As for the pixels in Part 1 (first image), I imagined that there should be a sense of covering all coloristic bases, and in this case, I chose a broad selection, not taking specifically from any shade, hue, or tint. As for the background in Part 2 (second image), I thought the deep red of the apple would act harshly on the eyes, so I took from the light that shines off it and added a lighter color that also has a kinder saturation.

From what I remember, Photoshop often utilizes rasterization as part of its transforming program. When I went to transform the pixels into a larger image, Photoshop didn’t account for the colors to be transformed as vectors, and created a compromise between compression and contrast.

These images could be used for advertising of some kind. I’m not completely confident on the type, but perhaps it could add an aesthetic edge to give the intended audience a viewpoint for the product by taking the image into account.

As for creating more imagery from scratch, there are plenty more ways, such as line art, gradient blending, and abstract clipping using the lasso tool. The list for potentials includes, but is not limited to, these techniques.