Section with miscellanies
Random stuff or quotations from below waiting to be added to relevant sections above with added authors and corrected. Possible duplicates. A mess, but potentially useful. Some stuff might be outdated. Random notes from our Discord server. Not working hyperlinks yet.
(added from the newest info to the oldest, currently 143 pages)
It didn’t fit the GDoc limit, so it was moved here.
Hall of fame
Great thanks to - Anjok, (creator of UVR), Aufr33 (UVR contributor - VR multiband implementation, phase fixer discovery, model trainer, uvronline owner), KimberleyJSN a.k.a. Kim (model contributor and MDX/Roformers support), viperx (our former heavy user, supporter and now private models creator), tsurumeso (the creator of VR arch base code), BoskanDilan (creator of the old UVR GUI), IELab a.k.a Kuielab & Woosung Choi (MDX-Net arch creators), ZFTurbo (creator of MVSEP, MDX23, and tons of models and training advice), GAudio (GSEP creators), Alexandre Deffosez a.k.a. Adefossez (Demucs creator), Bytedance with asriver (creators of Roformer arch and Ripple & Capcut app), lucidrains (for recreating the BS and Mel-Roformer code from released papers), jarredou (MDX23 fork, drumsep model, tons of support and work on Colabs), Bas Curtiz (model trainer, insane amount of testing and UVR5 settings guidance, tutorials with SDR evaluating, models creator), Captain FLAM (KaraFan creator), unwa (for his Roformer fine-tunes and training advice), Gabox (-||-), becruily (model trainer, tons of advice, phase fixer public implementation), Anvuew (a lot of de-reverb and vocal models), FoxyJoy (de-reverb, de-echo, denoise models), gilliaan (for currently the best phantom center models), mesk (metal models and training guide), jazzpear/Jasper (tons of SFX models), Aname-Tommy (a few of successful model fine-tunes) - thanks to all of these people for the best freely available AI separation technologies and models so far.
Special thanks for users of our Discord:
Not Eddy (UVR UI, Colabs, HF, KF fork), seiso/makidanyee/listra92 (for making our inference Colab great with model queuing), Sir Joseph (WebUI Colabs), SqA5uid/squiddagod (maintaining inference Colab fork), HV (older MDX and VR Colabs creator and UVR contributor), txmutt (Demucs Colab), CyberWaifu (lots of testing, some older Colabs), KoD/Mixmasher (first HV’s MDX Colab fork), dca100fb1/dca100fb8 (VR ppr bug, finding tons of UVR bugs and models testing, and feedback, titanic work on ensembles and tricky/AI-killing songs lists), isling (a lot of testing and suggestions), CyPha-SaRin (lots of models/UVR testing), BubbleG, ᗩรρเєг, Joe, santilli, RC, Matteoki (a.k.a. Albacore Tuna, our audio upscaling guru), Syrkov, ryrycd, Mikeyyyyy/K-Kop Filters, Mr. Crusty ᶜʳᵃᵇ (our mod; compensation values finding, MDX Colab mods and testing), knock (ZF’s MDX23 fine-tuning), A5 (lots of feedback on existing models), Infisrael (MDX23 guide and model testing), Pashahlis/ai_characters (WhisperX guide and script), Sam Hocking (our most valuable pro sound engineer and industry insider), Kubinka (for his Colabs and coding help), Vinctekan (one of our most valuable sound engineers and tools creator), CC Karaoke, vernight (both lots of testing and advice), hendry.setiadi, raiboomdash (lots of model tests with vast descriptions), wancitte, essid64, 5b, mohammedmehditber (tons of troubleshooting on 2GB GPUs, some tips and tricks, models instructions, feedback), dynamic64 (lots of model descriptions and testing).
- thanks to all of these people - for knowledge, help, testing and everyone whose advice, quotes and stuff appear in this doc. This guide wouldn't be created without you. If I forgot someone, forgive me.
__________________________________