Frequency Separation Technique: 8 Bit v 16 Bit

Written by: Glyn Dewis

Published: May 15, 2013

Category: General

Hey Folks,

Thanks for stopping by.

Now, today’s post is kind of a response to a few questions I’ve had about retouching in 8 Bit or 16 Bit.

Ok so here’s the thing…

Up until the last few months I was indeed retouching images in 8 Bit and there were lots of reasons for this…one being file size, especially now shooting with a Nikon D800. Now though everything has been upgraded and anything that hasn’t been is just about to be so things have changed and I do edit in 16 Bit; but…speed isn’t THE reason for the change.

I’m certainly not going to attempt to get all technical here but in it’s simplest terms there’s much more information in a 16 Bit image or to put it another way, you can ‘push’ an image that much more.

For example let’s take this Retouching Technique commonly used in Beauty Retouching called Frequency Separation; on the left hand side you can see the 16 Bit image and on the right hand side the 8 Bit image.

Now now other retouching has been done; all that’s been done is to have prepared each image using the steps in the Frequency Separation Technique and already there’s a difference (hopefully you can see it on your screen) i.e. the 8 Bit image has banding appearing on the dark grey background whereas the 16 Bit image doesn’t.

So in it’s simplest terms…16 Bit images give us a lot more information to work with or to put it another way…they can take a lot more punishment 🙂

If you haven’t seen the Frequency Separation Technique then be sure to check out this video tutorial that I recorded:

As always if you have any questions / comments then feel free t make use of the comments section below or email me direct to glyn@www.glyndewis.com

In the mean time have a great day and I’ll catch you back here tomorrow.
Enjoy,
Glyn

 

You may also like…

Interviewed by Ron Clifford
Interviewed by Ron Clifford

A few weeks back I was invited onto Ron Clifford YouTube LIVE Channel to speak with him about all things photography....

9 Comments

  1. DaveT

    I too find that using 16 bit files really helps to bring out the maximum information from the RAW files.

    I often shoot at dawn and dusk when there are deep shadows and often deep blue skies with differences of depth of colour between the horizon and the top of the image. It also helps with images that are really contrasty where there is a lot of highlight and shadow information to contend with such as interiors of building. Using 16 bit files helps to reduce artifacts when you have to bring out detail in the dark shadows.

    Dave

    Reply
  2. Sebastien Degardin

    So happy to see that it actually makes a difference. I do pretty much everything 16bit but I was starting do doubt …

    Thanks,
    Sebs

    Reply
    • Glyn

      Sebastien…Sure does make a difference especially when pushing for details in the shadows too.

      Reply
      • Glyn

        Absolutely Dave

        Reply
  3. jlua

    Great tip. But one question: I noticed that you had the “Use Legacy Settings” box set on the Brightness & Constrast adjustment dialog. Any particular reason for that?

    Reply
    • Glyn

      Sure thing. For this technique to work you have to put a tick in Use Legacy so that Photoshop uses it’s old algorithms and fairy dust to do it’s magic; if you don’t then you’ll see a difference in the image when you set it up.

      Reply
  4. CJ

    So how do you get your files into 16biy mode? I am assuming when I take a picture it is 8 bit with my canon?

    I open my images in Lightroom first and shoot both raw and jpg.

    Thanks

    Reply
  5. Roberto Palmari

    Great job as usual Glyn,
    just to give a geeky explanation to the banding effect on 8bit processed image just consider that digital images are made of a finite grid of pixels.
    The color of each pixel can be represented by 8 or 16bits.
    Since a bit can be either 0 or 1 you get 2 possibilities per bit, this means that with 8 bits you can represent 2^8 colors, 256 colors.
    With the same principle in 16 bits you can represent 2^16 colors, so 65535 colors.
    That’s why the number of possible shades of colors is much less in 8 bit than in 16 bits.

    Ok I know, that was really geeky and boring, I promise I’ll never do it again. 😉

    Cheers,
    Roberto

    Reply
  6. Troy D. Davidson

    Glyn, been playing around with this a lot! Thanks for the inspiration, and I seemed to have stumbled upon an even quicker way to build the layers. If I use high pass/-50 contrast layer once…all I have to do is copy that layer then control/I to invert the sharpness to smooth…and I never have to second guess that the settings match, because they are the same. No need for any other smoothing/blurring method when this one works out perfectly. Try it and hit me back on FaceBook…I think you’ll enjoy the simplicity. Mahalo to you now, and always, t-

    Reply

Submit a Comment

Your email address will not be published. Required fields are marked *