ATAS Footprint Chart Configuration

A footprint chart is only as trustworthy as its row aggregation. This chapter walks through the settings that decide how trades are bucketed, how imbalance is computed, and how much of the result is stored versus redrawn. The order of operations matters: instrument and tick size, then ticks per row, then imbalance, then profile overlays, then color.

Anatomy of a Footprint Row

Each bar on a footprint chart is a stack of rows. A row represents a slice of the value axis, and each slice carries two totals: volume executed at the bid and volume executed at the ask. The renderer prints those totals side by side, with the row's total often shown at the edge. When a bar has more rows than the available height, the chart does not hide rows silently; it reduces the row height until the text no longer fits, at which point values are abbreviated. Abbreviation is a display rule, not a data rule, so the stored totals remain exact.

The row boundary is defined by the instrument tick size multiplied by the ticks-per-row setting. A chart with a 0.25 tick and four ticks per row produces rows one full point wide. Every trade is assigned to the row that contains its execution value, using the same rounding rule across the session. If you change the multiplier, the assignment changes for every historical bar, and any saved template that references row-relative thresholds will appear to behave differently.

Row Height, Ticks per Row, and Aggregation

There is no universally correct row height. The useful range depends on the instrument's tick size, the typical range of a session, and the resolution of your screen. A row that is too thin produces unreadable text and frequent abbreviation; a row that is too thick collapses distinct activity into a single line and hides the structure you opened the chart to see. Start with a row height that shows the full session range in roughly one and a half screens, then adjust ticks per row until the busiest bar has between eight and twenty visible rows.

Aggregation also affects the imbalance calculation. Imbalance compares the dominant side of a row against the opposing side of a reference row, usually one row above or below on the diagonal. Because the comparison is diagonal, changing row height changes which rows are compared, and therefore changes how many cells qualify. Two charts with identical thresholds but different row heights will not highlight the same cells. This is expected behavior and the most common source of "the imbalance marker disappeared" reports.

Imbalance and Diagonal Comparison

The imbalance threshold is expressed as a ratio. A setting of three means the dominant side must be at least three times the opposing side for the cell to qualify. Lower thresholds produce more marks and more visual noise; higher thresholds isolate only the most one-sided rows. A useful default for liquid instruments sits between two and four, adjusted after observing how many marks appear on a typical session. Thresholds are evaluated per row, not per bar, so a bar can contain several qualifying rows on each side.

Diagonal comparison adds a second condition: the dominant side is compared with the opposing side of the neighboring row. This is what distinguishes a stacked imbalance from an isolated one. Stacked imbalances, where consecutive rows qualify in the same direction, are often the more interesting pattern, and most templates include a separate color for a run of three or more. The run length is a display property; the underlying qualification still happens row by row.

Volume Profile Overlay Settings

A profile overlay can be anchored to the bar, to the session, or to a fixed visible range. Bar anchoring produces a small profile beside each bar and is useful for comparing shape across a swing. Session anchoring produces one profile for the whole session and is better for locating value. The overlay inherits the chart's row grid, so a profile drawn on a chart with coarse rows will look blocky even if the underlying data is fine-grained.

Value area percentage controls how much of the session volume is enclosed by the highlighted band. The conventional setting is seventy percent, but the number is a convention rather than a law. On thinly traded instruments, a lower percentage can produce a more stable band; on very liquid instruments, a higher percentage can be justified. Record the setting in the template name so a later reader knows which convention produced the chart.

Applying Templates Without Breaking History

Templates store appearance and aggregation, not data. Applying a template to an existing chart rebuilds the visible bars using the new settings, which can take a moment on a long history. The rebuild is deterministic: the same data and the same settings produce the same rows. If a rebuild appears to change totals rather than regroup them, check whether the template also changed the session or the bar type. Those two fields alter which trades are included, and that is a different operation from regrouping.

Keep one template per instrument class rather than one per instrument. A small set of well-understood templates is easier to audit than a large set of near-duplicates, and it makes cross-instrument comparison meaningful. When a template is retired, keep a note of why; the reason is usually more valuable than the file.

Common Misconfigurations

• Mismatched tick size. A chart imported from another instrument keeps the old tick size and produces rows that do not align with executions.
• Threshold set below one. Every row qualifies, which reads as a solid block of color and hides the pattern.
• Session mismatch. The chart includes extended hours while the profile does not, so the profile appears to miss volume.
• Overlapping overlays. Two profiles with different anchors are drawn in the same space and visually cancel each other.

Each of these is a settings problem with a settings answer. None requires a reinstall, a new data source, or a change of machine.