This is the identical shot to the first 96th St. at Night photo, but post-processed in Photoshop differently.
For this version I converted the RGB colour file to Lab colour and processed the luminance channel only by expanding the range of luminosity levels to the full 0 to 255 possible distribution. This has the same kind of effect as increasing the contrast would, except that whereas changing the contrast discards some information, all information is preserved. Additionally, changing the levels of RGB proportionately results in a perceived colour-shift since the human eye is not equally sensitive to those three colours. By converting to Lab and using the luminance channel the perceptual colour information remains constant.
I also darkened the image in the other file by decreasing the brightness. Again, this discards information. I achieved a similar effect without information loss for this image by altering the gamma of the luminance of the image instead.
I also considered using an unsharp mask on this image but didn't like the result since a misty night picture seemed more realistic without sharpening. This image was saved at an identical size and jpeg quality as the first version.
I learned about these new (for me) processing techniques from reading the Usenet news group rec.photo.digital and by studying various instructional websites described by people in the newsgroup. Very helpful... high signal to noise.