AssetStoreまとめ 割引情報 beta

- アセットの価格推移をリアルタイムにチェック -

Twenty playful Halloween sounds: party horns, comic screams, cartoon frights, animatronic props, spooky organ and a wobbling theremin. Fun rather than fear.Twenty playful Halloween sounds, built as a working toolkit rather than a highlight reel: the comic end of the season, where the fright is a punchline rather than a threat. Dry enough to process, long enough to loop or trim.WHAT'S INCLUDED20 sound effects, uncompressed WAV, 44.1 kHz 16-bit stereo. Every file verified below -0.3 dB true peak after Unity's import re-encode, so nothing clips once imported. Consistent naming, and a README documenting every file.PARTY AND CROWDparty-horn-spooky, cackle-crowd, startled-yelp, comedy-scream, ghost-boo-playfulCOMIC FRIGHTScartoon-boing, slide-whistle-down, thunder-cartoon, jack-in-the-box-pop, creaky-floor-comicDECORATIONS AND PROPSanimatronic-servo, plastic-skeleton-rattle, rubber-bat-squeak, cobweb-string-snap, fog-machine-hissMUSIC AND TRANSITIONSorgan-flourish, organ-sting-spooky, spooky-music-box, theremin-wobble, spooky-whoosh-transitionLICENSERoyalty-free for commercial and non-commercial projects, including games you sell. Use in unlimited projects, no attribution required. You may not resell or redistribute the audio files themselves as a sound library or pack.Audio file types: WAV, uncompressedSample rate: 44.1 kHzBit depth: 16-bitLoopable: The fog machine hiss and music box sustain and loop; the rest are one-shotsAdditional: Stereo. 20 files. Every file verified below -0.3 dB true peak after Unity's import re-encode, scanned for edge clicks and DC offset, named consistently by object and action, and documented in an included README.All 20 sounds were generated with ElevenLabs Sound Effects, a text-to-audio model, under a paid commercial subscription that licenses the generated output for commercial distribution.Each sound started as a specific written prompt describing the object, action and character wanted - for example "a party horn blown once, comic timing, dry, no room". Generation was the starting point rather than the finished product. Every file was then screened and corrected by hand:Screening - each file was checked by ear and by measurement. Files that came back silent, flat, duplicated, or without usable dynamics were discarded and regenerated from a revised prompt.Level work - levels were measured with ffmpeg. Files that clipped were attenuated, and files left far below the pack's ceiling were raised. Sample peak alone was not trusted: every file was re-measured for inter-sample true peak, which is what actually clips on playback. All 20 files are verified below -0.3 dB true peak.Edge work - every file was scanned for hard starts and stops, DC offset and edge clicks, and a short fade was applied where a boundary was audible, with the fade length chosen per file by measurement rather than a fixed value.Curation and documentation - files were named consistently by object and action, grouped into party and crowd, comic frights, decorations and props, and music and transitions, and documented individually in the included README.The store artwork also uses AI: the backdrop is a generated image, over which the title, publisher line, rule and specification line are typeset programmatically.