Monday, February 18, 2013

256 Shades of Gray: A Ribald Tale of Bit Depth


Who knew working with image files could be so titillating? No one knew, because it’s not. But image files can have more than 50 shades of gray. They can have up to 256…and trillions of colors Now that’s exciting!

Before you start working with color, you should start thinking in black & white, or more specifically, grayscale. No matter what color mode or bit depth you choose, keep in mind that you are dealing with one channel for each color. Each of those channels consists of different grays. Keeping that in mind, if you focus on controlling the levels and dynamics of those grays in each channel, your final image will look the best it can for the intended design. More on this later.

This posting focuses on selecting which Color Mode & Bit Depth to choose when creating a file in Adobe Photoshop. 

Following this posting I will have additional postings focusing on the individual color modes in Photoshop.


THE COLOR MODES

When should you choose which mode when starting a new Photoshop document?

When you start a brand-spanking new document in Photoshop you are given the choice of the following Color Modes:
Bitmap: When you are only printing one solid color. It can be black, white or any single ink or spot color.
Grayscale: When you are only printing one solid color, but want tints or screens of that color. The color can be black, white or any single ink or spot color. But check with your print vendor to see how they handle whites. Not all printers are equal when it comes to printing white.
RGB Color: RGB is Red Green Blue, of course. It is the color space that we humans perceive. It is the color of light. Choose RGB if you want access to all of the filters in Photoshop. Definitely choose RGB if you will only be creating art for digital/on-screen viewing. But also choose RGB if you are printing to a large format printer, printing on a digital press, printing to an inkjet printer or printing in the Hexachrome (six-color CMYKOG) format. You may need to convert to CMYK before print and will deal with some color shifts (depending on how far out in the RGB spectrum some of your colors go). If your image has bright reds, blues and greens - use RGB.
CMYK Color: Cyan Magenta Yellow Black are the standard process printing colors. You can look at CMYK as the universal stock color mode. You can pretty much guarantee that any print vendor you select for your print job will have a four color press that is set up to run CMYK. Most print vendors will have 6, 7 or 8 color print presses which will allow you to add spot colors and varnishes…at a cost. If you’re not concerned with a wide range of bright colors - you are safe to choose CMYK from the start.
Lab Color: ‘Lab’ stands for ‘Lightness’, ‘a’, ‘b’ - which refer to the three channels available in the file. (The ‘L’ in Lab is sometimes referred to as ‘Luminosity’.) Lab color describes all of the colors that a person with normal vision perceives. This mode is intended to give you options to get the most out of the contrast and colors in an image. With the Lightness channel you adjust the grays on a 0 to 100 Scale. With the ‘a’ channel you adjust based on a Yellow to Blue scale. With the ‘b’ channel you adjust based on a Magenta to Green scale. The Lab Color Mode gives you the biggest color gamut. The best way to get familiar with the Lab mode is to work with an image which has wide varying contrasts and colors. It’s effects are not needed for all images, but you may be surprised how Lab really makes some images sing. Give it a try. You may be pleased with the results.

BIT DEPTH

Bit Depth Choices in Adobe Photoshop
Along with the color mode choices, when you start a new document in Photoshop, you have some bit counts from which to choose. The bit depth choices per Color Mode are as follows:
Bitmap: 1 bit
Grayscale: 8 bit, 16 bit, 32 bit
RGB Color: 8 bit, 16 bit, 32 bit
CMYK Color: 8 bit, 16 bit
Lab Color: 8 bit, 16 bit

The long and short of it is - The more bit depth, the more colors per byte, the more an image approaches the range of color and contrast the human eye can perceive.

So which Bit Depth do you choose? 
If you are creating a job for print, the short answer is - 8 Bit.
If you are creating a job for screen or web, the answer is - 16 Bit.
If you are creating a job for both print and screen, the answer is - 16 Bit with conversion to 8 Bit before releasing for print.

When, if at all, should you choose 32 Bit? 
The goal with 32 bit (HDR - high dynamic range) images is to come close to or simulate what the human eye sees. The human eye can perceive a much higher range of colors and contrast than a camera can. 32 bit gives you much more colors per digital byte with which to work. Some 32 bit images, on screen, can look almost freakishly unreal. But if you take the time to look around your world, when you are out and about, you will realize that you see many colors and contrasts, but you may choose to ignore them simply because you really don’t need to take all of that in to get through your day. The goal of 32 bit is to try to capture those real world colors and contrasts.

With any of the choices of Color Modes or Bit Depths available in Photoshop you have to consider, not just your design intent, but your source image. Don’t bother using 32 bit if you are generating a custom graphic which you create directly in Photoshop. If your source image has been shot in 32 bit, then try to keep it in 32 bit, or convert to 16 bit, for the editing process. You can then convert a copy to 16 bit or 8 bit after your final edits.

More choices of filters with 8 Bit compared to 16 Bit.
With 8 bit you will have more choices of filters to use on your image and any print vendor will be able to process an 8 bit image. But if your design includes photographic images which need to stand out using dynamic contrast and color, use 16 bit for the retouching and editing process. You can then convert to 8 bit to apply any additional filters and to send out the final file for printing. If you have direct access to the print vendor, ask them if they can handle 16 bit images. You may be able to release as a 16 bit file, if they can handle it.

Do you need to understand the math behind the Bits?
NO!
You can, if you want, but most explanations of bit depth that I’ve read go off on (what seem to me) to be scientific explanations geared towards those who have accomplished their math doctorates. Most of us who work in graphic design and production have no time or brain bandwidth to decipher these explanations. So I won’t try to explain too much of the math behind bits, bytes and binaries. I have a little of the math behind the bits at the end of this posting...for the morbidly curious. 

All you really need to know about bits, as I said above, is:
The more bits, the more colors per byte, the more an image approaches the range of color and contrast the human eye can perceive.


So where do the 256 Shades of Gray come in?

Each channel in a Photoshop file can be comprised of black, white or grays. You then combine or blend those channels (or plates - if your thinking along the lines of printing). Each channel is then assigned a different single color. Such as red, green, blue, cyan, magenta, yellow, black, spot PMS, etc. With Lab you work with grays or color ranges for the ‘a’ and ‘b’ channels. If you’re printing using spot colors you can add additional channels (plates) or swap out the standard RGB or CMYK channels. If you are working with 16 or 32 bit, you have more levels of grays between the white and black. With 1 bit, you just have black and white.

So when you think of color, think of the range or levels of grays that go into each channel. This is what really gives and image it’s dynamics.

THE BINARY SYSTEM, BITS & BYTES

First let’s quickly explain the basics of the binary system. The binary system is all the ‘0’s and ‘1’s you hear about. The ‘0‘ represents ‘black’. The ‘1’ represents ‘white’.

Bits

A ‘bit’ is that 0 or 1.

Bytes

A byte is 8 bits. It can represents 256 colors. That is arrived at by calculating 2 to the 8 power (2x2x2x2x2x2x2x2). That is to say - 1x2=2, 2x2=4, 2x4=8, 2x8=16, 2x16=32, 2x32=64, 2x64=128, 2x128=256. The more bits you have, the more these calculations expand to the higher dynamic range of color perception.




No comments:

Post a Comment