Editing Data Compression (section)

== <span style="color: #FFFFFF;">Understanding</span> ==
Data compression is understood through '''Redundancy Elimination''' and '''Perceptual Thresholds'''.

'''1. The Fight Against Redundancy (Lossless)''':
Most data is very repetitive.
* '''Pattern Recognition''': If a text says "The" 1,000 times, the computer doesn't need to store "T-h-e" 1,000 times. It stores "The" once and gives it a tiny "Shortcut" code.
* '''Statistical Probabilities''': Huffman coding uses the fact that some symbols happen more than others. By giving the most common ones the shortest codes, the average size of the message drops.

'''2. The Human Cheat (Lossy)''':
Our eyes and ears are imperfect.
* '''JPEG''': Your eye is great at seeing brightness but bad at seeing small changes in color. JPEG throws away 90% of the color data and your brain "Fills it in."
* '''MP3''': Uses "Acoustic Masking." If there is a loud drum and a quiet flute at the same time, you can't hear the flute anyway. MP3 throws the flute data away.

'''3. The Shannon Limit''': No matter how smart your algorithm is, you can never compress a file smaller than its "Entropy" (the pure randomness inside) without losing information.

'''Artifacts''': When you compress something too much (especially with lossy methods), you start to see "Blocks" in a video or "Blur" in a photo. these are called compression artifacts.
</div>

<div style="background-color: #8B0000; color: #FFFFFF; padding: 20px; border-radius: 8px; margin-bottom: 15px;">