DLOG Binary STDF Editing & Data Recovery Guide
Header Fix · Bin/Verdict Correction · Bulk Edit · Quick Fix · File Repair

1. The Need for Direct STDF Editing & Safe Architecture

In high-volume semiconductor manufacturing, STDF (Standard Test Data Format) is the binary raw log streamed directly by ATE testers. Operator errors (such as mismatched lot IDs or test program names that cause MES/EDC upload rejections), prober orientation mismatches, and bin classification glitches are common headaches for test and yield engineers.

Historically, fixing these issues required dumping multi-hundred megabyte STDF files into ASCII ATDF, manually editing line-by-line, re-compiling, or writing fragile custom Python scripts. DLOG eliminates this friction by enabling engineers to open binary STDF V4 files directly, modify records inside an intuitive GUI, and save clean, spec-compliant binary files back to disk.

Tools & Export Menu Hierarchy

The Tools and Export menus located in the top-right corner serve as the centralized command center for data correction and output.

Click to zoom

Click to zoom

Command Selection Guide by Scenario

2. Edit STDF Info (Lot Headers & Record Metadata)

The central dialog for inspecting and modifying core STDF V4 metadata. Select any category from the left navigation pane to update the corresponding record fields.

Click to zoom

8 Core Metadata Categories

3. Edit Result (Fine Part & Value Editing)

Fine-tune individual die test results on a wafer sort map or packaged final test units.

Click to zoom

4. Bulk Edit (Conditional Batch Editing & Coordinate Shifts)

Filter thousands of parts based on defect criteria, site, or parametric test ranges to batch-update bins, mark deletions, or invert/rotate wafer coordinate grids.

Click to zoom

3-Step Batch Workflow

1. Define Target Conditions (Target parts):
   Enable criteria such as Soft BIN =, Hard BIN =, Result =, Site =, Item value (Low~High range with Inside/Outside modes), or Item is FAIL. Parts must satisfy all enabled conditions (AND). Check Whole scope to select all parts.
2. Preview Targets (View targets...):
   Click View targets... before applying to inspect the exact part count and list.
3. Select Action & Apply:
   Select the operation, configure parameters, and click Apply.

Supported Batch Actions

5. Quick Fix (One-Click Data Contradiction Resolution)

A frequent issue in ATE datalogs is internal contradictions — such as measured values within spec limits marked as FAIL, or parts assigned to PASS bins having FAIL verdict flags. DLOG's Quick Fix engine scans the entire lot, groups inconsistencies into actionable solution cards, and recommends the safest path forward.

Click to zoom

Detected Discrepancies and Resolution Cards

Only cards corresponding to detected contradictions are shown (4 in the example above). If no discrepancies exist, "No consistency issues found." is displayed.

6. Retests & Duplicate Parts Purge

In high-volume wafer probing and package testing, contact failures, handler re-insertions, and re-transmitted files often cause duplicate records for identical dies or package units. DLOG provides two dedicated purification tools: coordinate-based Delete Earlier Retests and identifier-based Delete Duplicate Parts.

6.1 Delete Earlier Retests (Coordinate-Based Retest Purge)

In wafer sort testing where identical die coordinates (X, Y) are probed multiple times, this command preserves only the final touchdown (Last Insertion) and permanently purges earlier 1st/2nd attempts from memory.

Workflow & Operations

1. Launch Command (Tools > Edit > Delete Earlier Retests):
   DLOG scans multi-touchdown die coordinates per wafer and prompts with the exact count of obsolete records to be removed.
2. Preserve Final Insertion:
   Confirming removes earlier records from memory while retaining the final test result.
3. Real-Time Yield & Map Sync:
   Yield summaries, Pareto charts, and wafer maps update immediately. Undoable anytime via Ctrl + Z.

6.2 Delete Duplicate Parts (Identifier-Based Duplicate Purge)

Identifies and removes duplicate parts sharing identical identifiers (PART_ID or PART_TXT) regardless of coordinate availability. Essential for package Final Test (FT) where coordinates do not exist, as well as wafer sort lines tracking global 2D barcode / ECID serial numbers.

Click to zoom

Click to zoom

3 Core Duplicate Rules

Default values are applied on first launch. Clicking Delete remembers these options for subsequent sessions (Cancel discards changes).

7. Save STDF & Anonymize STDF (Binary Serialization & IP Masking)

Re-serialize modified in-memory lots into compliant binary STDF V4 files or create masked copies to share with external foundries or vendors without leaking proprietary intellectual property.

Click to zoom

Click to zoom

Save STDF (Binary Output)

Writes modified part verdicts, updated headers, and adjusted coordinates into a fresh binary .stdf file. A pre-flight Issues audit dialog pops up immediately prior to writing to confirm data consistency.

• Single File Prerequisite: Only available when exactly one binary STDF file is open (disabled for ATDF text logs or multi-lot aggregated documents).
• Retest Mode Integration: When the top Retest selector is set to Collapsed (default), earlier superseded retest rows are omitted, keeping output files lean. Switch to All insertions if full historical probing runs must be preserved.

Anonymize STDF (Intellectual Property Protection)

Generates clean, sanitized datalogs for external OSATs, third-party failure analysis labs, or equipment vendors.

• Anonymization Scope: Replaces Lot ID, Sublot ID, Test Program Name (and Revision), Device/Part Type, Node Name, Operator, Test Item Name, and Bin Name (name only; Bin numbers are preserved) with safe, non-identifying aliases. Single-wafer files also have their Wafer ID replaced.
• Preserves Parametrics: Real measured voltage/current values, distribution statistics, wafer die coordinates, and Cpk curves remain 100% unaltered for valid external analytical modeling.
• Automatic Key Mapping: A reversible mapping table (CSV) cross-referencing original identifiers with their synthetic aliases is saved alongside the anonymized copy. Keep the mapping CSV securely internally and never distribute it externally.

8. Repair STDF (Truncated Stream Recovery Engine)

Recovers truncated or structurally corrupted STDF datalogs caused by abrupt tester crashes, BSODs, network dropouts, or filled disks.

Click to zoom

Core Repair Capabilities

• Headless Direct File Execution: Operates directly on disk files, allowing recovery of severely corrupted files that cannot be opened in standard viewers. (Triggered directly via the STDF Repair Needed dashboard notification pill)
• Bit-Level Record Boundary Re-synchronization: Follows record header structures (2-byte length + 1-byte type + 1-byte subtype) to re-establish valid record boundaries, isolating corrupt byte sequences and salvaging as many valid part records (PIR/PTR/PRR) as possible.
• Rebuild Missing Summary Records: When truncated file ends lose MRR, PCR, SBR, HBR, and TSR summary records, DLOG recalculates statistics from the salvaged parts and synthesizes compliant summary records adhering to STDF V4 specifications.
• Transparent Recovery Reporting: The completion dialog details salvaged parts count, retained vs. synthesized records count, skipped corrupt byte count, and key findings (INFO logs document regenerated headers and summaries complying with the specification). If large gaps were skipped, an explicit warning alerts you that some parts may be omitted.
• Self-Validation Guarantee: Immediately re-parses the repaired output; if validation fails, the damaged file is discarded rather than leaving a faulty datalog.

9. Split & Merge STDF Datalogs

Partition massive lot logs into manageable subsets or consolidate multiple test run fragments into a single unified STDF file.

Click to zoom

Click to zoom

Split STDF (Subsetting Datalogs)

Split large single-file datalogs into independent files:

• By Wafer: Generates one file per wafer (parts outside wafer records are saved to NOWAFER).
• By Site: Isolates parallel multi-site test channels (Site 1, 2, 3...) to diagnose site-to-site hardware offsets.
• By Head: Splits multi-head tester sessions into individual test head files.
• Every N parts: Splits by uniform part counts while keeping retested die attempts grouped in the same file.

Records that do not belong to a specific part are copied to all output files. If all parts belong to a single group, DLOG will notify you that there is nothing to split. Files where wafers from two heads are active simultaneously or files with non-part PRRs (such as test plan delimiters) cannot be split at this time.

Merge STDF (Consolidating Datalogs)

Rather than crude binary concatenation, DLOG retains only the primary headers (MIR, WCR, SDR, pin map) from the first file, copies all part byte streams verbatim across all files, and comprehensively recalculates PCR (part count), SBR/HBR (bin totals), and TSR (per-test min/max/sum) summary records from scratch. Select two or more STDF files in the file dialog and specify the output name; when explicit chronological ordering is required, use the CLI's merge command (which merges in exact specified order).

10. Convert Datalog & CLI Headless Automation

Integrate DLOG's editing and conversion capabilities into production line servers, CI/CD pipelines, and automated yield analysis scripts.

Click to zoom

Click to zoom

Industry Standard Format Conversions

dlog-cli.exe Scripting

The bundled dlog-cli.exe executable runs headlessly in command prompts, PowerShell scripts, and automated cron jobs:

• dlog-cli info <inputs...> [--json]: Instantly print Lot ID, yield, part counts, and bin summaries
• dlog-cli convert <inputs...> --format csv,sinf,g85,e142,json: Batch convert datalogs to target formats
• dlog-cli merge part1.stdf part2.stdf --out merged.stdf: Merge files in exact specified sequence
• dlog-cli split <file.stdf> --by wafer|site|head|chunk [--out <folder>]: Split by wafer, test site, test head, or part count (--parts N)
• dlog-cli repair broken.stdf: Recover corrupt files, generating broken_repaired.stdf alongside
• dlog-cli anonymize lot1.stdf --out D:\share: Generate sanitized clone (_anon.stdf) and mapping CSV
• dlog-cli report <inputs...> --format pdf,xlsx,docx,image: Headless automated report generation

Inputs support individual files, folders (processing all *.stdf, *.std, and *.atdf within), and wildcards like *.stdf. Existing output files are not overwritten without --overwrite (exits with code 4). CLI messages are always emitted in English regardless of system locale. For complete command options and exit code definitions, refer to Tools > Help > CLI Usage within the application.

11. Frequently Asked Questions (FAQ)

Q1. When I modify an STDF in DLOG, does it overwrite the original disk file?

No, it never automatically overwrites your original file. DLOG utilizes a strictly non-destructive in-memory architecture. All edits exist only in PC RAM until you explicitly choose Tools > Output > Save STDF and select a destination. If you attempt to close the document or open another file with unsaved modifications, a confirmation dialog appears first. Note that if you explicitly select the original filename when saving, it will overwrite the file, so save to a new filename or keep a backup of the original.

Q2. Will STDF files edited and saved in DLOG open correctly in third-party commercial tools (JMP, YieldHub, Galaxy, etc.)?

Yes, they generally open without issue because they are saved in standard STDF V4 format. DLOG strictly re-serializes record lengths, type/subtype codes, and pin mapping tables in compliance with the STDF V4 binary standard, and automatically re-reads the saved file to verify structural integrity. However, because proprietary parsing nuances exist across various commercial software, we recommend verifying critical data in your target analysis tool.

Q3. What is the difference between Quick Fix and Bulk Edit?

Quick Fix is an automated data inconsistency resolution tool that detects logical contradictions (e.g., measured values within limits marked as FAIL, or bin nature mismatches) and provides one-click standardized solutions.
Bulk Edit is a batch operational engine that allows engineers to filter parts by custom criteria (specific sites, measurement ranges) to batch-reclassify bins or shift and rotate wafer coordinate axes.

Q4. Can DLOG recover STDF files if the tester crashed mid-wafer?

Yes. Repair STDF can recover most truncated files. DLOG strips incomplete terminal byte sequences, rescues intact part records, and synthesizes missing PCR, SBR, HBR, and TSR summary records to produce a fully validated STDF file.

Q5. What should I keep in mind when merging STDF files?

Merged files must come from the same test program configuration and share identical TSR test item lists and byte order. Additionally, ensure that serial PART_ID schemes do not collide to avoid downstream records being misidentified as retests.

Q6. How does Delete Earlier Retests differ from Delete Duplicate Parts?

Delete Earlier Retests is a coordinate-based cleanup tool for wafer sort testing that purges earlier touchdown attempts at identical die coordinates (X, Y), keeping only the final insertion.
Delete Duplicate Parts is an identifier-based cleanup tool that groups parts sharing identical PART_ID (serial number/barcode) or PART_TXT strings and retains either the last or first part. It works across both wafer sort and final test package data without requiring die coordinates.

↑ Back to Top