This time, I used ARA, which means I sang once, digitized it, and converted it into the voice of Sakurano Sora. This method makes it easy to produce a realistic nuance without complicated programming. It’s the evolution of the times. I was considering whether to use a de-esser, but if I tidy it up too much, it feels mechanical, so I deliberately didn’t use it.
For the accompaniment, I used Avenger2 for the kick and bass. If the tone is too heavy, it lacks the sharpness of the current times and makes you feel the age in a bad way, so I avoided it with rearrangement and tone selection. The illustration is by Dall-E3. I ordered this and that from AI, and at first, only pictures that were completely different from the image were made, so I asked for corrections many times. Once it becomes as per the image, it is helpful to draw many variations.