The window is the operating surface for preparing a source master for broadcast delivery. It is divided into seven annotated regions: the application header, the text-overlay control, the Scopes launcher, the QC Report control, the target audio sliders, the delivery preset cascade, and the QC dashboard. The player viewport, transport bar, waveform, Deep Analysis control, and Convert control sit around them.

1. Application Header

• The app wordmark and the "QC INSPECTOR" badge identify the screen and its purpose: an inspection and delivery-conformance workspace.
• Establishes the dark, video-tool visual environment the whole window inherits.

2. Text Overlay Control

• Appears only once a text-rectangle analysis has produced something worth drawing.
• Selects whether nothing, a cumulative figure of all text found, or the moving per-frame text map is drawn over the players. Changing it changes what is visible on the picture without affecting the underlying analysis.

3. Scopes

• Opens the floating Broadcast Scopes panel, which shows the source and the converted MXF side by side: histogram, waveform, vectorscope, and a legal-range verdict against EBU R103-2000.
• Sampling runs only while the panel is open, so a closed panel costs nothing.

4. QC Report

• Writes or reopens the paginated PDF describing the file, the delivery it was judged against, every check, and every place a fault was found.
• Once a report exists it opens in the in-app preview instead of writing a duplicate; any edit that changes what the report was built from invalidates it and returns the control to export. Requires a license to be written.

5. Audio Settings Sliders

• Four numeric controls that fix the audio the encode will target:
• Channels — the delivered channel count, applied to the layout and to the number of mono tracks written.
• Loudness (LUFS) — the integrated loudness the normalizer aims at.
• True Peak (dBTP) — the ceiling the normalizer will not exceed, and the figure the source is graded against.
• Dynamic Range (LU) — the loudness range; never overwritten by a preset, because no published specification states one.
• Channel count, loudness, and true peak are overwritten when a delivery preset is applied; the operator can move any of them afterwards.

6. Delivery Preset Cascade

• Region, Country, and (only for territories that publish more than one specification) Delivery select the territory whose published standard the encode will conform to.
• Applying a preset pushes its channel count, integrated loudness, and true-peak figures into the sliders and sets the video standard the encode targets.
• The summary line beneath shows the resolved video standard, frame rate, GOP length, and loudness recommendation, plus any caveat from the source document. An amber warning triangle marks figures that are approximations rather than published requirements.

7. QC Dashboard — Source File

• A severity-ordered verdict on the loaded master against the currently selected delivery.
• The verdict banner at the top states whether the source is within specification or requires conversion.
• Pinned warnings above the grid carry the findings that change what the operator does next: a standards conversion, a duration that is not on a whole-second boundary, a head/tail mute shortfall, missing-media cards, black or frozen passages, or a surplus of audio channels.
• The tile grid reports each measured figure with a colour-coded pass, deviation, fail, or corrected badge. Failing and deviating tiles are sorted above passing ones. Tiles that only have something to say when they are wrong appear conditionally; tiles backed by a Deep Analysis pass appear only once that pass has been run.
• In Converted mode the tiles the conversion actually changes report the target values with a corrected badge; the diagnostic-only tiles carry their source verdict across unchanged because the encode does not touch what they measure.

Player Viewport and Transport

• The viewport shows the source. Once a conversion has produced a preview, it becomes two synchronised viewports — source on the left with a blue accent, converted MXF on the right with a green accent.
• Dropping a MOV, MP4, or MXF file onto the viewport loads it as the new source, replacing the current file and starting a fresh measurement pass.
• The transport bar carries play, stop, the source timecode (which is never shifted to the in-point, so positions always name the same frame as the operator's timeline), the trim control, and a replace action.
• The trim control shows the current head and tail amounts without being opened, turns red when the two figures together would leave under a second of programme, and refuses that trim outright rather than exporting something nobody asked for.
• The stacked waveforms below show the source lane on top in blue and the converted lane below in green, sharing a single playhead. Discarded head and tail are veiled and hatched across both lanes. Dragging anywhere seeks within the trim range; the converted lane is narrowed and offset when a trim is applied so both waveforms describe the same instant at the same position.

Deep Analysis and Convert

• Deep Analysis opens a popover offering three optional, expensive passes: text safe, PSE flash, and missing media. Each pass states what it measures, against what standard, and what it will cost before offering to run. A pass that came back clean is treated as spent on the current file. The capsule shows up to three verdict dots for the passes that have run.
• Convert writes the broadcast-compliant MXF alongside a modal progress sheet reporting the current pass, a percentage, and a labelled time-left estimate. It is unavailable while a conversion is running or when the trim is excessive. An unlicensed copy still converts, but the resulting picture carries a visible watermark; a clean master is never produced without a license.

Broadcast Scopes Analysis Interface

This interface provides a real-time visual analysis of video levels, comparing a source master file against a converted, broadcast-legal version.

Global Controls

• Live / Freeze Selector: Controls the operational mode of the analysis engine. Toggling this either samples the video frames continuously during playback ("Live") or halts updates to hold the current measurement on screen ("Frozen"), allowing detailed analysis of a single frame without stopping the playback clocks.
• Sampling Rate Configurator: Defines the refresh rate of the scope rendering (measured in frames per second, such as 10 fps). Adjusting this value scales the processing resource consumption; lower frequencies reduce CPU usage on high-resolution sources, while higher frequencies provide smoother, more responsive visual feedback.

---

Side-by-Side Analysis Columns

The interface is divided into two symmetrical columns to compare the original video (Source MOV) on the left with the conformed video (Converted MXF) on the right.

##### 1. Compliance Verdict Strip

• Legal Range Verdict: Displays a color-coded evaluation of the frame's compliance. An orange warning designates an "OUT OF RANGE" signal, indicating luma or chroma excursions beyond acceptable thresholds. A green indicator designates a "LEGAL" or "LEGAL • R103" status, confirming compliance with EBU R103-2000 standards (where up to 1% of samples are tolerated outside strict limits to account for codec ringing).
• Signal Range Readout: Identifies the absolute minimum and maximum luma values detected in the active frame, expressed in IRE units.
• Excursion Percentages: Reports the exact percentage of pixels falling below legal black (code 16), above legal white (code 235), or outside the legal chroma gamut (code 16–240).

##### 2. Histogram

• Color and Luma Distribution: Graphs the normalized pixel-population curves for red, green, blue, and luma channels across the 8-bit code value spectrum.
• Illegal Zone Boundaries: Features shaded red regions on the far-left and far-right margins. These highlight where signal details extend into forbidden super-black or super-white regions, visually demonstrating the effect of the legalizer's clamping action on the converted output.

##### 3. Waveform

• Luminance Plot: Traces luma intensity across the horizontal width of the frame, mapped against a standardized grid ranging from 0 to 100 IRE.
• Excursion Alarms: Highlights out-of-range luma peaks with contrasting warning dots (colored red or yellow) near the 100 IRE ceiling, indicating exactly where highlights require clamping.

##### 4. Vectorscope

• Chrominance Coordinate Plot: Displays hue and saturation levels on a circular grid derived from the Rec. 709 color space.
• Color Targets: Provides reference boxes for 75% saturated primary and secondary colors (Red, Magenta, Blue, Cyan, Green, Yellow) to monitor color accuracy.
• Skin Tone Reference Line: Features a diagonal dashed line representing the canonical skin-tone axis, ensuring human skin colors are not shifted during standards conversion.
• Chroma Gamut boundary: Indicated by a dashed square that marks the absolute safe chrominance limits, illustrating where out-of-gamut colors are constrained.

Compliance and Range Status

• Compliance Verdict: Warns when the active video frame contains samples that fall outside standard broadcast-legal thresholds, violating EBU R103-2000 specifications.
• Range Readout: Identifies the exact minimum and maximum luminance values present in the analyzed frame, expressed in IRE units.
• Excursion Percentages:
• Black Excursion: Measures the percentage of luminance samples falling below the legal black threshold (code 16).
• White Excursion: Monitors the percentage of luminance samples exceeding the legal white ceiling (code 235).
• Chroma Excursion: Displays the percentage of chrominance samples lying outside standard broadcast color gamut limits.

Histogram Analysis

• Signal Distribution Plots: Visualizes the relative volume of red, green, and blue color channels, alongside an overlaid white line plotting overall luminance across the 8-bit code value axis (0–255).
• Legal Boundary Guides: Vertical red lines mark the strict legal boundaries (code 16 for black, code 235 for white).
• Illegal Zone Shading: Shaded red background bands on the extreme left and right highlight forbidden footroom and headroom areas, revealing when signal peaks spill into prohibited regions.

Waveform Monitoring

• Luminance Mapping: Traces luminance intensity vertically across the horizontal width of the video frame, showing where bright and dark details are located.
• Graticule Reference Lines: Horizontal markers placed at 0, 20, 40, 60, 80, and 100 IRE provide reference thresholds for signal levels.
• Clipping Alarm Overlay: Red speckled patterns highlight the exact areas where the video signal clips or exceeds the 100 IRE maximum, pinpointing the horizontal position of overexposed highlights.

Defines how much material is removed from the beginning and end of the source before any measurement, deep analysis, or export. The QC dashboard and the delivered MXF describe only the retained programme.

• Entry — selects the notation used to state the trim: seconds and hundredths. Changing it rounds both amounts to what the new notation can express. The choice is shared with Preferences, so it acts as a house convention rather than a per-file setting.
• Head — amount removed from the start of the source. The delivered file begins after this point. Current value: 4.00 s.
• Tail — amount removed from the end of the source. The delivered file ends before this point. Current value: 6.00 s.
• Source and delivered duration — shows the original length 00:05:37:01 and the delivered length 00:05:27:01. The delivered programme is therefore 10 seconds shorter.
• In / Out timecodes — source-relative positions of the retained range: in at 00:00:04:00, out at 00:05:31:01. These positions stay tied to the source so a reported fault can be found in the original master.
• Delivery behaviour — the MXF is written from this range and starts at 00:00:00:00 regardless of the source’s own starting timecode. Timecodes shown elsewhere still refer to the source file.
• Remove Trim — clears both head and tail amounts and restores the full source to the delivery.
• Active trim summary — shows the current reduction as −4s / −6s while a trim is in force.

Optional pre-delivery checks that go beyond the signal measurements already on the QC dashboard. They inspect the picture of the loaded master for faults the conversion cannot fix — a missing-media card, burned-in text outside the safe margin, flash sequences over the photosensitive-epilepsy limit — and report them so the file can go back to the editing timeline. Each check is run by hand and each costs real time, so every one states what it looks for and what it will spend before offering to run.

Missing Media

• Looks for error cards the editing software burns in when a clip's source file was missing at export: the flat red Final Cut Pro fill, and the "MEDIA OFFLINE"-type text Premiere, Resolve and Media Composer draw. A card covering only part of the frame counts too, where the missing clip was one layer among finished ones.
• Reports the only fault of the three that makes a master undeliverable outright: the file can be perfectly legal in loudness, levels and frame rate and still carry a card in the middle of the programme.
• Samples two frames a second; the rate and the size the frames are read at are configurable, and both decide how small a card can still be recognised.
• Scan for Missing Media — starts the pass over the loaded file, restricted to the stretch the delivery will contain when a trim is set. Replaced by Scan Again once a result exists.
• Footnote: how many wordings the text half of the pass reads, how many languages they cover, and how many of them are confirmed against real installations versus unverified translations — an unverified wording produces no answer at all on a file that contains a card, so the distinction is stated rather than assumed.

Text Safe Area

• Finds every block of burned-in text in the picture and checks it against the 5% text-safe margin. Samples the file a few frames a second, so it runs in seconds rather than minutes.
• Analyze Text Safe Area — starts the pass over the loaded file. Replaced by Analyze Again once a result exists.

PSE Flash Warning

• Checks the file against the photosensitive-epilepsy limits of ITU-R BT.1702: no more than three flashes in any one second, counting a flash as a 20 IRE swing over more than 25% of the picture, plus a stricter parallel test on saturated red.
• Decodes every frame, so it takes longer than the text pass.
• Analyze Flash Rate — starts the pass over the loaded file. Replaced by Analyze Again once a result exists.

Run everything

• Perform Full Deep Analysis — starts every check that still has something to do on this file, as independent passes reporting their own progress. Its label counts what pressing it will actually start ("Run 2 Remaining Checks"), reads "Analysis Running…" while any check is in progress, and reads "Nothing Left to Analyze" once all three have run.
• A check already running is left alone rather than restarted, and a check that came back clean is treated as spent on this file: its own action greys out and the control stays inert until a different file is loaded.
• A check that produced findings offers a link into the list of passages, where each finding can be reviewed frame by frame or marked ignored so it stops counting against that verdict.

Defines where the broadcast-compliant MXF is written relative to the source master. The choice applies to every subsequent export and is persisted immediately.

• Same as source — the delivered MXF is written into the same folder as the dropped source file. No additional location needs to be supplied.
• Subfolder of source — the delivered MXF is written into a named subfolder inside the source file's own folder. The subfolder is created automatically if it does not exist; if it cannot be created, the export falls back to the source's folder.
• Broadcast MXF — name of the subfolder created beside the source. Editing it changes the folder name used for every future export.
• Folder — the delivered MXF is written into one fixed folder, chosen once and reused for every delivery regardless of where the source files live.

Impact from the operator's perspective: this setting determines only the destination path. Filenames, versioning on collision, and the source file itself are unaffected, and an output path resolving to the source file is never used — it falls back to a numbered name instead.

Defines the house convention applied to the filename of every exported MXF, so a delivered file can be identified by name without opening it. Both entries decorate the source clip's base name; the original name is always preserved between them.

• Prefix — text placed immediately before the source clip's base name. A value such as a station code or delivery tag makes exported files sort and filter together in the destination folder. Leaving it empty adds nothing ahead of the name.
• Suffix — text placed immediately after the source clip's base name and before any collision-avoidance numbering. Used for markers such as a version label, a date stamp, or a delivery qualifier. Leaving it empty adds nothing after the name.

The two settings are consulted when the destination path for the finished MXF is built; changing either affects only files exported from that point on and never renames anything already written.

Determines how the application resolves a situation where the computed delivery filename already exists at the destination folder. This is a house-wide policy choice that applies to every MXF export, deciding whether an earlier export of the same name is disposable or must be preserved alongside the new one. The setting affects only the naming of the written file and never the encoding, the trim, or the audio and video targets.

• No versioning (overwrite previous file) — An existing file at the computed destination is replaced by the new export. This keeps the delivery folder to a single file per source name. Its one safeguard is that the source master itself can never be overwritten: because the file importer accepts MXF, an output that would resolve to the input path falls back to a numbered name instead, so a conversion cannot truncate the file it is reading from.

• Add "_01", "_02", … on collision — A numbered suffix is appended to the base name and incremented until an unused name is found. This preserves earlier exports rather than discarding them, so a previous delivery remains on disk for comparison or re-sending. It is the setting that guarantees no prior export can be silently destroyed by a new conversion.

The Trim panel defines the default head and tail amounts proposed when a master is loaded, how those amounts are written and entered, and the fallback frame rate used before a file’s own rate is measured. These defaults are applied to every analysis pass and to the exported MXF; the per-file trim can still be adjusted after loading.

• Entry: Determines whether trim values are expressed as decimal seconds, whole seconds, or timecode. Changing it also rounds stored default trim figures onto what the new notation can express.
• Frame rate: Sets the reference rate used to count frames for timecode trim when no measured rate exists. Once a file is analysed, its own rate replaces this reference.
• Head: Sets the default number of seconds removed from the start of every newly loaded file. The per-file trim starts at this value and can then be adjusted.
• Tail: Sets the default number of seconds removed from the end of every newly loaded file.
• Applies to every pass, everywhere: Indicates these defaults affect all analysis and conversion operations, not just the final export.

Information affordances beside each item explain the behaviour in more detail.

Controls how the converted MXF is finalized during export.

• Clamp R'G'B' into gamut — Determines whether the encode routes the picture through a full-range R'G'B' round trip so that any reconstructed component lying outside the preferred component range is pulled back inside before the video is returned to Y'Cb/Cr. When enabled, saturated colours that reconstruct to values the scopes flag as out of gamut are clamped; because the round trip resamples chroma, in-gamut material can be altered slightly as well, which is why it is off by default and left to the operator's judgement. The setting is consulted at the moment an encode starts and its outcome is recorded against the finished delivery so the QC report can state which way it went.
• Information — Explains the trade-off behind the gamut clamp so the operator can decide whether the chroma resampling it introduces is acceptable for the master at hand.

Configures where an exported QC report PDF is written, what happens immediately after it is written, and how many findings are illustrated with a frame grab in the report.

• Destination — selects the folder the QC report PDF is written to. It is deliberately independent of where the converted MXF goes, because the report and the delivery usually travel to different people.
• Ask every time — presents a standard save panel on each export, proposing the source's folder and a name derived from the source file. Chosen by default so the first few exports show the operator what a report looks like before automation is enabled.
• Same as source — writes the report beside the source file without prompting. Currently selected.
• Subfolder of source — writes the report into a named subfolder inside the source's folder, created on demand if it does not exist.
• Folder — writes the report into a fixed folder chosen once and reused for every report.
• Naming — reveals an explanation of how the report file is named when it is written without a save panel, and that existing reports are never overwritten outside the save panel.
• After Export — selects what happens the moment the report has been written to disk.
• Ask every time — presents an alert offering to open the report or show it in the Finder.
• Open the report — opens the finished PDF in the app's own preview window, which can hand it on to the system's default PDF viewer.
• Show it in the Finder — selects the report in a Finder window so it can be dragged to its destination. Currently selected.
• Do nothing — writes the file silently and carries on, for a house whose downstream workflow already collects the report.
• Frames — controls how heavily the report is illustrated.
• Per check — the maximum number of findings from each check that are given a captured frame in the PDF (currently 30). Each capture is a seek and a decode, so the limit keeps a master carrying dozens of findings from turning the report into a near-identical picture sequence; findings beyond the limit are still listed with their timecodes in a compact table. Setting it to zero captures every finding.
• An explanatory note beside the value states the meaning of the figure and the effect of removing the limit.

This pane configures how the optional Text Safe deep-analysis pass finds burned-in text and checks it against the 5% text-safe margin. Changes apply to future Text Safe analyses and persist automatically.

• Detector: chooses the text-finding strategy. Thorough performs full text recognition, catching captions, disclaimers, and small or low-contrast text even over busy artwork, but is much slower. Fast locates text shapes only, runs significantly faster, but may miss captions and disclaimers over busy or low-contrast areas. Impact: thoroughness trades analysis speed for fewer missed text blocks.
• Rate: sets how many frames per second are sampled for text detection. 3 fps examines three frames per second; All examines every frame. Higher rates catch very brief text but increase analysis time; lower rates shorten the pass but may skip text that appears briefly. Because on-screen text usually remains legible for about a second, the default is normally sufficient.
• Width: sets the pixel width frames are reduced to before text detection. 1280 scales the picture to 1280 pixels wide; Native leaves the original frame size. Larger or native widths help detect smaller text but cost more processing; smaller widths speed analysis but may miss small or thin text.
• Fast vs. Thorough: explains the tradeoff between the two detector modes, including when the faster mode is adequate and when the slower, more accurate mode is needed.

Configures how the photosensitive-epilepsy (flash) check reads the picture when measuring flash rates against ITU‑R BT.1702.

• Width — the horizontal resolution each frame is reduced to before consecutive frames are compared. A lower figure analyses more frames per second at the cost of precision on small flashing areas; a higher figure resolves smaller flashes but slows the pass. The default of 320 keeps the pass fast because the check only needs the proportion of the picture that changed, which survives downscaling.
• How it's measured — reveals the methodology behind the flash verdict: what qualifies as a transition, the luminance swing and affected-area thresholds, the stricter saturated-red test, and the one-second window the flash count is evaluated over.

Controls how long a stretch of black picture or a frozen picture must last before the QC dashboard flags it as a continuity fault. Both are reporting thresholds applied to passages already recorded during analysis, so changing either immediately re-judges the file currently loaded instead of requiring it to be measured again. Head and tail black are treated as programme material and never counted as faults, and a freeze that is also black is reported once, as black, rather than twice.

• Black — sets the minimum duration of an interior black passage that is reported as a finding. Values are 0.2 s, 0.5 s, 1 s and 2 s; 0.5 s is the default, chosen so isolated black frames from decoding or compositing are ignored while a genuine hole in the middle of a cut sequence is reported. Raising it makes the check tolerant of longer deliberate black; lowering it makes it stricter.
• Freeze — sets the minimum duration of a held, unchanging picture that is reported as a finding. Values are 3 s, 5 s, 10 s and 15 s; 5 s is the default, deliberately above the length of packshots, held end titles and legal cards that are ordinary in a commercial, and below what a failed render produces. Raising it stops ordinary held frames from being reported; lowering it catches shorter stalls.
• How thresholds are judged — explains that these durations are reporting thresholds rather than detection limits: every black and frozen passage in the file is recorded during analysis regardless of these values, and the numbers here only decide which of those passages the dashboard and the QC report present as findings.

Configures how the missing-media pass scans a master for editor error cards burned into the picture, and which wordings it matches against. Findings feed the Offline Media verdict on the QC dashboard and the Violations panel.

Detection

• Rate — how many frames per second the pass measures. A card lasts as long as the clip that should have been there, so 2 fps finds anything an editor would have noticed missing; a higher count costs proportionally more decode time and is used when hunting something briefer than that.
• Width — the pixel width frames are scaled to before the text test reads them. This figure decides how small a card the recognizer can still resolve: 640 suits a feature scan while a smaller on-screen card needs 1280 or 1920, and Native leaves the picture at its own size. Raising it costs more per frame.
• How Rate and Width trade off — reveals an explanation of the two figures' cost and reach against each other.
• Skip visually busy frames — withholds frames that contain no flat area from the text test. Off by factory, because the flatness measure covers the whole frame and a card occupying only part of a photographed picture reads as busy and is therefore never examined. Worth turning on for long-form material whose cards fill the frame, where it is a genuine saving. Its adjacent note explains the trade-off.

Bundled Dictionary

• Reports the coverage the pass ships with: 23 wordings across 5 languages, of which 11 have been read off real installations and 12 are unverified translations that may not match what a given copy of the software writes. An unverified entry does not produce a wrong answer an operator can argue with — it produces silence on a file that contains a card — so the split is stated rather than assumed.

Your Wordings

• Add Tag — adds a wording the editor burns in locally, for a localisation, a house build or a version newer than the shipped list. Each entry is folded to the form the matcher compares, must be long enough to be meaningful, and is deduplicated against the bundled list.
• PLUGIN (removable) — one added wording, carried in the list; removing it stops the pass looking for it. Entries sit alongside the shipped dictionary and are marked as operator-added rather than app-supported.
• Matching rules — reveals how a wording is compared against what the recognizer reads, including case folding and whitespace collapsing.

• Keep panels above other apps — Sets the window level for Broadcast Scopes, Waveform, and Deep Analysis Violations. When enabled, these utility panels remain above the windows of other applications. When disabled, they behave as ordinary windows and can be obscured by other windows. The setting takes effect immediately for all three panels.
• What "on" does — Explains the behavior of the enabled state, including that the named utility panels are kept above other applications.
• Focus and sampling — Explains how the preference affects focus and sampling for the utility panels.
• When it applies — Explains when the preference takes effect, including that open panels adopt the new level without restarting the application.

The panel is a read-only quality-control record of a source master and its intended broadcast delivery. It states whether the source passes, what the conversion will correct, what still needs review, and which optional checks were not run.

• Report heading: identifies the source file, the generation timestamp, and the app version that produced the report.
• Source verdict: overall source compliance. A failing state lists the checks that failed or need attention, indicating the master is not deliverable as-is.
• Delivered verdict: expected compliance after conversion. It names the checks the conversion corrects and the checks still requiring manual review.
• Incomplete-check notice: states that the QC is incomplete because certain deep checks were not run, and clarifies that the verdicts above cover only the checks that were run.
• Source summary: metadata measured from the source file.
• File name: the loaded master being assessed.
• Folder: location of the source file.
• Video: codec, resolution, frame rate, and scan type.
• Duration: total running time.
• Start timecode: source timecode at the file head.
• Audio: track count and channel count.
• Analysed range: the portion of the source actually measured, when a trim is applied.
• Delivery target summary: the selected broadcast specification against which the source is judged.
• Delivery: territory or preset name.
• Video: target standard, frame rate, field order, and delivery format.
• Loudness: target integrated loudness, true-peak ceiling, loudness range, and the applicable loudness standard.
• Audio: target track configuration and sample format.
• Specification note: any caveat or exact published figure attached to the chosen delivery.
• Checks summary: each row compares the measured source value against the delivery target and shows the expected result after conversion. The checks are ordered worst first.
• Loudness: measured integrated loudness against the target; source failure is corrected by conversion.
• Frame rate: measured source rate against the target rate; conversion changes it, but the result still requires review.
• Luma range: measured minimum and maximum luma against the legal range; source warning is corrected by conversion.
• Offline media: missing-media card check; not run, so no verdict is available.
• PSE flash: photosensitive-epilepsy flash-rate check; not run, so no verdict is available.
• Text safe: burned-in text margin check; not run, so no verdict is available.
• True peak: measured true peak against the ceiling; passes before and after conversion.
• Chroma range: measured chroma range against the legal range; passes before and after conversion.
• Audio layout: measured channel configuration against the required track layout; passes before and after conversion.
• Black frames: interior black-passage check; passes before and after conversion.
• Freeze frames: frozen-picture check; passes before and after conversion.
• Footer: identifies the generating build and the page position within the report.

This page reports how the loaded source measures against the selected delivery specification and what the conversion changes. The upper table summarises duration, loudness range, and format. The lower section explains each evaluated check in detail. Status badges appear before and after an arrow: the left badge is the source verdict, the right badge is the projected or actual delivered-file verdict.

• Duration — PASS to PASS. The source and delivered length are exactly 00:00:05:00, on a whole-second boundary.
• Loudness Range — INFO to INFO. The measured range is 0.70 LU against a 7 LU target. This is informational and not judged pass/fail.
• Format — INFO to INFO. The source is 1440×2560 ProRes; the delivery target is 1080/50i XDCAM HD422. This is informational.

Check Details

• Loudness — FAIL to PASS after conversion. Found integrated loudness of −16.20 LUFS, 7.8 LU above the −24 LUFS AGCOM target. Conversion applies two-pass normalisation targeting −24 LUFS with true peak held at or below −1 dBTP. Action: none for compliance; consider rebalancing the source mix.
• Frame Rate — FAIL to WARNING after conversion. Found 24 fps source against 25 fps delivery, a different broadcast family, so the delivered file will not be a native master. Conversion uses motion-compensated standards conversion. Action: recut at 25 fps for the best result; review fast motion, dissolves, and graphics for interpolation artefacts.
• Luma Range — WARNING to PASS after conversion. Found luma reaching 13 and 252, below the legal floor of 16 and above the legal ceiling of 235; one passage leaves the range. The source is 10-bit, measured in its own depth, with figures shown on the 8-bit scale. Conversion clamps luma to 16–235 on every delivered frame. Action: none for compliance; for a graded result, bring levels into range in the grade.
• Offline Media — NOT RUN. The file has not been checked for missing-media screens burned into the picture. Action: run the pass from Deep Analysis and export the report again.
• PSE Flash — NOT RUN. The file has not been checked for flashing and saturated-red sequences that can trigger photosensitive epilepsy. Action: run the pass from Deep Analysis and export the report again.
• Text Safe — NOT RUN. The file has not been checked for burned-in text crossing the text-safe margin. Action: run the pass from Deep Analysis and export the report again.
• True Peak — PASS. The highest true peak is −1.73 dBTP, at or below the −1 dBTP ceiling. Conversion’s true-peak limiter holds the delivered file at or below −1 dBTP. No action needed.
• Chroma Range — PASS. Chroma stays inside the legal 16–240 range on every frame, with lowest 25 and highest 221. Conversion clamps chroma to 16–240 on every delivered frame. No action needed.

This page reports the QC outcome for the converted master. It lists each compliance check with its measured status, details the luma-range finding, documents the delivered MXF, and summarizes the corrections applied during conversion.

• Report header — Identifies the report and the source file it describes, with the generation date shown at the top right.
• Audio Layout — PASS. The source carries 2 channels in 1 track, exactly what the delivery takes. The conversion writes them as 2 mono PCM tracks, one channel each, in source order.
• Black Frames — PASS. No black passage of 0.5 s or more inside the cut.
• Freeze Frames — PASS. No frozen picture held for 5 s or longer.
• Duration — PASS. The delivered range runs 120 frames, exactly 5 seconds.
• Loudness Range — INFO. Measured loudness range is 0.70 LU; highest sample peak is −1.7 dBFS. Reported for information because no delivery in the catalog sets a limit on it. The conversion passes it to the normaliser as the loudness-range target of 7 LU, so programme dynamics are only compressed if they exceed that.
• Format — INFO. Source is ProRes 10-bit, 1440×2560, 24 fps, progressive. The delivery standard is 1920×1080 and the conversion does not rescale, so the file is delivered at 1440×2560. Re-encoded to XDCAM HD422 (MPEG-2 4:2:2 Profile @ High Level, 50 Mb/s CBR, GOP 12) in MXF OP1a, BT.709, upper field first. Timecode restarts at 00:00:00:00.
• Luma Range finding — One finding. Every frame’s lowest and highest luma code value is compared against the legal 16–235 range. Out-of-range frames less than 1 s apart are grouped into one passage; the picture shown is the frame that goes furthest out. All are clamped by the conversion.
• Warning badge: Luma below and above legal.
• Source timecode: 00:00:04:00 – 00:00:09:00.
• File position: 00:00:04:00 – 00:00:09:00.
• Measured range: 13–252.
• 120 frames out of range; worst at 00:00:05:09.
• Thumbnail shows the offending frame with a text-safe overlay.
• Delivered MXF — Documents the file produced by the conversion.
• File: output MXF filename.
• Folder: full destination path.
• Format: MXF OP1a · XDCAM HD422 · MPEG-2 422P@HL · 50 Mb/s CBR · GOP 12.
• Video: 1440×2560 · 1080/50i · 25 fps · upper field first · BT.709 · luma 16–235, chroma 16–240.
• Audio: 8 mono PCM tracks · 48 kHz 24-bit · −24 LUFS · ≤ −1 dBTP.
• Timecode: starts at 00:00:00:00.
• Written: timestamp when the MXF was completed.
• Applied by the Conversion — Lists the changes the conversion made to the source.
• Trimmed to 00:00:04:00 – 00:00:09:00 of the source (5.00 s); nothing outside that range is in the MXF.
• Loudness normalised from −16.20 LUFS to −24 LUFS integrated (two-pass), true peak limited to −1 dBTP (source peak −1.73 dBTP).
• Luma clamped to 16–235 (source 13–252, 1 passage out of range) and chroma to 16–240 (source 25–221, 0 passages out of range).
• R'G'B' gamut clamp applied: the picture was routed through full-range RGB and clamped to 0–100% before legalising.
• Standards-converted from 24 fps to 25 fps by motion-compensated interpolation.
• Page footer — Shows the app version and the page number within the report.

The page presents the closing portion of the QC report: a continuation bullet from the applied-corrections list and a NOTES section that defines how to interpret every timecode, level, loudness and delivery statement elsewhere in the report.

• Applied correction — “Audio: 2 source channels padded to 8 with digital silence and written as 8 mono tracks.” This records the channel-layout change the conversion performed. From the operator’s perspective it confirms that the delivered MXF’s audio track structure matches the target delivery’s channel count, and that silence was used to fill unused channels rather than duplicating or dropping source content.
• Timecodes note — Defines two position systems: “Source TC” (the source file’s own timecode, counted from its start timecode) and “File position” (frames from the first frame of the file, as the players show). Both count at the source frame rate; the delivered MXF restarts its timecode at zero from the trim’s in-point. This matters because the report’s timecodes must be compared against the NLE’s displayed source timecode, not the MXF’s reset timecode.
• Trim note — States that only the analysed range is measured and delivered; nothing outside it is judged. Positions remain source-relative so a finding can still be located in the original master. The functional impact is that the operator knows the report describes exactly the portion that will be delivered, and any note about faults outside that range is irrelevant.
• Levels and loudness note — Explains that all level figures are 8-bit code values (legal luma 16–235, legal chroma 16–240) scaled from the source’s 10-bit measurement. Loudness is ITU-R BS.1770 integrated loudness, true peak and loudness range (EBU Tech 3342), with every audio track of the source merged before measurement. This defines the units and the measurement basis so the operator does not misread a level or assume per-track loudness.
• Delivery section note — Clarifies that the Delivery section states what the conversion was set to do and did; the MXF itself was not re-measured for the report. The checks judge the source file, and the “After” column shows which checks the conversion corrects. This prevents the reader from treating the Delivery section as a post-encode measurement of the delivered file.