Raster Images & Resizing

There are only two ways to resize a raster image. Change the pixel size (effective resolution) Change the number of pixels (resample)

Raster Image Resampling

Photoshop™ can resample images. (change amount of pixels)

Resampling renders new pixels from old, by interpolating original pixel data.

Raster Image Resampling

Beware! All of the pixels are NEW. They are not the same as the original capture. “Artifacts” can be created by the

resampling/rendering process.

Raster Image Resampling

Unwanted colors or effects may result from resampling

Raster Image Resampling(change the number of pixels)

Resampling Upsampling

Add pixels to increase resolution Downsampling

Delete pixels to decrease resolution Change size of an image and maintain the

original resolution

Raster Image Resampling

Downsampling Good for reducing the file size of a large image

Upsampling can increase the resolution of an image with resampling. Can reduce jagginess in CT images Trade-off is image sharpness

It will not help BIG enlargements.

Raster Image ResamplingUpsampling

No Upsampling Upsampled to 72 ppi

72 ppi 72 ppi

Resampling - Line Art

It cannot make a jaggie bitmap image smooth!

Capture at a high resolution initially

Line Art

100 PPI 200 PPI

Rasters & Rotation

Rotating a raster image results in a different pixel alignment caused by how the rotated image realigns to the image grid (raster).

Rasters & Rotation

Rotation triggers resampling.

Straight edges can become jagged.

Interpolation calculates the newly created pixels.

Original Image Location

Rotated Image New Pixel Location

S

Bit DepthRaster Image Characteristic

Raster Image Bit Depth

The pixels will either represent black & white, shades of gray, or colors.

The data collected for each pixel is stored as bits of computer data.

A bit is the basic component of all computer data.

B&W - Grays

Black & white pixels are stored with one bit of data per pixel.

The bit is either a 0 OR a 1.

0 1BlackWhite

B&W - Grays

More than one bit is needed for storing shades of

gray.

Two bits of data per pixel stores four shades of gray (black, dark gray, light gray, white).

B&W - Grays

Original PhotographGrayscale

Four Levels of Gray =

2 Bits of data per pixel

B&W - Grays

The number of bits per pixel as an exponent of 2 equals gray levels.

Normal GrayscaleIs 8 Bits/Pixel

2Number of bits

21 = 2 levels22 = 4 levels23 = 8 levels24 = 16 level28 = 256 levels

Digital Values

With 8 bits, 256 different values (grays or colors) are possible.

We can count from 0 to 255.

0 = Black or no color

255 = White or maximum color

Digital Color

Scanning red (R), green (G), and blue (B), a value is captured for each pixel.

Each R, G, and B value is captured and stored in 8 bits.

24 Bit = 16.7 million colors

Digital Color

Four Color Process is 32 bit

Does NOT reproduce as 400 million colors

Digital Color

16 Bit Images - High-Bit 48 Bit Color (16 bits per color channel RGB) Camera Raw .CRW For editing only - Export at 24 Bit

Digital Color

Indexed Color 8 bit - 256 Colors (out of 16.7million) Applied to GIF format PNG-8 Supports indexed transparency

16 x 16 = 256 Color Index

File Compression

Large raster files often require compression

Two basic types of compression Lossless - no data is lost in compression and

extraction LZW (TIFF, GIF) ZIP RLE (for line art)

"Lossy" - data is lost to achieve greater compression JPEG – compress once MPEG and other video CODECs