Bogner Goldfinger 45 Experiment

This capture set showcases a few different training architectures used on output clips which were generated with TTSv10 and 3D test signals, which should lead to less aliasing and slightly better accuracy. Only one settings was used per channel (Alpha The nomenclature is structured as follows: A = Alpha (Clean) Channel Fender Blackfaceesque O = Omega (Crunch) Channel Marshall Plexiesque V10 = Based on the TTSv10 input file 3D = Based on the 3D input file C = Complex Architecture (My usual Hyper Accuracy Training) C16 = Slightly Less Complex Architecture (Better CPU Efficiency) S = Standard Architecture (Tone3000 Default) XS = xSTD Architecture (Andrey's Standard Architecture) All the complex captures were trained with 3000 epochs and some of the standard captures were trained on 3000 and 6000 epochs. The omega channel standard captures were not trained locally but on tone3000 with 1000 epochs. Signal Chain: UA Apollo X8 -> Lehle P-Split III Reamp Box -> Bogner Goldfinger 45 -> Suhr Reactive Load Signal Chain Calibration: +12.4 dBu (Calibration Metadata Included)