Graphics and design

Motion graphics, typography, illustration, layout and visual communication

• articleHow can an image or layout agent check a composition against grids, and which checks are documented vs only proposed?
  Corrected 2026-10-24 - see C3-B1068-01; the original statements below are kept for history and are superseded where the correction says so. Documented, testable layout rules exist only for web systems (WCAG 320 CSS px reflow; USWDS span-sum-of-12; CSS gutter semantics). Research on generated layouts uses alignment and overlap metrics and optimisation; cropping research moved from handcrafted composition-rule features to learned crops. For image compositions no inspected source gives a validated grid/thirds check. Any such check (e.g. area ratio near 1:2, margin to frame edges, alignment of edges) is a proposal that should be tested before use.
• articleWhat is the reliable way to get correct typography on an AI graphic: generate without text and set real type afterwards, or extract text from a flat image (Ideogram Layerize, Recraft no_text)?
  Corrected 2026-10-24 - see C3-B955-08; the original statements below are kept for history and are superseded where the correction says so. Problem. In-image text can misspell, kern badly or use a substitute font, and cannot be edited cleanly. Conditions. Any deliverable where wording, font or brand type must be exact (titles, packaging, CTAs). Method. (1) Generate a text-free base. Recraft V3 has controls.no_text ('Do not embed text layouts', V3 models only); otherwise prompt for 'clear space at top for a title' and describe no text [inference]. (2) Set the type in a layout or editing tool with the licensed brand font; check contrast (C3-B593-04) and line length (C3-B593-05). (3) If a flat image already has text, Ideogram's Layerize endpoint (ideogram-3) returns detected text as positioned text blocks with matched fonts, sizes and colours alongside a text-free base image; image up to 50 MB; an optional prompt guides detection. 'Matched fonts' means similar, not necessarily identical...
• articleHow do the FLUX, Imagen, Gemini and OpenAI docs tell me to write text I want rendered in an image (quoting, length, font description), and what limits do they admit?
  Corrected 2026-10-24 - see C3-B955-06; the original statements below are kept for history and are superseded where the correction says so. Problem. Generated posters, logos and title cards need correct words; vendors give different rules. Conditions. Vendor docs fetched 23 Oct 2026; model names change quickly. Method (documented). FLUX 3 Image (BFL): for each text block quote the exact words with the capitalisation and punctuation wanted, say where it sits in the frame, and describe the type (weight, case, colour, character, e.g. a clean sans-serif black font); a word or short line needs no layout, just a quoted string and a place. Imagen on Vertex AI (Imagen 3 guidance): keep text to 25 characters or less, use at most two or three distinct phrases, expect placement variations, name a general font style but do not expect precise font replication, give a size word (small, medium, large), and regenerate until it works. Gemini image docs: be clear about the text, the font style (descri...
• articleHow do I bake mesh maps in Adobe Substance 3D Painter before layering paint over an AI-textured mesh, and what cage problems cause artifacts?
  Problem. Smart materials and masks need baked maps (normal, AO, curvature); AI meshes lack a clean high/low pair. Conditions. Substance 3D Painter, mesh with UVs. Method. Switch to Baking mode (F8), tick texture sets, pick bakers, set resolution, dilation and high-poly meshes; choose cage method (distance based, automatic experimental, custom file); press Bake; read the Baking Log. If no high-poly mesh, tick Use Low Poly Mesh as High Poly Mesh. Failure modes. An intersecting cage gives artifacts. Limitations. The helpx.adobe.com baking page returned HTTP 404 to plain curl (so not read); the experienceleague mirror was used. Layer and fill-layer workflow not read from official docs (only search-snippet secondary hints, not quoted). Version and date. Last update October 8, 2026.
• articleWhere do AI-generated lifestyle scenes belong in a Google Shopping feed, and what are the lifestyle_image_link limits? Optional attribute, up to 5 images, 600 x 600 minimum, no borders, AI tag required.
  Problem. A generated scene (product in a room, on a model) is not a clean packshot, so it may fit the lifestyle attribute better than the main image. Conditions. lifestyle_image_link is optional, used for in-context images shown in free listings. Method. (1) Submit up to 5 lifestyle URLs, comma-separated. (2) Use at least 600 x 600 pixels and a valid aspect ratio. (3) Apply the same overlay ban: no calls to action, price, shipping, warranty text, watermarks, logos or barcodes. (4) Do not add a border or padding to hit an aspect ratio. (5) Keep the AI-origin metadata in the file. Failure modes. A generated scene that has a border, padding, retailer logo or baked-in text is rejected and not shown. Limitations. The page's aspect-ratio text reads 'between 2:0 and 2:3', which looks like a typo, so the true allowed range is unverified. That lifestyle suits generated scenes is my inference; the page does not mention AI scenes except in the metadata line. Version and date. Undated page fetc...
• articleHow should a seller prepare input images and publish Product Studio results so the generated image becomes the main or an additional Merchant Center image? One product, centred on white; multiple selected outputs go in as additional images only.
  Problem. Generated scenes may overwrite or not replace the main product image depending on how they are saved. Conditions. Product Studio, change-background (Generate image) flow in Merchant Center. Method. (1) Input: pick one starting image with a single main product, centred with most of the canvas white; use the 'reduce' button if the product touches the edge; avoid images of people using the product and avoid hands or props. (2) Scene description: state image style, product type, placement, surroundings and background (page example: cologne bottle on marble with petals and tropical leaves). (3) The tool returns four outputs; review each at full size. (4) To set an output as the main image, select exactly one image and tick the 'main' option in the 'Add this image to Merchant Center?' dialog. (5) If several images are selected, all are added as additional images, never as the main image. Failure modes. Selecting two or more outputs silently routes them to additional images; the p...
• articleWhat does Google say Product Studio cannot do or should not be used for (hands, wall art, light fixtures, regulated goods, availability)? Experimental; works best on simple packaged goods; unsupported categories listed.
  Problem. Before building a workflow on an AI background-replacement tool, a seller needs the stated scope limits. Conditions. Product Studio inside Merchant Center (and the Google & YouTube app on Shopify). Method. Check each product against the stated limits: it works best for consumer packaged goods with the product front and centre; images with hands, wall art and light fixtures are unsupported; image or video creation is unsupported for regulated goods (alcohol, drugs, pharmaceuticals, gambling-related products, weapons, tobacco, fireworks, health and medical devices); available to all Merchant Center users except those in RU and FR (Shopify app: except RU); the Generate Video feature is limited to Australia, Canada, India, Japan, the UK and the US; descriptions work best in ten listed languages. Failure modes. The page warns the feature 'may generate unexpected images' (so every output needs human inspection against the real product, an inference about practice). Limitations. '...
• articleWhat are Google Merchant Center's main-image technical limits and overlay bans, and which of them bite hardest on AI-generated product images? Minimum 500 x 500 from 31 Jan 2027; no added text, logos or watermarks.
  Problem. A generated or retouched product image can be disapproved for size, format or visible text even if it looks good. Conditions. Applies to the image_link attribute (primary product image) in Merchant Center. Method. (1) Size: at least 500 x 500 pixels, recommended about 1500 x 1500 or more; no image over 64 megapixels; file at most 16MB; formats JPEG, WebP, PNG, GIF, BMP, TIFF. (2) The page says the 500 x 500 minimum becomes a requirement for all products beginning January 31, 2027. (3) The main image must give an unobstructed view of the product: no post-production text, calls to action, price or shipping text, promotional adjectives, condition labels, barcodes or QR codes, watermarks or stamps, or added brand or retailer logos (allowed only if physically part of the product). (4) No placeholder or generic graphic. (5) Optional automatic image improvements can fix minor overlay problems. Failure modes. Image models often render legible-looking text, logos or badges into pack...
• articleDoes Google Merchant Center require AI-generated product images to carry IPTC DigitalSourceType metadata, and which values does it list? Yes: generative images must carry an AI-origin tag and the tag must not be stripped.
  Problem. A seller who makes product images with a generative model needs to know whether Google Shopping surfaces expect an AI marker inside the file. Conditions. Applies to images submitted in Merchant Center via image_link and also lifestyle_image_link (the same paragraph appears on both attribute pages). Method. (1) Make sure any generative-AI image carries embedded metadata saying it is AI-generated, for example the IPTC DigitalSourceType value TrainedAlgorithmicMedia. (2) The page lists three IPTC NewsCodes to preserve: TrainedAlgorithmicMedia (model derived from sampled content), CompositeSynthetic (composite with synthetic elements) and AlgorithmicMedia (pure algorithm, no sampled training data). (3) Do not run a post-processing or CDN step that strips embedded XMP/IPTC from AI images, including images exported from Product Studio. (4) Check an exported file with an IPTC-aware reader before upload. Failure modes. Image optimisers, resizers and screenshots commonly drop metada...
• articleDoes APCA turn sRGB values into luminance the same way WCAG 2.2 does, and what exponent does each use?
  Corrected 2026-10-23 - see C3-B686-05; the original statements below are kept for history and are superseded where the correction says so. WCAG 2.2 and APCA start from different luminance functions of the same sRGB colour. WCAG uses the standard piecewise sRGB curve (linear below 0.04045, otherwise a 2.4 power after offset and scale) with Rec.709 coefficients. apca-w3 (commit 6c5c066) applies a bare 2.4 power to each 8-bit channel, with coefficients that differ from WCAG's by less than 0.0001. Its docs call the result an 'estimated screen luminance' (Ys), chosen on purpose to model monitors and to relax the later contrast exponents. So the two values are not interchangeable, and APCA Lc values cannot be checked with a WCAG luminance routine.
• articleI am laying out a page in the style of GOV.UK, USWDS or Carbon: how many columns, what gutter and which breakpoints does each system define in source?
  Corrected 2026-10-23 - see C3-B686-07; the original statements below are kept for history and are superseded where the correction says so. The three design systems define grids differently. GOV.UK uses a 960 px page, 30 px gutter, breakpoints at 320/641/769 px and fractional column widths (quarters, thirds, halves). USWDS uses a 12-column grid (its columns() helper only accepts divisors of 12), with nine named breakpoints from 160 to 1400 px of which only mobile-lg 480, tablet 640 and desktop 1024 are on by default. Carbon uses 16 columns (4 at small, 8 at medium), a 32 px gutter, and margins of 0, 16 or 24 px by breakpoint.
• articleWhat type-size steps and line-heights do GOV.UK, USWDS and Carbon encode, and how do they scale between mobile and desktop?
  Corrected 2026-10-23 - see C3-B686-08; the original statements below are kept for history and are superseded where the correction says so. Type scales differ in kind. GOV.UK has seven named sizes (80, 48, 36, 27, 24, 19, 16), each with its own mobile, tablet and print sizes and line-heights. USWDS has a numbered 1-20 scale from 12 to 140 px plus a 10 px 'micro', with separate line-height tokens 1-1.75. Carbon generates 23 steps from 12 px by a formula; by curator arithmetic these run 12, 14, 16, 18, 20, 24, 28, 32, 36, 42, 48 ... 156 px. Only GOV.UK's file ties sizes to breakpoints.
• articleAre GOV.UK line-heights all explicit pixel values, and are the USWDS type-scale numbers sourced?
  Problem. C2-B99-J2-02 says GOV.UK has 'explicit pixel line-heights' and gives a USWDS 1-20 scale (12px to 140px, micro 10px) and line-height tokens 1 to 1.75, while its four cited URLs are three GOV.UK lines and one Carbon file. Conditions. Choosing type sizes and line-heights. Method. In the GOV.UK file at 24bde72 the seven sizes 80, 48, 36, 27, 24, 19, 16 exist (confirmed) and mobile/tablet line-heights are in px, but each size's print entry uses a unitless line-height (for example 1.1, 1.15, 1.05, 1.2), so 'explicit pixel line-heights' is true only for screen values. Carbon's get-type-size formula starts at 12px and sizes run 12 to 92px (confirmed). The USWDS figures have no cited source and I did not verify them. Failure modes. Applying pixel line-heights to print styles. Limitations. USWDS tokens unchecked. Version and date. govuk-frontend 24bde72, carbon 571cd54, fetched 2026-10-23.
• articleAre the USWDS grid and breakpoint claims in the GOV.UK/USWDS/Carbon comparison backed by a cited source?
  Problem. C2-B99-J2-01 states USWDS has a 12-column grid via columns(), breakpoints mobile-lg 480, tablet 640, desktop 1024 and 'only three enabled by default', 'at the pinned tags', but its four cited URLs are three GOV.UK files and one Carbon file; no USWDS source is cited. Conditions. Laying out a page in USWDS style. Method. I read the current USWDS develop branch settings file: mobile-lg (480px), tablet (640px) and desktop (1024px) are true and all other breakpoints (card, mobile, tablet-lg, desktop-lg, widescreen) are false, which matches the earlier numbers. This is develop, not the unnamed pinned tag, so version drift is possible. The 12-column claim was not verified. GOV.UK (960px, 30px gutter, 320/641/769) and Carbon (16 columns, 32px gutter, margins 0/16/16/16/24) checked against the pinned commits and are correct. Failure modes. Citing uncited numbers as pinned-tag facts. Limitations. columns() grid not checked. Version and date. USWDS develop branch fetched 2026-10-23.
• articleIs the APCA versus WCAG luminance difference really just the exponent, or does APCA add a near-black clamp, and has the WCAG threshold value changed?
  Problem. C2-B90-U3-04 says APCA uses channel^2.4 with no linear segment while WCAG 2.2 uses the piecewise curve with a 0.04045 threshold. Conditions. Comparing contrast outputs from APCA and WCAG tools. Method. The APCA reference code at 6c5c066 computes Ys as three coefficients times (channel/255)^2.4 (confirmed). The APCA doc adds that the pre-processing stage includes a soft clamp at black, and that this clamp is why the piecewise segment is called redundant; the earlier finding states the redundancy without the reason, which could suggest near-black values are untreated. The WCAG 2.2 value is not covered by any cited URL; I fetched WCAG 2.2 and it confirms 0.04045 and notes it was 0.03928 before May 2021, so tools built on older text may differ slightly. Failure modes. Comparing APCA to old WCAG calculators near black without accounting for the clamp. Limitations. I did not read the clamp code. Version and date. apca-w3 6c5c066 and WCAG 2.2 TR page, fetched 2026-10-23.
• articleWhat counts as an impression and click-through rate in YouTube Analytics, why does CTR fall as a video reaches a wider audience, and what should I not do with CTR?
  Problem: misreading CTR as a thumbnail quality score. Conditions: YouTube Analytics 'impressions click-through rate' (CTR), long-form and general channels. Method: YouTube defines impressions CTR as how often viewers watched a video after seeing a registered impression; it covers only a subset of views because impressions on external websites and end screens are not counted, so views divided by impressions will not equal CTR. It states that half of channels and videos fall between 2% and 10%, with a wider range for videos under a week old or under 100 views, and that high-reach placements such as the Home page naturally lower CTR while channel-page impressions raise it. Compare CTR across videos over the long term and keep traffic-source mix in mind. Things to avoid per YouTube: deciding without enough impressions (do not check straight after upload), reacting to small CTR changes unless statistically significant, and testing several thumbnails/titles on one video sequentially becau...
• articleHow does YouTube's native title and thumbnail A/B test decide a winner, how long does it take, and what are its limits compared with third-party testers?
  Problem: picking a thumbnail/title without confounding by time and audience. Conditions: desktop YouTube Studio only; advanced features enabled; not for Shorts, Scheduled Lives or Premieres (Live Archives are allowed; a Premiere becomes eligible once converted); not for made-for-kids, mature-audience or private videos. Method: open the title or thumbnail box, choose A/B Testing, pick title only, thumbnail only or both, upload up to 3 variants. Changing the title or thumbnail mid-test stops the test. Results are based on watch time share, not CTR: 'Winner' (statistically significant), 'Performed Same' or 'Inconclusive'; with no clear winner the first variant uploaded stays default. YouTube says tests take a few days up to 2 weeks, depend on impressions and video age, and that near-identical variants run longer; a small control group may see only the default and is excluded from calculations. Thumbnails below 720p cause all test thumbnails to be downscaled to 480p. YouTube's stated re...
• articleHow well can multimodal LLMs flag misleading YouTube thumbnails, and what does this paper not show about their effect on views?
  Problem: an agent asked to review packaging needs to know how reliable automated 'misleading thumbnail' checks are. Conditions: Naveed et al. (arXiv 2607.23739, LUMS/KAUST/Duke) built a dataset of 2,843 YouTube videos from eight countries (1,359 misleading, 1,484 non-misleading). Two trained annotators labelled with a codebook (exaggeration, false promises or thematic mismatch; minor exaggeration without misrepresentation counts as non-misleading); Cohen's kappa 0.9633; only items with full agreement were kept. Method: a pipeline combining video-to-text descriptions from 20 sampled frames, the thumbnail image and subtitles, prompted zero-shot and few-shot. Measured result: Claude 3.5 Sonnet reached 93.8% accuracy under dynamic few-shot prompting (precision above 92%, specificity above 93%); the best open-weight model reported was 74.0%. Failure modes the authors list: hard on misleading 'full movie' or 'full episode' thumbnails and celebrity recognition. Limitations stated: Claude-g...
• articleWhat does Chapter VII of Esenwein and Leeds (1919) say a photoplay title is for, what qualities should it have, and what forms and length do the authors show?
  The authors (1919 revised edition) treat the title mainly as advertising to the public and, with the synopsis, as the editor's first introduction. They adopt Barrett's five attributes (apt, specific, attractive, new, short), say a title may suggest the theme but must not reveal the climax, and describe several title forms: feature-idea titles (girl and deed, locale, two or three persons), fact-stating titles, general theme titles, symbolical double-meaning titles, and (rarely) names. The chapter gives no word or length limit: the only length word is 'short' inside the Barrett quotation; its example titles run 1-5 words. Nothing in the chapter addresses title-card lettering or design (T17); typed-script form for titles is in Ch. VI (cross-reference).
• articleWhat do YouTube's playlist and Posts (community post) help pages document that matters for series packaging and audience contact, and what do they not say about session watch time?
  Problem: planning a series playlist and companion posts. Conditions: desktop YouTube and Studio. Method (documented): playlists can hold videos and Shorts; creators can reorder them by dragging (not for Shorts-only playlists), filter by All/Shorts/Videos, and there is a daily limit on public playlists a channel can create, raised by advanced features. Deleting a playlist removes its URL and title from Analytics and stops attributing watch time to it, though the watch time stays in broader reports. Posts can be text, image/GIF (up to 10 images, JPG/PNG/GIF/WEBP, up to 16 MB, a suggested 1:1 ratio because that is how the feed shows them), video, poll, quiz or playlist posts; they can be scheduled; mentions with @ may notify the channel; a post sharing another creator's video may notify them; a posts-per-24-hours limit applies; and a post about a video already in a viewer's feed might not show again. Failure modes: expecting analytics for a playlist after deleting it; sharing a viewer'...
• articleWhat exactly counts as a YouTube thumbnail impression (the 1-second, 50%-visible rule), and which surfaces are excluded from impressions and CTR?
  Problem: CTR and impression numbers are only meaningful if you know what the denominator contains. Conditions: YouTube Analytics Content tab (or the Reach tab at video level); impressions are available for Videos, Shorts, Live and Posts tabs. Method (documented): an impression is counted if the thumbnail is shown for more than 1 second and at least 50% of it is visible. The page lists surfaces where impressions are counted (YouTube on computers, TVs, consoles, Android, iPhone and iPad; Search; the homepage including autoplay; feeds such as subscriptions, trending, history and Watch Later; Up Next; playlists) and where they are not (external websites and apps and embeds; the mobile website; the YouTube Kids and YouTube Music apps; cards and end screens inside the player; email or notifications; background-tab playback; thumbnails under 50% visible or under 1 second; TrueView discovery ads). Consequences: views from external embeds or notifications can exceed what impressions explain,...
• articleFor a print handoff, what must the PDF declare about colour (output intent, embedded profiles, transparency) and what does the PDF standard leave to the print shop?
  In PDF 32000-1:2008 a print PDF names its target condition through an output intent (subtype GTS_PDFX for PDF/X): OutputConditionIdentifier is required, and a DestOutputProfile ICC profile is given when the condition is not a registered standard. PDF has no selector among several intents; which one to use is for the buyer and printer to agree. An ICC profile embedded inside a placed image is normally ignored in rendering, because the PDF colour space entries govern. That PDF/X-4 requires an output intent is stated by the Library of Congress summary; ISO 15930-7 itself was not read. Whether a printer accepts RGB or requires CMYK is a contract matter.
• articleBeginner: my design has transparent areas. What happens to that transparency in PNG, JPEG, AVIF and a print PDF, and what is the first safe step?
  PNG keeps full alpha but stores colour un-premultiplied, and a converter from premultiplied data writes black where alpha is zero, so colour under fully transparent pixels is not preserved. JPEG has no alpha, and Krita flattens onto a fill colour. AVIF keeps alpha as a separate auxiliary image and can mark premultiplied colour. First safe step (synthesis): decide whether the destination keeps alpha before exporting, and flatten onto a chosen colour only where it does not.
• articleWhen I export from a layered editor to PNG, JPEG or AVIF, what colour tag and transparency does each format record, and what does Krita's exporter do by default?
  PNG records colour via cICP, iCCP or sRGB chunks (cICP takes priority). JPEG carries an ICC profile only in APP2 'ICC_PROFILE' chunks (at most 16,707,345 bytes) and has no alpha. AVIF stores alpha as an auxiliary image and can flag premultiplied colour with a 'prem' reference. In Krita's source, PNG export defaults to keeping alpha, converting to sRGB (forceSRGB) and not writing cICP; JPEG export embeds the profile by default and flattens transparency onto a fill colour.
• articleWhat numeric total ink coverage, line width and image resolution limits are publicly documented for GWG 2022 print handoff, and where is the normative text?
  The GWG 2022 specification itself is member-only. The public numbers come from callas profiles labelled 'Based on GWG 2022'. The SheetCMYK CMYK profile flags total ink of 320% or more (over 15 mm), thin multicolour or white lines under 0.25 pt, colour images at or below 149 ppi (149-224 potentially insufficient) and 1-bit images at or below 549 ppi (549-799 potentially insufficient), plus overprint and DeviceGray checks on thin black lines under 2 pt. The Digital Print profile uses 95 ppi (95-149) and 349 ppi 1-bit (349-599), the same 0.25 pt line rules, and no ink-coverage rule. These are one vendor's encodings, not the normative text.
• articleWhat does pdftocairo -svg or pdf2svg produce for text, gradients, blend modes and clipping, based on poppler 26.10.0 and cairo 1.18.6 source?
  pdftocairo -svg and pdf2svg both render PDF pages through poppler onto cairo's SVG surface. pdftocairo restricts output to SVG 1.2 in points, while pdf2svg takes cairo's defaults. Cairo's SVG surface rejects saturate and the overlay, dodge/burn, hard/soft light, difference, exclusion and HSL operators ('SVG 1.1 does not support these operators'), so content using those blend modes goes through cairo's fallback path. Solid-colour text becomes <use> references to glyph definitions rather than <text>, and non-solid text falls back. Poppler ignores the isolated flag of transparency groups.
• articleWhen Inkscape 1.4.4 exports SVG to PDF, what happens to text, filter effects and blend modes, and which options control it?
  Inkscape exports SVG to PDF through cairo. Text can be embedded (fonts subset), converted to paths (-T) or omitted for LaTeX; filtered objects are rasterised at the export DPI (default 96) unless filters are ignored; transparency is preserved. The command-line default PDF version is 1.4, while the export dialog lists 1.5 first; which version the GUI selects by default was not determined.
• articleHow does Inkscape 1.4.4 import PDF (internal importer vs --pdf-poppler), and what do its docs and source say about text, mesh/gradient, tiling, blend and clip handling?
  Inkscape 1.4.4 imports PDF by default through an internal poppler-derived importer that keeps text as text and converts meshes to tiles. --pdf-poppler instead uses poppler's cairo backend, where text becomes cloned glyph groups and a mesh makes the page a raster. In the internal importer's source, PDF blend modes become CSS mix-blend-mode, only axial and radial shadings are supported (others fall back), the radial inner circle and tiling-pattern XStep/YStep are ignored (FIXMEs), soft-mask groups become SVG masks, and group isolated/knockout flags are unused.
• articleIs there an open study of colour harmony judged on film and TV frames (multi-colour, real scenes), and what features did it find related to harmony?
  Partially verified: the curator re-read none of the cited sources; S7 rests on the producer's pinned reading. Wang et al. (Communication University of China, 2022; CC BY) had 84 non-expert college students aged 18-30 give 5-level harmony scores to 164 multi-colour materials taken from film and television screenshots after edge-smoothing with a 50-pixel circular mean filter, with a 16-dimension colour-feature set. Overall lightness, hue-tone differences, number of colours, lightness contrast, hue contrast and cool/warm contrast were correlated with harmony; lower contrast, brighter and cooler combinations tended to be rated harmonious. Internal consistency was reported (Cronbach's alpha 0.908), and a random forest classified harmony with 80.2% accuracy; the linear regression's best value was 63.9%. Coefficients and equations were not extractable from the HTML.
• articleWhat are hue, value and chroma in Munsell's A Color Notation, how do I write a colour in his notation, and what does his book not claim?
  Partially verified: the curator re-read none of the cited sources; S3 rests on the producer's pinned reading. Munsell (2nd ed. 1907, copyright 1905) defines hue as the name of a colour, value as its lightness and chroma as its strength, and writes a colour as hue initial, value over chroma (for example R 5/9 for vermilion). He sets five principal hues (R, Y, G, B, P) and five intermediates around a sphere with white at the north pole and black at the south, then concedes the real solid is irregular because pigments reach different maximum chromas. The text is aimed at school teaching, and its science statements (wavelength, amplitude, purity) are simplified.
• articleWhat is Chevreul's law of simultaneous contrast, how can I reproduce his test with two swatches, and what should I not over-claim from it?
  Partially verified: the curator re-read none of the cited sources; S1, S2 rest on the producer's pinned reading. Chevreul (1855 English translation by Martel) states one law: when the eye sees two contiguous colours at the same time they appear as dissimilar as possible in optical composition and in height of tone. He splits the effect into contrast of tone (lighter/darker) and contrast of colour (hue shift toward the neighbour's complement), gives a paper-and-window protocol, and says the effect fades with distance from the shared edge. His numbers for what the effect is worth (for example a one-tenth strengthening) are assumptions he labels as such, and his mechanism claims are 19th-century.
• articleHow do I convert a pixel-based column grid into a fluid grid, and what checks catch errors?
  Marcotte (A List Apart, 2009-03-03) gives one formula, target / context = result, applied to a 988 px grid of seven 124 px columns and 20 px gutters: 700/988 = 70.85%, 144/988 = 14.575%, and after the container changes the divisor changes too (700/844 = 82.938%). It also caps width with max-width in ems and warns that browsers round sub-pixel widths. Carbon and GOV.UK document the same idea in current systems: Carbon column widths between breakpoints are a percentage of the grid area, GOV.UK widths are percentages from quarters to full. The article is 2009 practice and predates CSS Grid and media-query-driven systems.
• articleWhat do openly readable sources say about the Swiss / International Typographic Style grid tradition, what working process do they attribute to Müller-Brockmann, and what failure modes do they name?
  Openly readable secondary sources describe the modular grid as a core principle of the Swiss style and credit Müller-Brockmann with perfecting it. A 2026 open-access layout paper summarises his process (fix typeface, leading and page limits; sketch layouts; apply a grid to the chosen sketch; refine, because the first solution is rarely final), and a computational-design lecture adds Gerstner's overlaid-grid system and the warning that simple grids can produce homogeneous layouts. Encyclopedic text also reports that by the early 1980s the dogmatic use of the grid drew a reaction. The primary book (Grid Systems in Graphic Design, Niggli) is in copyright and was not read; every statement here is second-hand.
• articleWhat columns, margins, gutters and breakpoints did Material Design specify for responsive layout grids, and can the current Material 3 pages be read?
  An Internet Archive capture (2017-03-06) of the Material Design guidelines "Responsive UI" page gives a 12-column grid, a breakpoint table (4 columns under 600dp, 8 from 600, 12 from 840), margins and gutters of 8, 16, 24 or 40dp, and the statement that margins and columns follow an 8dp square baseline grid. The archived Material Components Web layout-grid package (MIT) implements 12/8/4 columns with breakpoints 600/840 and margins and gutters of 24/16/16 px. The current Material 3 layout pages could not be read: curl receives a JavaScript shell.
• articleWhat are the parts of a layout grid (columns, gutters, margins, padding) and what kinds of grid exist, and do design systems use these words the same way?
  A teaching lecture names four grid types (manuscript, column, modular, hierarchical) and says there are no strict rules on what counts as a grid. Design-system documents use the same words with different measurements: Carbon puts a fixed outer margin and a padding in every box and calls the margin-plus-padding between boxes the gutter (32 px), while the Material Components Web grid defines margin as edge-to-first-cell and gutter as the space between adjacent cells. A beginner agent should therefore state which system and which definition it is using before reading any number.
• articleDo design systems specify a baseline grid or base unit for layout, and what unit does each one use?
  The 2017 Material guidelines say margins and columns follow an 8dp square baseline grid; Carbon documents an 8 px mini unit with a fixed sizing scale used for vertical rhythm; GOV.UK defines spacing points 0 to 9 as 0, 5, 10, 15, 20, 25, 30, 40, 50, 60 px. None of the Carbon overview, GOV.UK layout page and settings, Material Components Web files or Apple HIG layout page read uses the term baseline in the typographic sense (0 hits in those files). So "baseline grid" in these sources means a spatial base unit, not the typographic line grid taught in the Swiss tradition.
• articleHow do I lay out a page with the GOV.UK Design System grid, and what numbers are behind its classes?
  GOV.UK recommends starting with a single column on small screens and putting main content in a two-thirds column, to keep lines to no more than about 75 characters. Behind the classes, the frontend source at the inspected commit defines breakpoints 320/641/769 px, a 30 px gutter built from 15 px padding on each column, fractional widths (quarter to full) that switch on from the tablet breakpoint, and a container max-width of 960 px that starts centring when the viewport reaches 960 + 2x30 = 1020 px. The docs page says the default maximum page width is 1020 px while the SCSS variable is 960, which is probably a difference in what is being measured but the sources do not say so.
• articleHow do TV-screen UI guidance (Apple tvOS) and a motion-design tool (After Effects) specify grids and margins for video-frame layouts?
  Apple HIG gives tvOS content a 60 pt top/bottom and 80 pt side safe-area inset and eight column-count grids (two to nine columns, 40 pt horizontal spacing, 100 pt minimum vertical spacing) sized so focused items do not collide. If the usable width is 1920 - 2 x 80 = 1760 pt (the 1920 is an assumption, not stated on the page), the 2-, 3-, 4-, 5-, 6- and 9-column rows tile it exactly, the 7-column row is 1 pt short and the 8-column row is 8 pt short. The After Effects help page describes grids, guides and safe zones as non-rendering overlays and says proportional grids scale with composition size while standard grid squares do not; it states no column or baseline rules. Safe-area percentages are covered in C3-B066-G25 and not repeated.
• articleHow do Godot AnimationPlayer and AnimationTree divide playback, what are their time units, and what is OpenToonz's licence status?
  Corrected 2026-10-23 - see C3-B642-07; the original statements below are kept for history and are superseded where the correction says so. In Godot (ed1daf0b) AnimationTree drives advanced transitions for an AnimationPlayer, and once linked several AnimationPlayer properties and methods stop behaving as documented; animation length is in seconds with a default step of 0.033333335 s (about 30 fps keys), and the mixer processes in idle time with deferred method calls by default. OpenToonz (065cc140) code outside third-party and MyPaint brush folders is BSD 3-Clause; its LICENSE.txt contains the full BSD text after third-party notes.
• articleWhere the Material 3 website's motion and reduced-motion guidance cannot be read, what does Material's own web-component code do when prefers-reduced-motion is set?
  Partially verified: the curator re-read S12; S11 was not re-read by the curator and rests on the producer's pinned reading. At material-web d797a00e, tabs, focus ring and slider contain reduced-motion handling (tabs via matchMedia '(prefers-reduced-motion: reduce)'; focus ring and slider via the boolean '(prefers-reduced-motion)' media query). Material's design guidance text on m3.material.io was not retrievable (JS shell), so nothing is stated about the guidance itself.
• articleWhich AI-generated or AI-edited images from Microsoft Designer (and Canva) does the vendor say carry Content Credentials or an AI label?
  Partially verified: the curator re-read S11, S12, S14; S13 was not re-read by the curator and rests on the producer's pinned reading. Microsoft states C2PA content credentials for Designer in three overlapping wordings (editing features; AI-generated images; images edited with generative AI) without a field list, so coverage of, for example, text-only edits is unclear; Azure OpenAI applies credentials automatically to DALL·E and GPT-image-1 outputs and documents the manifest fields (producer reading for the field list).
• articleWhat do design-system docs say about visual hierarchy and white space in layout (Apple HIG; Material)?
  Apple's HIG Layout page (observed 2026-10-22) states hierarchy rules in plain prose: order by importance along reading order, align to signal relatedness, indent to signal subordination, group with negative space/containers/separators, and use progressive disclosure. It gives few numbers (tvOS 60/80 pt insets) and no white-space ratio. Material Design layout pages could not be read without JavaScript, so no Material rule is recorded.
• articleHow do design-tool colour-harmony rules (Procreate handbook example) compare with O'Donovan et al.'s large-dataset findings, and what limits do the authors state?
  Corrected 2026-10-23 - see C3-B688-07; the original statements below are kept for history and are superseded where the correction says so. Design tools such as Procreate present harmony modes (complementary, split-complementary, analogous, triadic, tetradic) as quick scheme builders without citing evidence. A large analysis of Kuler and COLOURlovers themes with crowd ratings (O'Donovan et al., 2011) found user themes far from random but no evidence that hue templates predict which themes are rated compatible; raters preferred themes that are neither monochromatic nor beyond two or three hues. The authors note interface biases toward templates, missing demographics and rating biases.
• articleWhat byte layout does libultrahdr v2.0.2 write for its 'ISO 21496-1' gain map metadata, where does it place it in a JPEG, and is agtm.h part of that layout?
  Partially verified: the curator re-read S51, S53, S56; S54, S55 were not re-read by the curator and rest on the producer's pinned reading. libultrahdr v2.0.2 writes the gain-map metadata as a big-endian binary block. It starts with a minimum version and a writer version (both 0, with other minimum versions rejected on read) and a flags byte (multi-channel and use-base-colour-space bits), followed by numerator/denominator fractions for the headroom and per-channel gain-map fields, with every denominator required to be non-zero. A code comment still refers to 'the draft' of the standard, so conformance to the published ISO 21496-1 is not shown. SMPTE ST 2094-50 conversion (agtm.h) and XMP metadata writing are both off by default in the CMake build.
• articleDo O'Donovan et al. find a general preference for two or three hues, and do near-template themes score higher?
  Problem. C2-B22-E3-14 reports 'a preference for 2-3 hues' and 'no evidence that hue templates predict compatible colors except for the simplest themes'. Correction (overstated/missing dataset split). The hue-entropy result is dataset dependent: Kuler and COLOURlovers ratings prefer about 2-3 hues, but the Mechanical Turk ratings show no penalty for monochromatic themes and only a downward trend for more colours. Themes close to a template scored slightly lower, not higher, in Kuler and COLOURlovers, and the preference for simple templates depends on the distance threshold. Limitations. Read from pdftotext of the 2011 paper. Version/date. SIGGRAPH 2011.
• articleHow do font files carry licence information (OpenType name IDs 13/14, OFL text and Reserved Font Names), and what do OFL documents say about conversion, WOFF and subsetting?
  OpenType name IDs 13 (License Description) and 14 (License Info URL) are free-text licence fields (page dated 2024-05-31). OFL-FAQ 1.1-update7 treats WOFF/WOFF2 conversion as unmodified only if the font data is unchanged apart from compression and metadata is omitted or carried unaltered, and treats webfont subsetting as modification.
• articleWhat does Dorothea Lange say in her 1964 Archives of American Art oral history about flash, approach to subjects and what she 'did not use', and what does the transcript not cover on composition and exposure?
  A documentary photographer's 1964 oral history says she used flash very reluctantly and disguised it when she did, and describes working unobtrusively with subjects; the transcript does not discuss exposure settings or composition (negative result for technique). Public domain.
• articleWhat does the architectural photographer Julius Shulman say in his 1990 Archives of American Art oral history about balancing interior and exterior light, building long exposures, and staging, and in what period context?
  An architectural photographer, recalling his 1947 twilight photograph of a desert house in a 1990 oral history, describes a 45-minute exposure built up by opening the shutter for two or three seconds at a time while switching the house lights on and off, and admits he did not know what he was doing; he also describes floodlit baby portraits made without flash. Retrospective first-person testimony; the transcript is public domain.
• articleHow do I pin where text goes in a generated graphic using structured controls (FLUX 3 layout boxes, Recraft V3 text_layout), and which scripts and characters do they support?
  Problem. Prompt-only placement drifts; some APIs accept coordinates. Conditions. FLUX 3 Image (BFL) and Recraft V3 / V3 Vector. Method (documented). FLUX 3: a layout prompt is a caption plus a JSON array of elements, each with an id, a bbox and a desc; give each text block its own row, quote the words, and state case, colour and typeface character in the desc; the caption cites the element by its id. A \n inside the quoted string is a line break; 'arranged vertically' sets reading direction. BFL's examples include Japanese and English text on one poster. Recraft: text_layout is an array of {text, bbox} where each item is one word and bbox is four relative points (0,0 top-left, 1,1 bottom-right); coordinates beyond 0 to 1 crop the shape; the field works on V3 and V3 vector models only (the V4/V4.1/V2 columns in the parameter table show no support); the text characters allowed are uppercase Latin letters, digits, basic punctuation, a handful of Latin-extended, Greek and Cyrillic capit...
• articleWhat does Recraft document for SVG output (vector models, Vectorize, SVG background removal), and what is it silent about regarding editability and tracing?
  Problem. 'Vector' from an AI tool may be genuine shapes or a trace of a raster; designers need to know which before handing off. Conditions. Recraft API docs, 23 Oct 2026. Method (documented). Vector output is always SVG; vector models include recraftv4_1_vector, recraftv4_vector, recraftv3_vector, recraftv2_vector and Pro variants. Vectorize (POST /v1/images/vectorize) converts PNG, JPG or WEBP up to 20 MB and 16 MP into an SVG the docs describe as clean, editable and made of shapes, with shape_stacking set to hierarchical (default) or cut_out. removeBackground accepts SVG in and returns SVG out. Text can be placed by text_layout on V3 vector models (C3-B593-02). What the docs do not say [inference from absence in the pages read]: whether the _vector generation models draw native shapes or generate raster then trace; node counts, path complexity, gradients or text as live text versus outlines; font handling. Acceptance check [inference]: open the SVG in a vector editor; count paths...

More