Text this: High-fidelity single-frame computational super-resolution using signal-preserving denoising-enabled deconvolution.