Held: Wednesday, 18 September 2013

Back to Outline 11 - A Design Perspective. On to Outline 13 - Pause for Breath.

Summary

We consider the basics of raster graphics, a common technique for describing and representing images. We also consider RGB colors, one of the most common mechanisms for representing colors digitally.

Related Pages

• Assignment: Exam 1: Scheme Basics
• Reading: Raster Graphics: Images from Pixels and Colors
• Reading: RGB Colors
• Lab: Raster Graphics and RGB Colors
• EBoard (Source) (PDF)

Overview

• Representing images, revisited.
• Pixels and colors: The basics.
• RGB colors.
• Those weird numbers.

Administrivia

• We seem to have a misunderstanding about workload. If you're spending more than five hours on a homework assignment, something has gone wrong and we should talk.
  
  • Preferably, you would realize that you were going to spend that much time before, say, 8pm on Tuesday night.
  • If you hit more than five hours, it's making sense, and you really want to do the last bit, that's fine.
  • But if you're stuck, you should come to see me.
• Lab writeup: Extra 1 from today's lab, due Friday before class.
• Reading for Friday: Transforming RGB Colors
• It sounds like there may be room for one more person in Dr. Davis's 11:00 a.m. section of 151. So, if you'd like to switch sections, let her know asap.
• I plan to email you your current grade info by Friday.
• SACNAS Ice Cream Social Thursday, September 19th at 8:00PM in the Biology Commons
• EC Opportunities:
  • CS Extras Thursday @ 4:30: Kim Spasaro on Linguistics Programming
  • CS Table Friday (Pair programming)
  • Other?
• I also encourage you to attend any of the other Title IX talks/events.
• We're going to try an experiment today. Instead of giving my start-of-class lecture and clicker questions, I'm going to let you ask about five questions on the readings.
• Exam 1 distributed and quickly discussed.
  • There will be time at the start of class for questions on Friday, Monday, and Tuesday.
  • I'll also add some Q&A to the exam itself. Please check it regularly.

Raster Graphics

• Yet another model of images.
• This one corresponds closely to how computers actually represent images.
• The image is a grid of colors.
• You already know about the numbering of the grid points:
  
  • Rows and columns numbered starting at 0
  • Rows numbered from top to bottom
  • Columns number from left to right
• It's painful to have to set individual pixels, but it can be useful.
  • At some point "under the hood", most programs eventually deal with individual pixels.

Colors

• Many ways to represent colors, too.
• Goals: Unambiguous, fast to process, compact
• The color names we've been using are
  • Ambiguous
  • Slow to process
  • Long
• Whoops!

RGB Colors

• The most common internal representation of colors on computers.
• We think of a color as the combination of three primaries: red, green, and blue.
  • These are the primaries for the so-called additive colors
  • You are probably used to the primaries being red, yellow, and blue, but those are the subtractive colors
• On computers, we represent each component as a number between 0 and 255, inclusive.
• It turns out that you can shove four numbers, each between 0 and 255, into the internal representation of an integer.
  • So, the red, green, and blue components are three of those numbers.
  • What's the forth? In multi-layer images, it's the alpha channel.
• We use rgb-new to create these colors.
• We use rgb-red, rgb-green, and rgb-blue to extract the corresponding components.

Those Weird Color Numbers

• We're representing colors as three integers, each in the range [0..255].
• Yet when we ask MediaScript for a color, we tend to get a single integer which is rarely in that range?
• What's going on?
• Well, most computers have different techniques for representing really small integers (particularly integers in the range 0 to 255, inclusive) and standard-length integers. We're using the smaller representation, but shoving three of those into a standard-length integer.
  • We use that representation because it makes everything faster.
• Even those of us who designed and implemented that representation can't read the values presented in that representation.