Even though we ARE Nikon lovers,we are NOT affiliated with Nikon Corp. in any way.

English German French

Sign up Login
Home Forums Articles Galleries Recent Photos Contest Help Search News Workshops Shop Upgrade Membership Recommended
members
All members Wiki Contests Vouchers Apps Newsletter THE NIKONIAN™ Magazines Podcasts Fundraising
kmh

Los Alamos, US
134 posts

Click to send email to this author Click to send private message to this author

"RE: Scanner Math"

kmh Silver Member Nikonian since 04th May 2008
Wed 17-Oct-12 06:51 PM | edited Thu 18-Oct-12 04:00 PM by kmh

Here a few comments concerning the number of bits used in a digital image. In a JPEG image, the intensity of each color channel is quantified by an 8-bit number. That means that a color intensity can have values between 0 and 255. Only integer values are possible, so a value of 128.3 is not allowed. In other words, the intensities are quantized; the minimal increment in intensity is one. As a result, a JPEG image of a continuous ramp in intensities is actually a step function with at most 256 steps.

The effect of using a limited number of bits to represent an image is shown in the following two figures:

Click on image to view larger version

8-bit per channel image


Click on image to view larger version

3-bit per channel image


The first figure shows a portion of a JPEG photo, which uses 8 bits per channel or 24 bits per pixel. At this intensity resolution, the figure appears to be a continuous image. The second figure shows the same scene, but using only 3 bits (corresponding to eight intensity levels) per channel.

In the second figure only specific colors are displayed. Thus, a scene with smooth variations in tone and color produces abrupt jumps. These steps are most easily seen near the top of the image where the cranes in the background are severely blurred. This effect is often referred to as posterization or contouring.

The same effect occurs for the preceding 8-bit image, but the jumps in tone are smaller than we can detect with our eyes, so smooth variations in tone appear continuous. That is why an 8-bit per channel JPEG protocol is acceptable for printing digital images.

However, using an 8-bit representation in an image-editing program, such as Photoshop, can have unfortunate consequences. The reason is that each image manipulation, such as things as simple as brightness and contrast adjustments, can transform integer steps in intensity to non-integer steps in the resulting image. When converted back into an 8-bit per channel image, the result no longer corresponds to a good approximation of the intended continuous image. When one makes numerous adjustments an 8-bit per channel image, the resulting image can be severely degraded.

The bottom line here is that one should try to work with images with more than 8 bits per color channel. That applies to image capture and editing. However, 8 bits per channel are acceptable when you print an image.

Hope this helps,
Ken

Cheers,
Ken Hanson

My gallery

A topic tagged as in need of help Scanner Math [View all] , Bug Gold Member , Wed 10-Oct-12 01:58 AM
Subject
ID
Reply message RE: Scanner Math
1
Reply message RE: Scanner Math
3
Reply message RE: Scanner Math
2
Reply message RE: Scanner Math
4
Reply message RE: Scanner Math
5
Reply message RE: Scanner Math
7
     Reply message RE: Scanner Math
10
          Reply message RE: Scanner Math
12
Reply message RE: Scanner Math
6
Reply message RE: Scanner Math
8
     Reply message RE: Scanner Math
9
Reply message RE: Scanner Math
11
Reply message RE: Scanner Math
13
Reply message RE: Scanner Math
14