The model is trained to produce natural noise, camera grain, motion blur, and a realistic photo look. Please set the adapter weight/strength between 0.6 and 0.8, as this range yields the best results. The trigger word to place at the beginning of the prompt is 'badqual'.