Discrete cosine transform
A discrete cosine transform is a math process that can be used to make things like MP3s, and JPEGs smaller. It does this by breaking the sound or picture into different frequencies.
One way to calculate a discrete cosine transform is to use the Fourier transformation. "Discrete" means that it works on discrete-time signals (sampled data).
Sounds
For sounds, frequencies are the same as simple tones. You can make any sound by playing several tones at the same time. The discrete cosine transform is a way to find out which tones to play in order to make a given sound. The only difference between the tones is their pitch.
Human ears are good at hearing low pitches, but bad at hearing high pitches. If you use the DCT to break a sound into tones, you do not need to be as careful when playing the higher tones because people cannot hear them as well. MP3 encoders (programs which make MP3s) use this fact to make the sound smaller.
Pictures
With pictures, frequencies are like stripes and checker patterns. Just like with sound, you can make any picture out of different frequencies. Also like sound, people's eyes are not as good at seeing high frequencies (hard edges and fine grain) as low frequencies (solid colors and smooth shades). JPEG encoders use this fact almost exactly the same way as MP3 encoders do.
Discrete Cosine Transform Media
The basic stages of computing VR DIF Algorithm are given in the block diagram
The single butterfly stage of VR DIF algorithm is given.
Two-dimensional DCT frequencies from the JPEG DCT
An example showing eight different filters applied to a test image (top left) by multiplying its DCT spectrum (top right) with each filter.
Original size, scaled 10x (nearest neighbor), scaled 10x (bilinear).
On the left is the final image. In the middle is the weighted function (multiplied by a coefficient) which is added to the final image. On the right is the current function and corresponding coefficient. Images are scaled (using bilinear interpolation) by factor 10×.
