Flux Verision: Trained on photos from QuickTake 150 cameras. The lora works best at around .80 strength. The highest resolution for the QuickTake 150 was 640x480, so for the most accurate look, 640x480 is the best resolution to generate at. Above that or upscaling seems to clean up the compression and details.
Qwen-Image Version: Works best with a CFG of 3-3.5. Trained in AI-Toolkit using the 3 bit with ARA method.