articles

API: /v1/records?type=article

• How is SMPTE LTC structured per libltc documentation and code, and where do 23.976, 29.97 and 59.94 behave differently? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  libltc encodes LTC as 80-bit frames (2400 bit/s at 30 fps) with drop-frame at bit 10 and colour-frame at bit 11, and sets the drop-frame bit only when the frame rate rounds to 29.97 (rint(fps*100)==2997, in src/ltc.c); so 23.976 material is labelled as non-drop 24-frame counts (code reading).
• How do iXML SPEED tags express NTSC frame rates and pulled-up recordings (for example 48048 Hz labelled as 48000), and which field should a conform tool trust when they disagree? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  iXML SPEED gives rates as integer ratios (e.g. 30000/1001 rather than 29.97), separates timecode display rate and flag, and adds file, digitiser and timestamp sample rates so pulled-up recordings (48048 Hz labelled 48000) can be identified; readers must take the authoritative rate and timestamp from the fmt and bext chunks, not iXML copies.
• What does the BWF bext TimeReference field actually store, and what does FFmpeg's WAV code do with it when I copy, re-encode or resample a recorder file? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  A BWF bext TimeReference is a 64-bit count of samples since midnight at the file's own sample rate, with no frame-rate field; FFmpeg's WAV demuxer (4bef4614) reports that raw count as time_reference metadata without converting it and does not parse iXML, so tools must divide by the file sample rate (and know the project rate) to get timecode.
• What does an external lens calibration file for FreeD look like, and how does it compare with Unreal's Lens File model? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Ventuz's lens file keys OpenCV calibration points by the FreeD zoom value and allows up to 14 OpenCV distortion coefficients (k1-k6, s1-s4, Tx, Ty; missing ones taken as 0), while Unreal's Lens File keys five Brown-Conrady parameters by focus and zoom; the files describe different models and cannot be swapped directly (conversion would need refitting or truncation).
• What does an Unreal Engine Lens File store, what inputs does it need to be calibrated, and how is it evaluated at runtime? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Unreal's Lens File stores five Brown-Conrady distortion parameters (K1, K2, P1, P2, K3), camera intrinsics and nodal point offsets keyed by focus and zoom, interpolated between calibrated points; Epic says to make a new Lens File for each lens and camera body combination.
• How does Unreal nDisplay layer colour grading and OCIO for LED panels and the inner frustum, and what does the SMPTE RIS colour paper say it covers? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  In nDisplay, colour grading settings stack additively in order (a viewport red and a cluster blue give purple), while OCIO overrides are not additive; for in-camera VFX each camera and LED panel combination typically needs its own destination colour space.
• What do public docs say about synchronising an LED volume's render nodes, cameras and trackers, and how is tracker latency/offset represented? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Epic says genlock plus framelock is the most desirable nDisplay setup for LED virtual production, genlock alone is not fully reliable, and without hardware sync expect at least a one-frame delay; SMPTE's camdkit tracking metadata records the synchronisation source (genlock, video in, PTP, NTP) so a tracker's timing can be checked later.
• Which language-code system applies where: DCNC title-string codes versus CPL metadata tags, and what extra rules does ISDCF Doc 7 add? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Two layers carry language in a DCP: DCNC codes in the title (case carries subtitle-rendering meaning) and BCP 47 tags in CPL metadata, which ISDCF Doc 7 says must be treated case-insensitively; tools converting between them must not copy case semantics across, must keep region subtags (es vs es-419) and should use RFC 4647 matching for playback selection.
• How do I read the audio-language and subtitle-language parts of an ISDCF digital cinema package name, and where do the same values sit in the CPL? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  In ISDCF digital cinema naming the language field lists the audio language first and subtitle languages after it; upper case means the projector renders the subtitles and lower case means they are burned in; the same languages are mirrored in the CPL (MainSound and MainSubtitle Language, MainSubtitleLanguageList).
• How can an archive verify that an FFV1/Matroska transcode is lossless, and what do MediaConch policies and FADGI say? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  To verify a lossless transcode, compare decoded-frame hashes: FFmpeg's framemd5 gives per-packet MD5, while its crc and framecrc muxers use Adler-32 (not CRC-32) on frames converted to raw; MediaConch's sample digitization policy checks Matroska v4+, a unique ID, FFV1 3.4+, GOP size 1 and slice CRCs, plus SD geometry.
• What fixity and error-detection does FFV1 (RFC 9043) and Matroska (RFC 9559) build in, and how do documented tools use it? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  FFV1 (RFC 9043) supports frame fixity: ec=0 puts a 32-bit CRC only in the configuration record, ec=1 also in every slice, using the IEEE polynomial 0x104C11DB7 with initial value 0 and no pre- or post-inversion; so per-slice damage is detectable only when ec=1.
• How does Perfetto's FrameTimeline tell me whether a late frame on Android was the app's fault or the compositor's? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Perfetto FrameTimeline (Android 12+) flags a frame as janky when its actual present time differs from the scheduler's predicted time, shows expected and actual timelines per app and SurfaceFlinger, and classifies jank (app deadline missed, buffer stuffing, SurfaceFlinger, prediction error) so app-caused and compositor-caused jank can be separated; SurfaceViews are not supported.
• After a Yarn Spinner for Unity project has translated CSVs, what happens to the translations when I edit, add, delete or reorder lines in the .yarn file? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  YarnSpinner-Unity keys translated lines by line tag and stores a Lock (first 8 characters of the SHA-256 of the base text); when the inspector update runs, changed base lines get '(NEEDS UPDATE)' prefixed to the translation, new lines arrive empty and removed lines are dropped. Editing base text without the update leaves stale translations unflagged.
• Which continuity and report documents do script-supervision job descriptions and training outlines say go to editorial, production, camera and VFX, and where do they disagree? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Script supervisor paperwork serves two audiences: editorial (marked-up/lined script pages, editor's daily log, facing pages, selects or print log, VFX notes) and production (timings and page counts for progress reporting). ScreenSkills assigns these to the supervisor; IATSE Local 634 gives the editor's lined script and page recalculation to the script assistant. No source publishes field-level templates.
• What reasons do cinematographers give first-hand when they commit to one focal length (or a narrow lens set) for a whole project, and where do secondary retellings go beyond what they said? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Cinematographers give different practical reasons for shooting a whole film on one focal length: Lewis (Adolescence, 32mm on a gimbal) balanced hiding rigging and booms against disorienting movement; Mukdeeprom (Call Me By Your Name, 35mm) wanted close-ups and characters in their surroundings from one lens; Suschitzky used a single 21, 24 or 27mm per film with Cronenberg. These are first-person accounts, not rules.
• What does open-access scholarship say about griot-informed oral storytelling in West African screen narratives, and what limits does it state? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  A 2022 University of Cape Town PhD thesis reviews scholarship that likens West African filmmakers to griots and proposes a model of how filmmakers mediate oral tradition, analysed through three films chosen because they adapt literature drawn from oral tradition (Xala, Karmen Gei, Invasion 1897). The griot comparison is a cited scholarly interpretation, and the selected sample cannot stand for West African cinema generally; no production workflow is documented.
• What do working cinematographers say about exposing for HDR when a proper on-set HDR monitor is not available, and where do they disagree? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Cinematographers writing in 2016-2018 agree that meters and exposure knowledge still matter in HDR but split on how much practice changes: some say to expose as for film with testing and an ACES pipeline ("use a meter ... read shadows, read highlights"), others say highlights need deliberate placement, HDR should be monitored with periodic SDR checks and a spot meter is the fallback without a monitor, and an ASC paper recommends an on-set HDR reference monitor with a waveform that validates rather than replaces exposure readings. These are practitioner accounts that pre-date current LED-volume workflows.
• How do I read a Kodak or Fujifilm camera-negative datasheet: what do the characteristic curve, spectral sensitivity and dye curves show, and under which conditions were they measured? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Kodak's VISION3 500T sheet gives exposure indexes of 500 (tungsten) and 320 (daylight with an 85 filter) and warns that its curves are representative of production coatings, not specifications. Kodak plots log exposure in lux-seconds (the 2015 sheet also states ECN-2 processing and Status M densitometry in its text; the 2022 text layer does not), while Fujifilm's ETERNA manual plots against camera stops with Status M densities, so the two makers' curves cannot be overlaid without converting the exposure axis.
• What do WPT results for the CSS features that WebVTT and IMSC rendering rely on (ruby, writing modes, bidi, emphasis) show per engine? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  The CSS layout tests beneath subtitle styling tell a different story from the WebVTT cue reftests: css-writing-modes bidi tests pass 88 of 94 in all four engines, while css-ruby passes 114, 114, 111 and 64 of 179 and text-emphasis tests 77, 90, 125 and 20 of 145 (Chrome, Edge, Firefox, Safari). So engine differences in WebVTT bidi are not explained by generic CSS bidi support; whether ruby results explain the cue failures is not established.
• What does the DNG 1.7.1.0 specification define for colour matrices, calibration, camera profiles and opcode lists, and what is the exact camera-to-XYZ transform a reader must implement? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  DNG 1.7.1.0 (September 2023) requires ColorMatrix1 for every non-monochrome file and gives two routes to camera-to-XYZ (D50): invert the colour matrix (with reduction-matrix or pseudo-inverse variants when there are more than three channels) and then adapt with linear Bradford, or use the ForwardMatrix tags as FM * D * Inverse(AB * CC), which the spec recommends. Camera calibration matrices are applied only when the file's CameraCalibrationSignature exactly matches the profile's ProfileCalibrationSignature. Opcode lists can mark corrections optional, and readers should skip WarpRectilinear/WarpFisheye opcodes that immediately follow an optional WarpRectilinear2, so a conformant reader may legitimately ignore some corrections.
• What do ARRI's look-file documents say an ALF2, ALF4 and ALF4c contain, how the look travels with clips, and what does not carry over? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  ARRI's ALF2 look files bundle creative intent with a display conversion, whereas ALF4 (ALEXA 35/265) is log-to-log: optional ASC CDL plus a Creative Modification Transform, with the display transform applied downstream; ALF4c carries custom display transforms and leaves colour management to the user. Looks travel in each clip's header, so transcoding loses them (link rather than import in Avid); spatial tools cannot live in a 3D LUT; and LogC3 LUTs cannot be converted to LogC4, so looks must be recreated. The 2016 note that ALF2 needs Resolve Studio is dated.
• What do two BAFTA screenwriters say about adaptation decisions: fidelity, structure and when the source form must be reinvented? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Two screenwriters' BAFTA lectures describe adaptation as service to the new film rather than transcription. One warns against being slavish to the source, describes drafts that dropped the novel's voice-over and later drifted back towards the book's own section structure, and a 'flash-sideways' device for scenes shown from two perspectives. The other says her first draft adapting her own play failed because tone and language did not translate, so she restarted from scratch, and advises adapting only material one relates to. These are the speakers' own accounts of their process.
• What do the Sami Ofelas (Pathfinder) guidelines say about consultation, ownership, script review and who they apply to? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  The Pathfinder (Sámi Film and Culture Advisory Board with the International Sámi Film Institute; version 22 February 2021) asks productions with Sámi content, or shot in Sápmi, to use it as early as possible and in any case before the script is completed. An expert group evaluates the project and ISFI can then confirm that it is culturally sensitive and respectful. Its checklist asks about Sámi ownership through a producer or co-producer, Sámi writers, meaningful consultation, Sámi actors for Sámi characters, Sámi language, community benefit, and whether anyone has said No; consent is mostly obtained in face-to-face consultation. It states Sámi ownership of their culture and stories, says it does not answer every question, and points to a lawyer for legal matters.
• What do the BBC Casualty accessible-audio trial, a Fraunhofer-linked thesis and Dolby's white paper report about how much viewers adjust dialogue-versus-background balance, and how variable it is? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Older listeners in one small study (11 listeners aged 66-75, no hearing devices; 2020 bachelor thesis using the Fraunhofer IIS listener pool) chose speech-to-background differences far above the 8.4 LU default: about 15.3 LU with separately produced objects and 13.5 LU with dialogue-separated material, with wide interquartile ranges (11.8-20.8 and 9.8-17.1 LU). The BBC's 2019 Casualty trial let viewers slide between the broadcast mix and a simplified mix; Dolby reports widely differing preferences; SVT chose one fixed Enhanced Speech track. Together they support personal control where separated or object audio exists, with a pre-rendered alternative as the simpler delivery (cross-source synthesis, not a measured comparison).
• When an agent follows the repository ids, version names and sample counts printed in the 1X cards and README, which of them have drifted (renamed ids, v1.1 vs v2.0, 450 vs 250 test samples) and what must be pinned? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  At observation, the ids printed in the cards (1x-technologies/worldmodel and worldmodel_raw_data) answer with HTTP 307 to world_model_tokenized_data and world_model_raw_data, so a pin should use the sha, not the name. The GENIE cards and the README leaderboard refer to challenge v1.1 and data/val_v1.1, while the tokenised-data repository (sha 42e3e12f..., lastModified 2025-04-20) describes v2.0 as replacing v1.1 and still lists the v1.1 folders. The two dataset cards give different test-set sizes (450 vs 250 samples). These are observations of text and listings, not a history of changes.
• How do WorldRoamBench's single-H20 throughput measurements for LingBot-World, HY-World 1.5, Matrix-Game 2.0/3.0 and Yume 1.5 compare with the authors' stated FPS, and what can and cannot be concluded? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  WorldRoamBench v5 (AMAP CV Lab, Alibaba Group; not the authors of the open models) timed 10 open models on one NVIDIA H20, 20 runs, as chunk frames over wall-clock chunk time. Measured: Matrix-Game 2.0 6.26 Hz (640x352, 12-frame chunks), Matrix-Game 3.0 5.34 Hz (1280x704, 40), Yume 1.5 6.49 Hz (1280x704, 32), LingBot-World 1.79 Hz (832x480, 12), LingBot-World 2.0 1.75, HY-World 1.5 1.30 Hz (832x480, 16). Against the authors' figures (LingBot 16 fps on one GPU node; Matrix-Game 2.0 25 FPS on one H100 per C3-B062-W13-05) these are lower, but the hardware, GPU count and settings differ, so none of the gaps shows an author figure to be wrong.
• What do contributors and users report when they time LingBot-World-Fast themselves on 8xH100 (chunk times, cold start, whole-clip time), and how does it compare with the authors' 16 fps / under-1-second statement? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Non-author timing logs for LingBot-World-Fast (14B, 480x832, 81 frames, 8xH100) come from a contributor's profiling pull requests: steady chunks about 1.0 s, chunk 0 about 7.2 s, VAE encode+decode about 5.6 s (26% of generate()), whole generate() 22,146 ms unoptimised, 13,523 ms after three merged PRs and 11,668 ms with a fourth, unmerged PR. If a chunk is 12 frames (WorldRoamBench's chunk size; the PR sample uses chunk_size=3), steady-state is about 12 frames/s, below the authors' 16 fps; a separate user reproducing on 8xH100 with NVLink and FA3 saw about 2.5 s/chunk around chunk 7 (about 4.8 frames/s on the same assumption), twice the 1.21 s another user reported. The discrepancy was not resolved in the thread. No key-press-to-frame measurement was found.
• What does Learned Motion Matching (Holden et al. 2020) replace, what data and training did it use, and what limits do the authors report? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Learned Motion Matching (2020) replaces the stored animation database and nearest-neighbour search of motion matching with three small networks (Decompressor, Stepper, Projector) so the runtime keeps no pose database; memory is the stated motive. Training runs on CPU for up to 500,000 iterations (a few hours for results, overnight for best). Reported scenarios span 22,398 to 694,272 frames, but the 694k Bear set was augmented from about one minute (3,581 frames) of key-framed animation. In the user study subjects struggled to identify the reference system, with a slight responsiveness preference for basic motion matching and no quality preference; the authors found Projector accuracy had the largest impact on quality.
• At the archive handoff (access file or package to FFV1/Matroska, BagIt and PREMIS/OAIS), what does each primary text guarantee, which package-level information has no home in the archive container, and which records must be written explicitly? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Integrity checks work at different layers: FFV1 with ec=1 stores a CRC in every slice (MediaConch's sample digitisation policy requires per-slice CRCs), ffmpeg's framemd5/framehash hash decoded frames so two files can be compared by content, and a BagIt 'valid' result only says every manifest checksum matched the stored files. PREMIS 3.0 leaves which events to record to the repository and provides eventOutcome/eventOutcomeDetail for results, so validation outcomes are not automatic archive metadata. Matroska (RFC 9559) can flag hearing-impaired, visually-impaired and text-description tracks, so access roles have a place if a transcode step sets them; the LoC FFV1/Matroska page says timecode support is still in development, so DCP/IMF structure and timecode need separate records (no inspected source maps CPL reels/edit rate/KDM into Matroska).
• What is the difference between scene-referred log footage and display-referred Rec.709 video? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  ARRI Log C stores relative scene exposure, logarithmic in stops (about 78 10-bit code values per stop), and ARRI describes two ways to display it: the camera's tone-mapping or a linear conversion with display correction, where values above 1.0 may clip. Rec.709 video is adjusted so the picture looks right on a BT.1886 reference monitor (EOTF exponent 2.404) in a BT.2035 environment. Both can sit in the same pixel container, so treat a file as log or as display video according to its signal metadata or documentation, not by appearance. H.273 code 2 means the transfer characteristics are unspecified and left to the application. The words scene-referred and display-referred are not used in these texts; they are convenient shorthand for ARRI's scene exposure versus BT.709's viewed picture (F1, confirmed by curator search).
• What do colourists themselves say about grading skin tones (when to use scopes, balance-first or skin-first, warm or accurate), and where do they disagree? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Colourists quoted in a 2017 bachelor thesis do not share one method for skin: one grades the base for skin when a face fills most of the frame and grades the rest first with skin in a secondary on wide shots; vectorscope skin-line use ranges from habitual early in a career to final passes only; stated preferences range from slightly warm to plain accuracy without colour contamination. An interviewed colourist describes a scene where skin and walls were all beige and the work was separating them. The thesis is secondary evidence of what each said (questionnaires cited as personal communication; one quote from a 2016 publication).
• GPT Image 2 prompt structure from fal's guide: five slots, anti-slop rules, change-vs-preserve edits, labelled multi-image roles active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Template: Scene / Subject / Important details / Use case / Constraints, with line breaks; the Constraints slot ("no watermark, no logos, preserve face, preserve layout") is where unbounded prompts drift. Anti-slop rules: replace praise words (stunning, epic, masterpiece) with visual facts (overcast daylight, brushed aluminium, 50mm feel); give style tags concrete visual targets; name the real object (boarding pass, transit kiosk); in edits separate "Change" from "Preserve" and repeat the preserve list each iteration; treat text as typography (quote it, specify font, size, colour, placement; spell hard words letter by letter); one revision per turn. Three modes: generate, edit one image, combine multiple images with each input labelled by role (Image 1: base scene to preserve; Image 2: ...). Consistent with OpenAI's own cookbook (C3-B406-01) and Google's positive-framing rule (C3-B408-01).
• How do HDR guides say exposure and lighting decisions change (or not) compared with SDR, and what do the PQ numbers imply for scopes and nit targets? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Practitioner guidance conflicts: ARRI says HDR exposure and lighting need not differ in general (watch clipped highlights, bright practicals, noise) and that there is no consensus on grading HDR or SDR first; Messerschmidt (ASC) treats HDR as a different format and places practicals at about 300-400 nits with a background window at 800 nits on the display. Netflix adds that light-meter ratios and scope readings need reinterpretation (producer reading).
• What do ARRI, Netflix, Sony and a working cinematographer say about whether and how to monitor HDR on set? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Netflix (as archived 2025-01-23) encourages but does not require HDR monitoring on set, warns that 'Bright SDR' monitors marketed as HDR-capable are unsuitable, and asks that monitors and LUT boxes support at least 33-point 3D LUTs; ARRI recommends HDR monitoring with a true HDR monitor to catch HDR-only problems early.
• What does the Hollywood Industry-Wide Labor-Management Safety Committee's portable-power bulletin say lighting crews must do on set, and how binding is it? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  CSATF Bulletin #23 (2013, recommended guidelines, not law) says only authorised employees connect or operate distribution, rig de-energised and check for shorts before energising, connect grounds, neutrals then phases (reverse to disconnect), avoid tie-ins to energised premises power (qualified person required), never use alligator clips, and use GFCI protection for non-double-insulated tools on construction sites.
• Which Open RV v4.0.2 render-path code decides which paint properties are read, and what happens to extra per-stroke properties? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Open RV's PaintIPNode recompiles drawings when a pen/text/rect/line/arrow/ellipse/frame component changes and reads a fixed property list; uuid, softDeleted and undo/redo data are stored in the session but never read by this node, so they have no render effect here (deleted shapes disappear because they are left out of the order list).
• How do the properties written by the Python OTIO annotation hook differ from those written by the Mu Annotate mode for the same pen stroke or text box? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  The OTIO import hook writes pen strokes with fixed constants (debug 1, join 3, cap 1, splat 0), copies uuid and soft-deleted state from the OTIO layer, treats invisible layers as deleted and leaves them out of frame:N.order, and writes no undo/redo stacks; the interactive Mu mode generates uuid4 ids and keeps per-user undo/redo stacks. Imported drawings therefore are not undoable like hand-drawn ones.
• What does PDF/X-4 require for print and how does it handle transparency, fonts and colour? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Per the Library of Congress format description (secondary), PDF/X-4 (ISO 15930-7:2010, PDF 1.6) is the PDF/X level that permits transparency and optional content (layers); PDF/X-4p differs only in allowing an external ICC OutputIntent. The ISO text was not read.
• Which SVG features are documented as unsupported or unreliable when rasterising or converting SVG, and how mature is SVG 2? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  resvg (commit 5192641d) declares SVG fonts, external-SVG use, colour profiles and the direction/unicode-bidi attributes unsupported, and its SVG 2 checklist marks 31 items done and 69 not done; its claim that browsers fail some static SVG cases is the project's own (README not re-read).
• What does OpenEXR say deep images represent and how are deep IDs and subsample alpha meant to interchange between tools? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Splitting an OpenEXR deep sample recomputes alpha as 1-(1-alpha)^x, which equals un-premultiplying and re-premultiplying colour, with alpha 0 handled as a special case. Deep IDs store 32-bit or paired 64-bit hashes per sample (MurmurHash3 suggested, not required) with a manifest, keeping per-sample colour and alpha so isolated objects keep correct edges; Cryptomatte stores coverage only.
• Are blend-mode formulas identical in W3C Compositing 1, PDF and OpenRaster, and in what colour encoding are they evaluated? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  W3C Compositing 1 defines blend modes as a mixing function applied in place (weighted by backdrop alpha) followed by a Porter-Duff composite, with exact formulas (soft-light uses a polynomial below Cb 0.25 and sqrt above); OpenRaster adopts these by reference. Neither text says whether blending happens on encoded or linear values; PDF 1.7 makes the blending colour space explicit (producer reading).
• How do variable-font axes and OpenType features behave in CSS, and what does CSS Fonts 4 say is ignored or clamped? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  In CSS Fonts 4 (Working Draft) font-variation-settings ignores axes a font lacks (other axes still apply), clamps out-of-range values, uses case-sensitive four-character tags, and does not influence fallback font selection; authors should set wght, wdth, slnt/ital and opsz through font-weight, font-width, font-style and font-optical-sizing instead.
• What must an agent get right for mixed-direction text and line breaking, and where do standards say algorithms are not enough? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Line breaking (UAX #14) is a tailorable rule set: Thai, Lao and Khmer need dictionary or language analysis outside the standard, and in bidirectional text breaks are decided before bidi rule L1 without using directional properties. Bidi overrides are discouraged for security reasons (UAX #9). HarfBuzz shapes one script/direction/language/font run at a time but reads five characters of context on each side, so pass surrounding text when shaping a substring.
• Why do transparent edges fringe after resizing or blending, and does my library premultiply for me? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Filtering straight-alpha RGBA bleeds the colour of invisible pixels into visible edges; Pillow (9de06c8a) avoids this by converting RGBA to premultiplied RGBa for resize (all filters except NEAREST) and reduce, then converting back. Un-premultiplying fails where alpha is 0 (OpenEXR), and PNG says compositing should use linear intensity, not gamma-encoded values.
• Which image formats store straight versus premultiplied alpha, and what does each specification say an agent must do when moving pixels between them? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Alpha conventions differ by format: PNG and WebP store straight (unassociated) colour, OpenEXR stores colour premultiplied by convention; TIFF signals associated vs unassociated per file with ExtraSamples, and AVIF signals premultiplication with a 'prem' item reference (both from producer reading). Never assume one convention across formats; read the format rule or the per-file flag.
• What is a storylet or quality-based narrative in plain terms, and what did Failbetter's own early write-up and Emily Short say about structures like questicles and the Midnight Staircase? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  A storylet is a unit of content with prerequisites on world-state qualities and effects on those qualities; it can be paper-prototyped on index cards. Failbetter names reusable gating shapes built from qualities (Faust's Tea Party social exchange; Midnight Staircase accumulated preparation cashed in at any time). These are 2010/2019 designer accounts, not measurements.
• How are creative looks carried from set to post under ACES (Look Transforms, AMF, CDL plus LUT), and what do the ACES docs say is not portable? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  ACES Look Transforms are whole-image ACES-to-ACES transforms carried as ordered Transform IDs in the ACES Metadata File; CDL values chosen under an LMT are invalid without it; non-ACES looks/LUTs are hard to translate (rebuild them in ACES); archive original ACES files plus grading data, the ordered LMTs and the ODT. 'Well-designed LMTs are interchangeable' is a stated design aim, not a test result.
• What do first-person storyboard accounts say about shot numbering, camera-move arrows and shot labels, and where do they disagree? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Practitioner accounts tie board panels to script or breakdown numbering but differ: scene number plus a letter per shot (No Film School), or sequential numbers never renumbered with letter inserts such as 12A (storyboard instructor); camera moves drawn as start and end frames joined by one arrow and one label. A feature artist reported avoiding arrows (producer reading).
• What breaks when an ACES 2.0 (or 2.1) OCIO config is opened by OCIO builds older than the config's ocio_profile_version? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  OCIO's loader rejects a config whose profile version is newer than the library supports ('not able to load that config version'), so an ACES 2.0 config (profile 2.5) is expected to fail in OCIO 2.4 or older (code inference, not run); unknown keys produce warnings rather than failures; and the ACES 2.1 config's AMF transform IDs are documented as preliminary in 2.6.0, so mappings keyed on them are not stable.
• What is the status and content of the IETF AI Preferences working group drafts (vocabulary and attachment), and what do they claim about the effect of expressing a preference? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  The IETF aipref drafts are working documents, not RFCs: vocab-08 defines usage categories (train-ai, ai-use, search) and a serialization and states it does not reflect working-group consensus; attach-05 defines a Content-Usage HTTP header and robots.txt rule and would update RFC 9309 if approved. Preferences are statements; enforcement is out of scope.

More