Bit depth and preferences

Bit depth specifies how much color information is available for each pixel in an image. More bits of information per pixel result in more available colors and more accurate color representation in an image, thereby affecting the image size. Simply put, the file size of an image increases with bit depth because more color information is stored per pixel in an image having higher bit depth.

For example:

• An image with a bit depth of 1 has pixels with two possible values: black and white. 
• An image with a bit depth of 8 has 2⁸, or 256, possible values. 
• Grayscale mode images with a bit depth of 8 have 256 possible gray values. 
• RGB mode images are made of three color channels. An 8‑bit per pixel RGB image has 256 possible values for each channel which means it has over 16 million possible color values. 

RGB images with 8‑bits per channel (Bits/Channel or bpc) are sometimes called 24‑bit images (8 bits x 3 channels = 24 bits of data for each pixel). In addition to 8 Bits/Channel images, Photoshop can also work with images that contain 16 or 32 Bits/Channel.. Images with 32 Bits/Channel are also known as high dynamic range (HDR) images.

To convert your bit preference, do any of the following:

• To convert between 8 Bits/Channel and 16 Bits/Channel, choose Image > Mode > 16 Bits/Channel or 8 Bits/Channel.

• To convert from 8 or 16 Bits/Channel to 32 Bits/Channel, choose Image > Mode > 32 Bits/Channel.

Photoshop provides the following support for working with 16 Bits/Channel images:

• Work in Grayscale, RGB Color, CMYK Color, Lab Color, and Multichannel modes.

• Use all tools in the toolbox, except the Art History Brush tool with 16 Bits/Channel images.

• Apply color and tonal adjustment commands.

• Work with layers, including adjustment layers, in 16 Bits/Channel images.

• Apply numerous Photoshop filters that can be used with 16 Bits/Channel images.

  To take advantage of certain Photoshop features, such as some filters, you can convert a 16 Bits/Channel image to an 8 Bits/Channel image. It’s best to do a Save As and convert a copy of the image file so the original file retains the full 16 Bits/Channel image data.