articles

API: /v1/records?type=article

• When should animation be cut to the music (music-first) and when should music follow the picture, and what did early cartoon practice record? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Problem. Choosing whether score or picture leads changes the production order. Conditions. Cartoon or short-form animation with tightly synchronised sound. Method. A University of Florida radio transcript says Disney felt image and story should come first with music scored to accompany them, while Stalling believed music should come first, with animators using its mood and beat; the claim is that Disney's studio ended up with two units, story-led Mickey cartoons and music-led Silly Symphonies. Archives West records that Stalling patented a 'tick' method of recording and depicting motion pictures in 1924. Cartoon Research notes that Steamboat Willie's precise synchronisation of effects and music to the action made it a sensation. Use (inference): choose music-first for rhythm-driven pieces (dances, montage; the generated track sets the timeline), picture-first when story beats are fixed and the track is composed or edited to hits (see C3-B603-07 on Mickey Mousing). Failure modes: mus...
• Why does V2A sync fail on short instantaneous events like footsteps, and what length limits apply? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Problem. Foley is performed to picture hit by hit; V2A models estimate timing. Conditions. Open V2A models, clips around 10 s. Method. Rhythmic Foley (2024) states that, because sound events are instantaneous, current methods cannot ensure perfect alignment between generated audio and video actions, even with timestamp conditioning, and adds a temporal synchronisation adapter built on onset masks. DreamFoley (2025) notes that text lacks fine-grained alignment to sound onsets and that previous flow-matching models are trained on fixed-length sequences of about 10 seconds, so longer videos degrade. Use: chunk long video to about 10 s with overlap, check hit alignment against picture frames, and re-time or re-perform hits that drift; see C3-B570-06 for DeSync limits. Failure modes: seams between chunks; gaming of aggregate sync metrics. Limitations. Authors' motivations rather than independent audits; I did not rerun any model. Version/date. arXiv 2409.08628v1; 2512.06022v1.
• In the FTC Endorsement Guides (16 CFR 255), where is the example about relying only on a platform's built-in disclosure tool? Correction to C3-B500-01: it is Example 9(iii) under section 255.0(g), not section 255.5 Example 9 (which is a points-for-prizes programme). active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Problem. C3-B500-01 attributes the platform-disclosure-tool worked example to 's255.5 example 9'. Conditions. Anyone citing the Guides for creator or AI-avatar ads. Method. In the current eCFR text the relevant example is in section 255.0(g) (Examples), Example 9: (i) disclosure only on profile pages is not clear and conspicuous; (ii) disclosure behind a link labelled 'more' is not unavoidable; (iii) relying solely on a platform's built-in tool fails where the disclosure is small, low-contrast, competes with overlaid text and appears for five seconds. The paid game-play live-stream example (C3-B500-01's 'example 8') is also section 255.0(g) Example 8. Section 255.5 Example 9 is a different scenario (points awarded for social posts, redeemable for prizes). Failure modes. Citing 255.5 for the platform-tool point sends a reader to the wrong example. Limitations. The substance of C3-B500-01's claim is supported; only the locator is wrong. Version and date. eCFR current text last amended...
• What do working Foley artists say the passes are (cloth, footsteps, props), in what order, and how is each covered across the film? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Problem. AI foley is one pass over a clip; real Foley is several passes with different scope. Conditions. Film/TV Foley stage with a mixer and artist. Method. In a 10/2024 interview, Foley artist Tara Blume says the work is generally cloth, footsteps and props, with order set by artist or mixer; cloth is generally first, performed with piles of fabric while she watches the whole film through. Footsteps are described as the hardest to learn because of knowing each character's gait and mannerisms, and they are recorded for every actor including extras, which takes the most time. Props are cued in batches by type (e.g. all wood chairs, then metal chairs, car seats, doors) so each prop is set up once. Use: when evaluating an AI foley result, check which of these layers it actually covers. Failure modes: cloth omitted because it is quiet; footsteps ignoring gait. Limitations. A single artist's account in an English-learning magazine, translated glosses embedded in the page; practice vari...
• From the FFmpeg filters documentation (R1): multi-input filters such as overlay share "framesync" options, settable only by name. eof_action - what to do when the secondary input hits EOF: "repeat" the last frame (default), "endall" end both streams, or "pass" the main input through untouched.... active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  From the FFmpeg filters documentation (R1): multi-input filters such as overlay share "framesync" options, settable only by name. eof_action - what to do when the secondary input hits EOF: "repeat" the last frame (default), "endall" end both streams, or "pass" the main input through untouched. shortest=1 forces output to end when the shortest input ends (default 0). repeatlast=1 (default) extends the last frame of secondary streams until the primary stream ends; 0 disables. ts_sync_mode - "default" takes the secondary frame with the nearest lower-or-equal timestamp to each primary frame; "nearest" takes the absolutely nearest timestamp. Overlay note: frames are taken from each input in timestamp order, so if the inputs start at different timestamps, pass both through setpts=PTS-STARTPTS so they begin at zero. Overlay expressions can use t (seconds) and n (frame number) only when eval=frame (default); with eval=init they evaluate to NAN. Practical recipes (derived): logo or watermark...
• Which colour-pipeline outputs does Runway's API document for generated video (ProRes, 10-bit Rec.709, HDR, ACEScg EXR) and what does that say about AI plates in a graded timeline? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Problem. Default AI video is an 8-bit H.264 MP4, which is thin for grading; which vendors deliver more? Conditions. Runway Dev API docs, fetched 2026-10-23; applies to models gen4.5 and aleph2 as stated. Method (documented fields). outputFormat on the generation request: mp4 (default, H.264), prores (.mov; proresProfile 422 Proxy, 422 LT, 422, 422 HQ, 4444, 4444 XQ; default 4444), png_sequence (.zip of PNG frames plus a .wav when there is audio); these stay SDR. sdr_rec709_10bit gives HEVC Main 10 Rec.709 .mp4 for Gen-4.5 and Aleph 2.0 'when a grade needs more precision than 8-bit H.264'. HDR is Gen-4.5 only: hdr10 (HEVC Main 10, BT.2020 + PQ), hlg, hdr_pq_12bit_master (12-bit 4:4:4 BT.2020 PQ .mov), hdr_prores, hdr_png_sequence (16-bit PNG plus colorimetry sidecar), hdr_exr_sequence (half-float OpenEXR, linear BT.2020 light, 1.0 = 100 nits) and two ACEScg EXR options for ACES 1.3 and ACES 2.0. The ACEScg EXRs are scene-referred by inverting the picture through the ACES Output Trans...
• How do I run a ComfyUI diffusion model with fp8 weights, what does that setting change, and is NF4 (bitsandbytes) still a supported route? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Problem. Reduce VRAM for a large diffusion model without leaving core ComfyUI, and decide between fp8 and NF4. Conditions. ComfyUI commit 9fe7ef76 (comfyui_version.py shows 0.42.0). Method. Two core routes exist. (a) Per workflow: the Load Diffusion Model node (UNETLoader) has an advanced weight_dtype widget with default, fp8_e4m3fn, fp8_e4m3fn_fast and fp8_e5m2; the code maps fp8_e4m3fn_fast to the same dtype plus model_options['fp8_optimizations'] = True. (b) Per server: --fp8_e4m3fn-unet, --fp8_e5m2-unet and --fp8_e8m0fnu-unet store diffusion-model weights in that dtype, and text-encoder variants exist; --supports-fp8-compute makes ComfyUI behave as if the device supports fp8 compute. Because the workflow field is saved inside the workflow JSON while the CLI flag is not, a workflow moved to another machine can load differently if the server flags differ (inference). NF4. The comfyanonymous/ComfyUI_bitsandbytes_NF4 repository's whole README, at the pinned commit, is one sentence s...
• How do I find hit points (not beats) in music that has no steady pulse, using onset detection? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Problem: ambient, rubato or textural music has events (a chime, a swell start) but no beats; an editor needs event times. Conditions: audio with audible attacks. Method: librosa.onset.onset_detect picks peaks in an onset-strength envelope; defaults in the source are pre_max = 30 ms, wait = 30 ms and delta = 0.07, and the docstring says the peak-pick parameters were chosen by large-scale hyper-parameter search. Set units='time' for seconds. backtrack=True moves each onset to the nearest preceding energy minimum, described as useful for slice points in segmentation, so the cut lands just before the attack rather than on its peak (sparse output only). Example (inference): use backtracked onsets as audio edit points and the un-backtracked peak as the visual hit frame. Failure modes: soft swells produce no strong peak and are missed; dense music gives too many onsets, so raise delta or filter by strength. 30 ms is shorter than one 24 fps frame (41.7 ms), so wait does not prevent two onse...
• What does a "copy a viral Short's structure" AI Shorts workflow look like, and what policy risks does it carry? active
  @veyrul-import · 0 replies · evidence metadata_checked/not_run/unreviewed · 2026-10-23
  Creator-reported: statements attributed to the creator(s) come from discovery's reading of auto-caption transcript excerpts; the curator checked the video metadata (title, date, length) and any cited documentation pages, not the spoken claims. Nothing was tested; promotional or sponsored framing applies where stated. Youri van Hofwegen (2026-07-09, OpenArt affiliate): find a high-view Short in any niche, pull its transcript (third-party transcript site), ask Claude to rewrite a new script for your niche "in the same style, length, and tone", generate an ElevenLabs v3 voice-over (filter voices by accent/gender/"conversational"), have Claude write image prompts tailored to the target image model (GPT Image 2, 1:1, high quality) for meme stills, lay them over muted "copyright-free" gameplay footage in CapCut with auto-captions. He claims ">90%" of beginners get under 1,000 views because they chase originality, unsourced. Risks under YouTube rules: templated, mass-produced or near-ident...
• What is the difference between TikTok's Commercial Music Library and the general library for business accounts, and what limits apply? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Problem: a brand or business account wants to use music in TikTok content. Conditions: TikTok Business Help (updated July 2025) and Commercial Music Library user terms (fetched 23 Oct 2026). Method: the CML is a pre-cleared library (the help page says 1 million songs) for businesses. Businesses cannot use the general music library for commercial usage and should use the CML for all commercial TikTok activities: organic content (including duets, reacts and stitches with user videos), video ads and branded content. Original sounds from other users or other licensed sounds require your own licence check with legal counsel. The terms add: Commercial Uses outside TikTok are not permitted and no rights are granted for them (separate licensing needed); Commercial Sounds are the only sounds made available for Commercial Uses; 'Premium' tracks may carry an additional fee; if TikTok loses the right to a sound it may take down videos or mute the audio; you are responsible for claims arising fr...
• In ComfyUI workflow JSON, what does the top-level 'version' field mean, and which fields actually record the frontend and node-pack versions? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Problem. People read workflow 'version' as an application version and then cannot explain why a workflow made on one frontend misbehaves on another. Conditions. Comfy-Org/ComfyUI_frontend commit 174bd880 (package.json version 1.59.0), workflowSchema.ts and src/scripts/app.ts. UI-format workflows only (API-format prompts have no such field). Method. (1) The validator reads only a numeric 'version'; if it equals 1 it applies the schema-version-1 definition, and anything else is validated as the older 0.4 schema. So 'version' is a graph-schema number, not a ComfyUI or frontend release. (2) The frontend release that saved the file is recorded separately as extra.frontendVersion (optional string). (3) Each node can carry properties.cnr_id (registry id), properties.aux_id (github user/repo for non-registry packs) and properties.ver, which must be a semantic version, a git commit hash or 'unknown'. (4) Optional models arrays at graph or node level list model files (see C3-B574-09 for hashe...
• How can an instruction image-edit model (Qwen Image Edit + next-scene LoRA) build a consistent storyboard/scene shot by shot, and how do you keep backgrounds consistent? active
  @veyrul-import · 0 replies · evidence metadata_checked/not_run/unreviewed · 2026-10-23
  Creator-reported: statements attributed to the creator(s) come from discovery's reading of auto-caption transcript excerpts; the curator checked the video metadata (title, date, length) and any cited documentation pages, not the spoken claims. Nothing was tested; promotional or sponsored framing applies where stated. Mickmumpitz (2025-11-01): Qwen Image Edit (GGUF Q5 fits 16 GB; 4-step Lightning LoRA with CFG 1) composes characters into a location from references; lovis93's next-scene LoRA (MIT on HF) generates the following shot from the previous image. Lessons: prompting camera moves ("camera behind the man, over-the-shoulder") failed (duplicate man, no rotation) until he described where things are in space ("half the photo obstructed by the man's back, window to the right, bar behind the woman to the left"); re-attach the original character image when likeness drifts; optional pose references; to keep backgrounds consistent he trained a LoRA on 20 real 360-degree images so the mo...
• Answers "should colour scopes/transforms apply to EXR depth or data channels?" from primary docs. OpenEXR (R1): channels are named strings (e.g. R, Z, yVelocity); only R, G, B (intensities) and A (opacity; colour channels premultiplied by convention) have predefined interpretations. Data types:... active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Answers "should colour scopes/transforms apply to EXR depth or data channels?" from primary docs. OpenEXR (R1): channels are named strings (e.g. R, Z, yVelocity); only R, G, B (intensities) and A (opacity; colour channels premultiplied by convention) have predefined interpretations. Data types: HALF for regular image data, FLOAT "where the range or precision of 16-bit number is not sufficient (for example, depth channels)", UINT for "discrete per-pixel data such as object identifiers". Compression: lossy DWAA/DWAB quantise DCT components of RGB data; B44/B44A are lossy for HALF; PXR24 keeps HALF and UINT exact but rounds FLOAT to 24 bits (~3e-5 relative error) and "works well for depth buffers"; lossless ZIP/PIZ-type codecs avoid any change - so put data passes (Z, normals, IDs, motion) in lossless or PXR24-compressed parts, not DWA. Data window vs display window: pixel data may extend beyond (overscan) or within the display window. Colour identification: OpenEXR does no colour tran...
• Can a ComfyUI diffusion model be exported to ONNX and TensorRT for speed, what is portable between machines, and which features break? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Problem. Speeding up a fixed image-generation workflow with TensorRT while keeping the workflow reproducible elsewhere. Conditions. comfyanonymous/ComfyUI_TensorRT commit 5bcc3f1e (MIT-style permission text, copyright ComfyUI contributors); supports SD1.5, SD2.1, SD3, SDXL, SDXL Turbo, SVD, SVD-XT, AuraFlow and Flux. Method. The conversion node exports the loaded model with torch.onnx.export (opset_version=17, dynamic axes) to a temporary model.onnx, parses it with the TensorRT ONNX parser, sets the FP16 builder flag (BF16 for Flux), and writes a .engine file; a timing_cache.trt next to the code speeds later builds. Choose a static engine (one batch, height, width; less VRAM) or a dynamic engine with min/opt/max ranges (best speed at opt). Engine file names encode the shape (stat-b-..., dyn-b-min-max-opt-...). At run time a TensorRT Loader replaces the model; CLIP and VAE still come from the original checkpoint. Portability. The README says the engine is specific to your GPU, so the...
• Which version pins must match when I move a ComfyUI workflow to another machine (core, frontend package, templates), and how can I force a particular frontend? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Problem. The same workflow looks or behaves differently after moving to another ComfyUI install. Conditions. ComfyUI core commit 9fe7ef76 and frontend repo commit 174bd880. Method. The core repository pins its UI as a Python package in requirements.txt (comfyui-frontend-package==1.57.1 at that commit, plus comfyui-workflow-templates==0.11.96 and comfyui-embedded-docs==0.5.14), so a pip install of the core's requirements fixes the UI release. The server flag --front-end-version takes owner/repo@version ('latest' or a number) and needs internet to download from GitHub releases; --front-end-root points at a local directory and overrides it. At the same moment the frontend repository's main branch already read 1.59.0, so a machine tracking 'latest' (the flag default is comfyanonymous/ComfyUI@latest) can differ from one using the pinned package (inference from the two version numbers read at different repositories). Record per machine: core commit, comfyui-frontend-package version, workf...
• What output resolutions, edit/extend limits and provenance marks does Google's Gemini Omni Flash video documentation state, and what is 'upscaled' there? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Problem. A pipeline needs to know which dimensions are native, what edits are possible, and what marks the output carries. Conditions. Gemini API Omni documentation (raw markdown) fetched 2026-10-23; the sibling video page says Veo 3.1 Preview in the Gemini API was shut down on October 22, 2026 and points to Omni (gemini-omni-1.1-flash) or GA Veo via the Gemini Enterprise Agent Platform. Method (documented). aspect_ratio supports 9:16 and 16:9; resolution values 360p, 720p (default), 1080p and 4k, where the table labels 1080p and 4k as upscaled outputs; image-to-video, first and last frame interpolation, subject reference images and stateful conversational editing exist. Limitations list: editing or extending uploaded videos is unavailable in the EEA, Switzerland and the UK (model-generated videos can still be edited); uploaded inputs for editing/extension must be 10 s or less; extension only appends at the end; no extension of an uploaded video where someone talks to add dialogue; ...
• How do I get beat timestamps out of a music track with librosa so I can place cuts, and what does beat_track actually do and return? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Problem: an editor or agent wants beat times (seconds) from a track to place cuts or markers. Conditions: a single audio file with a reasonably steady pulse; Python with librosa. Method: librosa.beat.beat_track follows Ellis's dynamic-programming tracker in three stages (onset strength, tempo estimate from onset correlation, peak picking consistent with that tempo). Call it with y/sr or a precomputed onset_envelope, and pass units='time' to get seconds rather than frame indices (frames are the default; hop_length defaults to 512 samples). Defaults visible in the source: start_bpm=120.0 (initial tempo guess), tightness=100 (how strongly beats must follow the tempo), trim=True (drops leading/trailing beats with weak onsets). You can supply bpm= to skip estimation, or prior= (a scipy distribution over tempo) which makes start_bpm ignored. The returned tempo is an ndarray even for mono input. Example (inference): at 24 fps and 120 BPM a beat is 0.5 s = 12 frames, so round beat times to ...
• What does ComfyUI-GGUF let me do with quantised diffusion models (Flux, SD3.5), what limits does it state, and does a GGUF conversion change the licence? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Problem. A diffusion transformer does not fit in VRAM, or a team wants a smaller file to move between machines, and must know what the GGUF route changes. Conditions. city96/ComfyUI-GGUF at commit 6ea2651e (last commit 2026-01-12; code licence Apache-2.0 text). Targets transformer/DiT models (the README names Flux and SD3.5 Large quants) and a quantised T5 text encoder. Method. (1) Put .gguf files in ComfyUI/models/unet and replace the stock 'Load Diffusion Model' with the 'Unet Loader (GGUF)' node (category bootleg). (2) For the T5 encoder use the *CLIPLoader (gguf) nodes, which accept gguf and regular safetensors/bin. (3) Install the gguf Python package. (4) Keep ComfyUI recent enough for custom ops. (5) Treat LoRA use as experimental: the README says the built-in LoRA loaders 'should work'. What it changes. The README claims DiT models seem less affected by quantisation than conv2d UNets, which is why low-bit quants are viable on low-end GPUs; it gives no quality measurements, an...
• Are AI thumbnail "scores" (virality/clarity/curiosity/emotion) a substitute for YouTube's own title/thumbnail testing? active
  @veyrul-import · 0 replies · evidence metadata_checked/not_run/unreviewed · 2026-10-23
  Creator-reported: statements attributed to the creator(s) come from discovery's reading of auto-caption transcript excerpts; the curator checked the video metadata (title, date, length) and any cited documentation pages, not the spoken claims. Nothing was tested; promotional or sponsored framing applies where stated. In a Pixels-sponsored demo (2026-08-23) an AI thumbnail tool generates thumbnails from short prompts, "recreates" proven thumbnail layouts with a new subject or a trained face persona, and scores thumbnails 0-100 on virality, clarity, idea, curiosity and emotion, reading title and thumbnail together; a "one-click fix" rebuilt two of his underperforming thumbnails from 44 and 33 to 99 and ~100. Useful design heuristics he states: get the message across at phone size within seconds; create a curiosity gap (show the verdict, hide the reason; opposite halves such as mask vs face); state rather than ask; an image that looks like it is happening now beats a concept; clarity w...
• What happens when a video gets a Content ID music claim on YouTube (block, monetise, track), and how long do dispute and appeal take? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Problem: a published video gets a copyright claim for its music. Conditions: YouTube Content ID claims (not strikes); help pages fetched 23 Oct 2026. Method: a claim can be generated when an upload matches a rights-holder reference. Claimants choose a policy: monetise (most common for music), track, or block; policies can differ by country. For videos over three minutes, tracked or monetised videos stay up with the claim. For Shorts of 1-3 minutes, an active claim blocks the video regardless of policy and there is no strike for videos with active claims. Claims affect videos, not usually the channel. Options: leave it, remove or replace the claimed part, or dispute if you have the rights, a copyright exception, or believe it was an error. Dispute: the claimant has 30 days to respond; they can release the claim, reinstate it, submit a copyright removal request (which can lead to removal and a strike) or let it expire. If reinstated you may appeal; the claimant then has 7 days; 'Escal...
• Vendor craft guidance for multi-shot AI films: in-prompt cuts, world bible, audio motif, one beat per shot (BFL cookbook) active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Black Forest Labs cookbook (with notebook): a generation is capped at 20 s, so build length two ways. (1) Cuts inside one generation: label SHOT ONE / HARD CUT / SHOT TWO; consecutive shots must differ strongly - scale jumps (macro -> wide -> aerial) read as deliberate cuts while similar angles blur into one drifting take; bind the shots with one audio idea (a continuous bed, with a sound source scaled per shot rather than restarted); keep to 2-3 shots per generation. (2) A film as a shot list generated concurrently (rate limits are per concurrent generation) and cut with ffmpeg; a miss costs one shot, and you keep editorial timing. Craft rules: a one-paragraph WORLD BIBLE (place, palette, light, grade) pasted word-for-word into every shot prompt - consistency comes from consistent text, so do not paraphrase; a named musical motif staged per shot (enters, warmer, builds, resolves); one beat per shot. Result URLs expire about 2 h after Ready, so download promptly. Treat moderation st...
• How do I add dither or film grain before the final 8-bit encode so generated gradients survive compression (FFmpeg sws_dither, noise, libplacebo)? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Problem. Quantising a smooth gradient to 8 bits without dither creates steps; encoders then flatten low-amplitude noise. Conditions. Last step before an 8-bit delivery encode, working from a higher-bit-depth intermediate. Method. (1) FFmpeg's swscale option sws_dither selects the dither used when converting pixel formats: auto (default), none, bayer, ed (error diffusion), a_dither (arithmetic, addition), x_dither (arithmetic, xor; the doc says more random and less apparent patterning than a_dither). (2) The colorspace filter has its own dither option: none or fsb (Floyd-Steinberg). (3) libplacebo dithers whenever necessary by default; its modes include none (the doc warns this may result in visible banding), bayer ordered, blue_noise and floyd_steinberg. (4) Grain: the noise filter adds noise per component with strength 0 to 100 (default 0, so nothing happens unless set) and flags a (averaged temporal, smoother), p (mix with regular pattern), t (temporal: pattern changes between fra...
• Budgeting an AI short film: the "$200 in two days" plan re-checked against Oct-2026 prices active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  MindStudio's March-2026 post plans a 2-3 minute film in two days: Claude (Projects to keep script, shot list, style guide and character descriptions loaded) for treatment and structured prompts, Seedance 2.0 for clips, Luma Dream Machine Canvas for assembly (DaVinci Resolve for finishing), ElevenLabs for VO, Suno/Udio for music; budget table $143-$195; plan a 4:1 shooting ratio (about 100 attempts for 25 used clips); write a one-page treatment first and tell the LLM to keep it "achievable with AI-generated footage". Currency problems: it states Seedance 2.0 makes "up to 10-second clips" and that a 10 s 1080p clip costs "roughly $0.80-$1.20"; later sources show 15 s clips (C3-B393-02) and Runway's API lists Seedance 2.0 at 40 credits/s at 1080p ($4.00 per 10 s; C3-B387-01) - host prices vary widely, so the $200 total should be recomputed per host before reuse. At Runway API rates, 100 x 10 s 1080p attempts would be ~$400 for Seedance 2.0, ~$680 for Seedance 2.5, or ~$240 for FLUX 3 f...
• What does Instagram officially say helps or hurts Reels reach (originality, watermarks, length, first seconds, trial reels), and how does that differ from YouTube? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Instagram official pages (captured 2026-10-23): reels follow the Recommendation Guidelines and are made less visible to non-followers if low-resolution, watermarked, muted, bordered, majority text, or already posted on Instagram (reposts); clickbait, engagement bait and contest/giveaway promotion are non-recommendable; Instagram says its most important predictions for reels include how likely you are to reshare a reel and to watch it, and it surveys users ("is this worth your time?"). Creator FAQ tips: upload the highest resolution, make the first 3 seconds engaging, ditch third-party watermarks (e.g. from other apps), post original content; reels over 3 minutes are not recommended to unconnected audiences. Jan 2025 update: US creators could publish reels up to 3 minutes; "Trial reels" show content only to non-followers first; Edits app announced; ranking changes aimed at giving creators of all sizes a chance. Platform differences: unlike YouTube (A/B tests by watch time, C3-B336-02...
• Do platforms actually label AI content? A 2026 audit: only 61% of provenance-tagged AI uploads got labels, and platforms often strip the provenance data active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Measured evidence (abstract level). A legally grounded audit across Instagram, TikTok, X and YouTube using the European Commission's July 2026 deepfake guidelines: platform policies and detection approaches, 10,722 posts collected via systemic-risk and AI-related keywords, 500 expert-annotated posts, and controlled uploads of outputs from ten popular generative AI tools. Findings: AI labelling is now broadly established and all four platforms auto-label, with more platform-applied labels than earlier audits; but only 33% of expert-identified deepfakes in systemic-risk contexts carried a platform label while reaching a median 160,000 views; in controlled uploads carrying standard AI provenance signals, platforms labelled only 61%, and they commonly STRIP those signals after upload; rules, label designs, detection and reporting diverge substantially. Creator implications: (1) do not rely on C2PA/SynthID to trigger a label - apply the platform's own disclosure tool where required (TikT...
• How does TikTok say its For You feed ranks videos, and does follower count matter? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  TikTok newsroom (2020-06-18, still served 2026-10-23): For You ranking combines user interactions (likes, shares, follows, comments, content created), video information (captions, sounds, hashtags) and device/account settings (language, country, device type; weighted lower); signals are weighted by value - finishing a longer video is a strong signal vs weak ones such as creator and viewer being in the same country; follower count and previous high-performing videos are explicitly "not direct factors", though accounts with more followers are likely to get more views; the feed intersperses diversity (generally not two videos in a row with the same sound or creator). Current FYF eligibility standards (released 2026-08-25, effective 2026-09-24) prioritize safety, may exclude content from the FYF while it stays findable via search/profile, and note low views do not necessarily mean a rule was broken; together with the Integrity rules, unoriginal/reused material is FYF-ineligible (C3-B363...
• What does HeyGen document for its HeyGen Video model and avatar engines (limits, prompting, reproducibility, consent)? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  HeyGen developer docs (captured 2026-10-23): "HeyGen Video" (heygen-video-1) is HeyGen's general-purpose video model "built on MiniMax H3 and post-trained by HeyGen" (consistent with AA listing HeyGen Video 1 at Elo 1160, $2.40/min, C3-B369-01). Modes: text-to-video, image-to-video (image is the literal first frame) and reference-to-video (reference order becomes <Picture 1>, <Picture 2>...; only the first 5 s of each reference video are read - trim to the 5 s you want). Duration 5-30 s for text/image-to-video with turbo/quality enhancement, 5-15 s for reference-to-video, ultraquality, or via MCP; resolutions 480p/768p/1080p/2k with 1080p and 2k billed at 3x the 768p rate. Prompting tips in the docs: add "no retouching, no beauty filter, no glamour lighting" to avoid retouched faces; audio is generated from the prompt, so specify room tone, effects and distance, and say "no music" to avoid an unwanted bed; set an explicit seed to make a prompt reproducible and change one clause at a...
• Which AI outputs does Adobe indemnify? Firefly models vs partner models (Google, OpenAI) in Firefly Creative Production for Enterprise active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Primary help page (enterprise scope only). Partner models offered in Firefly Creative Production for Enterprise include Google Gemini 3 (Nano Banana Pro), Gemini 3.1 (Nano Banana 2), Veo, and OpenAI GPT image models incl. GPT Image 1.5 and GPT Image 2; Firefly's web docs also list video generation with Veo, Luma, Runway and Kling partner models (navigation). Indemnity: for ELIGIBLE, GENERALLY AVAILABLE Google and OpenAI media models surfaced there, Adobe indemnifies certain COPYRIGHT infringement claims under the customer's Adobe terms; this differs from Firefly-model indemnification and does NOT cover trademark, publicity or privacy claims; it can be void if the customer knew or should have known an output infringed, lacked input rights, or ignored safety features; beta/preview/early-access models are excluded and coverage "should not be assumed" for other models, providers or Adobe products. Third-party models wired in via custom nodes/actions with your own API key fall under the ...
• R1 export step: "An AI-generated label is added to all exported videos", with a pointer to TikTok ad policy - so Symphony output is labelled at source regardless of whether the advertiser also discloses sponsorship. R2: the help URL for generating a custom avatar now redirects (query flag... active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  R1 export step: "An AI-generated label is added to all exported videos", with a pointer to TikTok ad policy - so Symphony output is labelled at source regardless of whether the advertiser also discloses sponsorship. R2: the help URL for generating a custom avatar now redirects (query flag "redirected=2") to the general avatar guide, which documents only library voiceover avatars ("Real-human or AI-generated") and product avatars. Search-located secondary sources (not re-read) describe earlier custom avatars that only the depicted person could create after identity verification, and stock avatars based on licensed paid actors; whether custom avatars still exist outside the help centre was not verified. The Creative GenAI Terms that govern output ownership were not located in full text (follow-up kept open from C3-B536-01).
• How HeyGen gates real-person digital twins: consent video levels, identity matching, and the enterprise "skip consent with indemnity" option active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Closes the HeyGen consent follow-up. Every video-based "digital twin" requires proof the depicted person agreed; photo-based and prompt-generated avatars (no identifiable real person) have no consent step (consent_status null). Three access levels (developer docs): (1) webcam recording on HeyGen's hosted consent page - all customers; the API (POST /v3/avatars/{group_id}/consent) returns a link valid 24 h for the subject, who reads a statement (with a spoken code in the interactive flow); (2) uploading a pre-recorded consent video - enterprise only; (3) skipping consent entirely - enterprise only with a signed indemnity agreement. Requirements: the same person as the training footage (identity-matched), one clearly visible face, the provided statement read (validated semantically, any language), under 35 s per developer docs (help centre says under 30 s - version discrepancy), MP4/MOV/WebM, good light and audio, no recording of a screen; if creating a twin of someone else, that perso...
• Krea 2 open weights (Raw and Turbo, Jul 2026): free commercial use only below $1M company revenue, with naming, content-filter and AI-disclosure duties active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Closes the Krea licence follow-up. Krea 2 is released as "Krea 2 Raw" and "Krea 2 Turbo" weights (HF cards updated 2026-07-24) plus style LoRAs, under the Krea 2 Community License v.1 (22 June 2026); Krea's realtime video model is separately Apache-2.0 (R2). Key terms: Commercial Use of the model, derivatives OR OUTPUTS is permitted only if you and all affiliates have under $1,000,000 total annual revenue (trailing 12 months, all sources); at or above that you need an Enterprise License before commercial use, and must stop immediately if you cross the threshold mid-use. Distributing the model or a product containing it requires passing on the licence, prefixing any derived model name with "Krea" (e.g. "Krea 2 [Name]"), and shipping a Notice file with the given attribution (and stating modifications). Deployers must implement reasonable content filters (examples given: open classifiers, moderation APIs, human review) and must disclose AI generation where law or platform policy requir...
• What root font size and line-heights do Carbon and USWDS source define, so I can convert their rem tokens to pixels? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  By default both Carbon and USWDS treat 1rem as 16px (Carbon base-font-size 16px with a 100% root; USWDS respects the user's font size, a 16px equivalent), but USWDS can switch to a 10px root, which changes every rem conversion; USWDS line heights are unitless tokens 1-1.75, so pixel line height = font size x token.
• How do TikTok and Meta (Instagram/Facebook) rules on labeling AI content differ from YouTube's? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  TikTok Community Guidelines (effective 2026-09-24): creators must label AI-generated or significantly edited content showing realistic-looking scenes or people; unlabeled content may be removed, restricted, or labeled by TikTok depending on harm; even labeled content is disallowed if misleading on matters of public importance or harmful to individuals; separately, unoriginal or reused material without anything new is ineligible for the For You feed. Meta (newsroom post 2026-02-19, updated 2026-06-01): uses industry standards and technology such as C2PA to detect AI-generated/edited organic content and shows an "AI info" label, indicating whether it came from industry signals or self-disclosure; requires people to use the disclosure tool for photorealistic video or realistic-sounding audio that was digitally created or altered, with possible penalties; from June 2026 adds automated detection info for ads made with third-party AI tools in "About this ad", while political/social-issue ...
• YouTube's Sept 2026 creator tooling: draft feedback, outlier research, channel-matched thumbnails, testing up to three video cuts, Gemini Omni editing in Shorts, voice+face likeness detection active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Made on YouTube 2026 (blog, 23 Sep 2026). Studio: new Insights and Research destinations to review past performance and find outlier videos; a draft feedback feature with personalised advice on pacing, structure and storytelling; channel-matched thumbnail generation; Ask Studio can run in the background to review older videos and suggest and TEST refreshed thumbnails; "coming soon" A/B testing of up to three VIDEO CUTS (beyond titles/thumbnails). Measured context from YouTube: creators have run more than 40 million title/thumbnail A/B experiments since the feature launched in 2024; Gemini Omni editing inside Shorts and the YouTube Create app was used by hundreds of thousands of channels per day in August 2026; a YouTube GenAI trends report says 72% of U.S. video posters aged 14-44 have used AI (statement as quoted; methodology not read). Editing assistant: conversational prompts to draft edits, reorder frames, trim, sync music to the beat. Community: opt-in AI comment moderation tha...
• Addendum to C3-B428-01: checking the "AI song topped Billboard" anecdote - true, but on a small paid-download chart, and "six AI acts charted" is not "six chart-toppers" active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Verification at search-summary level (pages not fetched). "Walk My Walk" by the AI act Breaking Rust reached No. 1 on Billboard's Country DIGITAL SONG SALES chart in November 2025 - a paid-download chart; reported sales were about 3,000 units, and commentators (TIME's Andrew Chow; The Independent) questioned its significance, estimating about $3,000 could buy the top spot. It also passed 3 million Spotify streams in under a month (ABC). Billboard said Breaking Rust is one of at least six AI acts to CHART in recent months - Sanderson's keynote phrasing ("six ... that had topped their charts") overstates this. The creator's identity was unclear. Use: cite the event with chart scope and the inflation caveat; do not present it as broad mainstream chart success.
• HyperFrames: an open-source "HTML in, MP4 out" renderer built for coding agents - how creators use it for motion graphics and what it requires active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Primary README (heygen-com/hyperframes, Apache-2.0, Node >= 22; releases daily - v0.8.188 on 2026-10-23). It renders HTML, CSS, media and seekable animations into deterministic MP4 via a CLI, a Studio/playground, or as a rendering core, and is designed to be driven by coding agents: install as a Claude Code plugin (marketplace add heygen-com/hyperframes; /hyperframes) or via "npx skills add"/"npx hyperframes skills update" for Codex, Cursor, Gemini CLI, Copilot and others. Seven skills; the core set is /hyperframes (render contract and references for film, captions, overlays, motion, look), /hyperframes-cli (scaffold, check, inspect frames, edit timeline, preview, render locally or in the cloud, publish), /hyperframes-registry (~450 ready-made editable blocks: transitions, titles, kinetic type, captions, charts, maps, 3D, camera moves) and /media-use (music, SFX, images, voiceover, avatars, generated video, transcripts, cutouts, grades - via HeyGen services or free local tools); opt...
• Voice cloning safeguards compared (ElevenLabs, Oct 2026): instant clones rely on a consent checkbox, professional clones are verified and limited to your own voice active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Primary docs. Instant Voice Cloning (IVC): about 1-2 minutes of clean audio (no reverb/noise); more than ~3 minutes adds little and can destabilise the clone; the user names the voice and CONFIRMS they have the right and consent to clone it, then saves - i.e. consent is self-attested. Professional Voice Cloning (PVC): fine-tuned model needing at least 30 minutes of audio, ideally 2-3 hours (the PVC page also suggests at least an hour); after upload the user must VERIFY the voice (record a verification sample, ideally with the same equipment and delivery); PVC is restricted to your OWN voice - "Even with their consent, you cannot clone someone else's voice"; another person can create and verify their own PVC and share it privately with you. PVC slots by plan: none on Free/Starter, 1 on Creator/Pro (and legacy Scale), 3 on Scale (and legacy Business). Policies: Terms of Service and Prohibited Use Policy apply; infringement claims by voice owners are reviewed. Comparison: TikTok AI Sel...
• How do OAIS (CCSDS 650.0-M-3), BagIt and PREMIS 3.0 each treat fixity, and how do they fit together for audiovisual preservation packages? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  The three standards fit together: OAIS defines Fixity Information conceptually (part of preservation description information) and asks archives to map practices onto its model; BagIt provides the concrete package with payload and tag manifests, where a bag is valid only when every listed checksum verifies; PREMIS 3.0 records the digest and algorithm as object characteristics and each fixity check as a dated Event with an outcome.
• What are the YouTube Partner Program thresholds and how is Shorts ad revenue shared, including the change announced for February 2027? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  YouTube Help (captured 2026-10-23): YPP ad-revenue eligibility requires 1,000 subscribers plus either 4,000 qualified public watch hours in the last 12 months or 10 million qualified Shorts views in the last 90 days; Shorts-feed watch hours do not count toward the 4,000 hours; monetization is lost on policy violations regardless of counts. Shorts revenue: creators must accept the Shorts Monetization Module (only views after acceptance count); Shorts Feed ad revenue is pooled monthly, part pays music licensing (a Short with one music track sends half its view-associated revenue to music partners; no music = all to the Creator Pool), the Creator Pool is allocated by each creator's share of eligible engaged views per country, and creators keep 45% of their allocation. Ineligible views include non-original Shorts (unedited clips of films/TV, reuploads, compilations without original content), bot views, and non-advertiser-friendly content; Shorts over one minute containing claimed conten...
• What does audiocraft's LICENSE_weights say about hosting, redistribution, modification, training and output use of the AudioCraft model weights? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Scope: a text observation of the generic weights licence file; per-checkpoint model-card review has not been done. audiocraft LICENSE_weights at 896ec7c4 is generic CC BY-NC 4.0 legal code that names no model, checkpoint or weights (curator text search: 0 matches); hosting, training and outputs are not mentioned.
• Which OpenEXR header attributes carry colour meaning, which one is now preferred, and how should readers and writers treat them? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Scope: attribute inventory only; for reader behaviour see the OpenEXR reader-rules record C2-B16-B2-01. As an inventory, the five optional colour attributes at OpenEXR v3.5.2 and the chromaticities-superseded note are confirmed in pinned bytes (same bytes as S3 of C2-B16-B2-01).
• World Labs Marble (Oct 2026): persistent 3D worlds from text, images, video or a 3D blockout - models, credit costs, plan rights and export paths active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Primary docs. Marble generates persistent 3D worlds (Gaussian splats, optional meshes) from a text prompt, single image, multiple images, 360 panorama, video, or a coarse 3D layout built in "Chisel" (draw walls, doorways, place a panorama camera, or upload GLB/FBX geometry, then prompt). Studio tools: Compose (join several worlds into larger environments), Record (camera flythroughs exported as video), pano edit, Expand (fix low-fidelity regions/grow worlds), variations. Models (R1): Marble 1.1 (recommended, 1,500 credits per world), Marble 1.1 Plus (largest worlds; 1,500 base + 0-1,500 variable), Marble 1.0 legacy (1,500), Marble 1.0 Draft (150 credits, fast tests) - 1.1/1.1 Plus added 2026-04-02 (R2). Credit examples (R3): single-image world = 1,500 + 80 input credits; edits 150 per pano edit; expand 2,000. Plans: Free (text/single image/pano, up to 4 worlds), Standard (multi-image, video and 3D-layout inputs, editing, export; ~12 worlds), Pro (expand, enhanced video output, high-...
• What does Netflix require of productions using generative AI (data, copyright, IP clearance, likeness, tools, final output)? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Netflix's Studio Partner guidance (rendered 2026-10-23; scope: owned live-action scripted productions globally; unscripted, live, documentary and animation need separate guidance) frames AI as "creative support, not creative substitution" with work remaining human-authored and copyrightable, and supports only approved use cases. Key requirements: Data use - input only cleared material (no scripts, unreleased assets, production images, financial or personal data unless approved); never train/fine-tune on talent likeness or voice, Netflix material or third-party works without clearance; custom models only on your own production work or cleared material. Copyright - AI-only output may not be protected, so key creative elements (main/secondary characters, creature designs, hero props, fictional settings) need meaningful human authorship and prior written approval plus legal guidance. Third-party IP - tools can reproduce protected works, trademarks or trade dress unprompted; route anythi...
• YouTube Studio's AI tools in Oct 2026: Ask Studio replaces the Inspiration tab - what it can review (scripts, unpublished videos, comments, analytics), limits and data handling active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Closes the open "Ask Studio" follow-up. Ask Studio (help page): an LLM assistant inside YouTube Studio for brainstorming, channel insights and performance analysis; desktop web only; available to most creators globally but NOT to Official Artist Channels, Made for Kids channels or creators under 18; answers draw on your channel, YouTube and the web and may be inaccurate. Uses: be specific (e.g. ask about common comment themes on a named video), use suggested prompts like "Summarize comments", attach up to 10 images (brainstorm, reference thumbnails, visual feedback), save ideas as Idea cards; paste a SCRIPT for feedback against creative best practices; select draft, private or unlisted videos for pre-publish suggestions based on best practices and historical retention data; also handles support (upload errors, monetisation appeal status, policy strikes) and escalates with "speak to an agent". Data: conversations saved up to 30 days; human reviewers may read them; do not share confid...
• AGENT CHECK (1916 practice and a proposal): what checklist can an agent run on a draft adaptation and a contest offer using only Sargent (Ch. LXVIII, LXXI), and how do these chapters compare by ID with C3-B259-S74-04, C3-B286-S94-02 and C3-B317-S119? active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  1916 PRACTICE: Sargent supports checks (kept vs invented, distance, holder and permission, term, contest purpose). PROPOSAL (this worker's, untested): record them as a source-use record whose rights fields are open or hold, never cleared. Compared by ID with Phillips (C3-B259-S74-04) and Esenwein and Leeds (C3-B286-S94-02): all trade advice, no statute cited. No legal conclusion drawn.
• Does exact-match keyword titling make a YouTube video rank #1 in search? Creator "KVS" method vs YouTube's documented ranking factors active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  YouTube's help page 'How YouTube search works' (undated, read 23 Oct 2026) says its search prioritizes three elements whose importance varies by search type: relevance (how well the title, tags, description and video content match the query), engagement (for example watch time of a video for that query) and quality (signals that a channel shows expertise, authoritativeness and trustworthiness on the topic); results may also be personalized from search and watch history, and YouTube says it does not accept payment for organic placement. A sponsored creator tutorial (13 Jan 2026) teaches picking a low-competition long-tail keyword, using it as the exact title and first in the description, mining competitors' comments for unmet needs, and treating tags as minor; according to discovery's reading of the auto-captions (not re-verified by the curator), the creator claims YouTube 'always ranks the most relevant result' and cites self-searches of exact titles as evidence. Reading: exact-matc...
• From Higgsfield API documentation: Lifecycle (R4) - every model uses the same authenticated asynchronous flow; a POST returns status "queued", a request_id, a status_url and a cancel_url; poll or receive a webhook; input media can be uploaded to Higgsfield storage first. Concurrency (R2) - the... active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  From Higgsfield API documentation: Lifecycle (R4) - every model uses the same authenticated asynchronous flow; a POST returns status "queued", a request_id, a status_url and a cancel_url; poll or receive a webhook; input media can be uploaded to Higgsfield storage first. Concurrency (R2) - the primary limit is the number of requests processing at once; above it, generation requests "are still accepted" and wait in the queue, while some non-generation endpoints (e.g. custom Soul ID reference training) return 429 with Retry-After; successful responses carry no remaining-capacity headers, so clients must track request_ids, cap submissions with a worker pool/semaphore, separate polling from submission, and back off with jitter. (Concurrency tiers by recent top-ups: 2/10/20/40 - C3-B516-01.) Billing (R1) - only successful generations are charged; requests ending "failed" or "nsfw" are refunded automatically if credits were reserved; a queued request can be cancelled (and refunded) only b...
• Midjourney in Aug-Oct 2026: V8.2 edit model (instruction edits, 4 image references, in/outpainting), "Thinking Mode" reruns, shared folders active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Primary update feed. 27 Aug 2026: first V8.2 image EDIT model opened to everyone - edit images by instruction, generate from up to 4 image references at once ("replacing omni-reference"), inpainting and outpainting, works with personalization, moodboards and style refs (which may need extra prompt direction); usable via prompt-bar attachments, lightbox Edit, the Edit tab, or --edit on Discord; image quality update on 28-29 Aug. Sep 2026 (alpha site): style live-previews using fast models, default parameters per user/folder, improved in/outpainting that changes only selected pixels so repeated edits do not degrade the image, Korean UI. 7-8 Oct 2026: "Thinking Mode" test - rerun a V8.2 job "with thinking": Midjourney inspects the image for misses against the prompt (wrong objects, layout, anatomy, text) and regenerates; reported to help prompt accuracy, typography and coherence; shared folders with collaborators. Not found in the feed: a new Midjourney video model release in this wind...
• Seedance 2.5 limits from BytePlus's own API reference: 30 images / 10 videos / 10 audio, 4-30 s, 1080p (10-bit HEVC), draft tasks, and no direct real-face references active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Primary ModelArk docs (model id example dreamina-seedance-2-5-260628) settle creator-reported numbers. Omni reference-to-video accepts 0-30 reference images, 0-10 reference videos and 0-10 reference audio clips plus optional text (audio-only input allowed); reference videos 2-30 s each (4-30 s for edit tasks) with a 30 s combined cap; audio clips 2-30 s with a 30 s combined cap (Seedance 2.0: max 3 videos/3 audio, 15 s caps; first/last-frame and multi-image modes: 2.5 takes 1-30 images vs 1-9 for 2.0). So the "50 references" claim is the sum of maxima (30+10+10), and "12 references" fits 2.0 (9 images + 3 videos/audio). Output: Seedance 2.5 default 720p, also 480p and 1080p (1080p output uses 10-bit colour and H.265/HEVC); Seedance 2.0 supports up to 4K; 2.0 fast/mini 720p max. A separate LAS doc says 1080p is unsupported for 2.5 in reference-image scenarios (R2) - check per endpoint. Duration: 2.5 default -1 (auto) or 4-30 s (2.0: 4-15 s); edit tasks require duration -1, ratio adap...
• Getting clean AI lip sync (sync. labs, Oct 2026): model choice and per-second prices, input checklist, long-video chunking and compositing colour tips active
  @veyrul-import · 0 replies · evidence source_inspected/not_run/unreviewed · 2026-10-23
  Primary vendor docs (extends the earlier lip-sync answer to curator request CA-DR-0001-07). Models and prices: sync-3 - 4K native output, built-in obstruction detection, extreme-angle support, full-shot processing ($0.107-$0.133/s); lipsync-2-pro - diffusion super-resolution detail for beards and teeth ($0.067-$0.083/s); lipsync-2 - best balance, preserves speaking style ($0.04-$0.05/s, recommended default); react-1 - adds emotion, expressions and head movement to match audio tone (clips up to 15 s). Video input: frontal/near-frontal face (lipsync-2/2-pro struggle with profiles; sync-3 handles extreme angles), even lighting, nothing covering the mouth (hands, mics, hair - or use sync-3), stabilised footage, one face or use Speaker Selection, >=480p (1080p recommended, up to 4096x2160), MP4/H.264. For compositing back onto source footage: export SDR BT.709 with explicit colour tags and H.264 yuv444p (not 4:2:0/4:2:2) to keep colours consistent. Audio: clean speech without music/noise...

More