ATAS vs TradingView: footprints, feeds and custom studies

What the visible chart does and does not prove

ATAS Cluster Statistics exposes bid volume, ask volume, and their difference as delta. TradingView displays categorized buy/sell volume, delta, and volume by price. Both can help inspect a candle, but the displayed labels alone do not specify how transactions were assigned to either side.

The classification method can change delta

TradingView documents classification by intrabar price direction: rising intrabars contribute buy volume and falling intrabars contribute sell volume. Flat intrabars use the previous close and, if necessary, the previous category. Historical calculations can use coarser intrabar intervals farther back. ATAS also needs a connection-specific check: its IBKR documentation says the feed does not transmit trade direction, so ATAS derives it from Bid/Ask and delta can change after refresh.

• Hypothetical example: one rising intrabar contains 100 contracts. A separate trade-side record assigns 60 contracts to ask-side executions and 40 to bid-side executions. Its trade-side delta is 60 - 40 = +20.
• Classifying all 100 contracts by that rising intrabar gives buy volume 100, sell volume 0, and delta +100. The totals agree at 100 while delta differs by 80. This illustrates two classification methods, not an observed ATAS or TradingView session.
• A finer interval may reduce aggregation differences but does not by itself prove that the classification matches an exchange aggressor flag. Ask which fields the connection actually delivers.

Match row size and the rule before comparing markers

Use a fixed price increment for a controlled comparison. For a hypothetical instrument with a 0.25 minimum tick, two ticks per row gives a 0.50 row width. Combining rows can hide opposing activity even when every trade is retained.

• The merged row has 160 contracts and zero net delta; it does not mean no trading occurred. Preserve the original rows when the research question concerns activity at a specific price.
• For an explicitly defined diagonal rule, compare buy volume at one row with sell volume at the row below. If these are 120 and 40, their ratio is 3. A rule using ratio >= 3 passes; ratio > 3 does not.
• Define zero handling. A denominator of zero cannot produce an ordinary finite ratio. Decide whether to skip that row or apply a separate rule, then verify the platform behavior.
• TradingView warns that realtime footprint calculations can differ from later historical recalculation. Save the alert-time configuration and values; a later screenshot is not sufficient to reconstruct the original signal.

Exchange access is separate from the platform label

ATAS documents a default delayed futures connection and separate broker or provider connections for live data. Its dxFeed connection has chart-type restrictions: the documented footprint support is for minute and second candles. TradingView also publishes exchange-specific data coverage. Check your actual entitlement and instrument on each service before treating two charts as a synchronized test.

• Record the exchange and exact contract month. A continuous futures series, an expiring contract, and a broker CFD are different inputs even when their names resemble each other.
• Request the fields your study needs: executed trades, trade direction, price-level history, or depth updates. Receiving last price does not establish that the other fields are available.
• Confirm live versus delayed status, professional or non-professional classification, historical coverage, and the connection permissions. Budget software and required market-data charges separately; this page does not quote prices.
• Check a recent completed interval and an older interval. Match timestamps, sessions, volume units, filters, and row settings. Investigate the first disagreement before judging a whole platform.

Custom studies need compatible inputs, not matching names

ATAS documents C# indicator access to candle price levels, online trades, aggregated trades, and market-depth updates. Pine now supports footprint data requests, including row data and volume statistics. TradingView lists Premium or Ultimate access for the footprint features in its announcement. These capabilities were checked on October 17, 2026; confirm current access before purchasing.

A small acceptance fixture for your own study

Before migrating a larger study, specify this hypothetical output: after a bar closes, mark it when total volume is at least 100 and delta divided by total volume is at least 20%. Use the categories supplied by the selected platform and label them in the output. This is a formula test, not a recommendation to trade the condition.

• Buy 60 and sell 40 gives total 100, delta 20, and 20 / 100 = 20%: mark the bar. Buy 59 and sell 41 gives 18%: no mark.
• Buy 6 and sell 4 also gives 20%, but total volume is only 10: no mark. With total zero or missing inputs, return no decision and avoid division.
• Require a completed bar. Compare the raw input totals, decision timestamp, and expected mark before comparing a strategy return.
• Give Pineify the explicit rule, input definitions, and these expected outputs to draft a Pine study and request syntax diagnostics. Confirm its behavior in TradingView; a successful syntax check does not prove data equivalence.