White Balance

Forums Forums Get Technical Farktography tech talk White Balance

Viewing 8 posts - 16 through 23 (of 23 total)
  • Author
  • #43889

    If you convert the raw NEF files to 16-bit PNG you wind up with a format any image manipulation program will read without the loss of dynamic range inherent in 8-bit JPEG’s. It’s also a lossless format, unlike JPEG, so you don’t get any extra distortion on top of stepping down the dynamic range. The files are big though, so that’s the tradeoff. And I’ll toss out the “disk is cheap” cliche before someone else does…

    And if you keep the raw files around (disk is cheap remember?) then you can always go back and re-process them to PNG again if you decide you want to tweak things like white balance in the raw image before converting. Basically, you have an archive of the best quality image around you can use as a source.


    You will be shocked how much you can recover from a RAW file, shots you were forced to toss in jpeg because you shot fast and didn’t adjust exposure enough now become completely recoverable. When I am shooting street work I just Sunny 16 and shoot in RAW, so even if I stepped into the shade and shot from the hip without checking exposure, I still have a usable shot.

    I highly encourage you to pick up Adobe Lightroom, which is both a catalog and a power editor. It runs around 200 for purchase and or…err…less if you don’t mind sacrificing your ethics.

    You can shoot me an email/PM if your ethics are feeling flexible.


    The biggest advantage to RAW and its additional color depth is the amount of levels stretching that can be done while still maintaining smooth gradients across the dynamic range. As an example, take an 8-bit image (24bit color) and pull up the levels dialog in your editor of choice and move the gamma slider about half way or more towards the left. This brightens up the shadows quite a bit, and recovers a lot of image data. Close the levels dialog and re-open it then look at the histogram. The left side of the histogram now looks like the teeth in a comb or similar, where the gaps represent now-missing intensity information; ie you now have the original 128 possible values below the midpoint being occuppied by only 64 or 32 (or some arbitrary lower number depending on how far you moved the gamma). Those gaps lead to sharp changes in color where the remaining values meet in the image and you wind up with an image that has obvious, visible banding or posterization.

    For each one of the intensity values in an 8-bit image, a 16-bit image (48-bit color) doesn’t have just twice the number of possibilities, but 256 possible values. This allows for some very aggressive levels tweaking without introducing the gaps in the histogram. Things that would destroy an 8bit image still look smooth with a 16bit image. White Balance, EV, gamma, contrast (which ties in with sharpness), HDR and any other level adjustments all depend on color depth information so the more, the better.

    The catch is obviously software… GIMP doesn’t handle 16bit natively (although it’s in the works), but PS users have the ability to do full editing in 16bits/channel. If you’re using open-source, KRITA wil do basic levels and adjustments in 16bit, but doesn’t have the editing features of a full-blown image editor like PS or GIMP. I do levels tweaks and corrections in KRITA after running my images thru whatever 16bit CLI processes I need then export to GIMP for final editing if needed or *minor* level corrections.


    I’ll second the investment in Lightroom. Holy crap, it’s amazing. I’m now slowly working on going back and reprocessing this year’s pictures because I can actually save some that were unrecoverable before.

    Unfortunately I don’t know anything about the catalog feature, how to use it, etc. and it’s not very bonehead friendly. The processing part is very friendly. And all the plugins you can get for it…


    It’s been said so many times, but ditto on the RAW shooting. I’ve thought about getting a grey card, but it’d be something more for portrait work, or if you know you’re going to have the exact same lighting conditions for many photos. If I’m walking around just exploring, there’s no way I’ll take the time to set up a grey card shot every time I walk through shade.

    Shoot RAW, have big memory sticks, and keep empty spares just in case. I’ll also second the “disc is cheap” – I never delete or edit my original RAW files. I’ll store the originals in a safety file, then copy them over into another folder to play with the copies before converting them to a webpage friendly format. You can always go back and change things around if you wish, and you never lose any quality.


    I never delete or edit my original RAW files. I’ll store the originals in a safety file, then copy them over into another folder to play with the copies before converting them to a webpage friendly format.

    You probably don’t have to worry about editing the originals, at least when using Canon’s DPP on a RAW file. I don’t remember if you’re a Canonite or a Nikonian, but I’d guess that this also applies to Nikons and other software. DPP stores the changes separate from the actual RAW data (what they call a “recipe”) so that you can always go back to SOOC with just a ctrl-shift-Z. You can even store the recipe separate from the RAW file itself if desired so that it can be applied to multiple files.


    Nikonian here, but I’m still paranoid about some things like that 😉

Viewing 8 posts - 16 through 23 (of 23 total)
  • The topic ‘White Balance’ is closed to new replies.