In this paper, we present a fast end to end deep convolution network to map a low resolution color image to a higher resolution color image. Furthermore, a comparison of how using different loss measurements such as mean squared error effect the outcome of the network. Along with that a discussion of how slicing the image in various ways can effect the overall quality of the output image.