Well, as I’ve been using various programs, it is pretty clear that you have to be pretty careful about using the different programs. Some things you want to do earlier and some later. For instance, you probably want to sharpen last as this really changes the bits. “Larry Gerbrandt”:http://photos.larrygerbrandt.com/2007/07/24/hello-world/ has some recommendations on workflow that are in short:
# Shoot in RAW and JPG. This produces NEF for Nikon or CR2 for Canon. It usually contains one stop more information. I actually typically shoot both since many times I just need a photo quick and don’t have time to process it. Technically speaking, with JPEG you have 8-bits of information per color (24 bits total), while with RAW in most cameras you ahve 12-bits or with later cameras even 14-bits of information. That’s way more.
# Use DxO first on the raw image. This correct distortion and vignetting, so it should be the same across a group of images. Don’t use this program for sharpening and so forth. It produces TIFF and DNR files which are Adobe device-independent raw files.
# Photomatix. This takes multiple bracketed images (usually these should be at least +2 and -2EV, but correctly done, should be bracketed so the brightest shot is all the way to the right, really overexposed and the darkest shot is all the way to the left in the histogram all taken aperture priority to ensure the same focus). This produces .EXR files which are the raw high dynamic range and also .TIF files. You have to do it in this order because for some reason TIF files Photomatix produces, doesn’t have enough information for DxO to match the lense with it
# Noise Ninja. This is the last thing you use, it should be done in Photoshop because Noise Ninja standalone only read uncompressed TIF and TIFs are big enough.
