
No intervals. No record is not darkness, just no data.
Manual recorder
Hold the button down (mouse, finger, or Space/Enter on the button) while you see light, and release it in darkness. The first press starts the record with light. Reaction delay adds error to every change of state, which is why there is a tolerance. You do not need this to learn the reading, and the recorder uses neither sound nor flashing.
Not recording. Recording time: 0.0 s
Limits: fitting a model is not identifying a lighthouse
What “fits” means
- The record agrees with one of the four conventional models (Fl 6 s, Oc 6 s, Iso 6 s, Fl(2) 10 s). This is not a catalogue of lights.
- Real lights have other periods, groups, colours and sectors. Their parameters are given in the official list of lights (in Poland, according to V2, it is published by the Hydrographic Office of the Polish Navy; this needs checking in the current edition).
- This page is not a navigation aid and does not tell you what object you are seeing.
- The start phase changes only the moment the observation begins, not the definition of the pattern.
How agreement is calculated
- For every tenth of a second of the record we check whether the same state occurs in the pattern within the tolerance, and vice versa. This way a wide tolerance will not hide an entirely missing flash.
- Right at the start and end of the record (within the tolerance) the observation window is cut off and it is not known what came before and after the record, so locally the “pattern → record” direction is not checked there. When the record is at least one pattern period long, what is checked instead is whether each state of the pattern (light and darkness) occurs in the record at all. For that reason a record at least one period long that contains no flash at all (or no darkness at all) does not fit, even with a tolerance of ±1.0 s. A shorter excerpt without a flash, on the other hand, may fit as an incomplete window.
- A rule of the exercise, not a nautical regulation: a record shorter than the period is incomplete; only a record covering a period can honestly be called consistent over a full period; two periods show repeatability.
- The tolerance is a setting of the algorithm, not a norm of perception.
What changed relative to V2
- Removed the flashing lamp and the cycle playback. They were replaced by a static score on an explicit scale of seconds (a simpler way to learn the rhythm, without a stimulus).
- Removed the choice of a “hypothesis” and the verdict “correct refusal”. Instead, a single list compares the record with all four models, and incompleteness follows from the coverage of the period.
- Removed saving to localStorage: nothing needs to be remembered between visits, and the page stores no data.
- Removed the photograph of Rozewie (an external image, unrelated to the models) and the link to the withdrawn attempts.
- Kept the four samples A–D, the tolerance, the recorder, the interval editor, the phase shift and the distinction between “consistent with the model” and “consistent with the definition”. The Space key now acts on the focused button, not on the whole page.
- The page has no animation or sound, so reduced-motion mode hides nothing.
Sources
Links carried over from V2. Their content was not re-checked in this version, so this is material to verify, not confirmation. The times in the models are the author's own teaching schematic.
- IALA dictionary: Rhythmic Light — repeatability of the cycle (to be checked).
- IALA NAVGUIDE — definitions and classification of Fl, Oc and Iso lights (to be checked).
- IALA R0110, 5th edition — abbreviations and group structure, including Fl(2) (to be checked).
- Raport o polityce morskiej RP [Report on the maritime policy of the Republic of Poland] — status of the official lists of lights of the Polish Navy Hydrographic Office (to be checked).