Fig.1 : Original Test Tones (50-100-250-500-1000-5000-8000-10,000 Hz sine waves at -3db) Linier Fade with 1sec gap.

---

Fig.2 : How those test tones look like after posted on YT

---

Fig.3 : Test Tones encoded by Riva from Mpeg to FLV locally with 320x240, 128kbps

---

Fig.4 : How original sine wave looks like at the attacking point maintaining constant -3db

---

Fig.5 : How attacking point of sine wave was clipped/distorted and boosted by YT over -3db

---

Fig.6 : How original sound should decay in linier fashion smoothly

---

Fig.7 : How YT compression changes decaying behavior compare to original

---

Fig.8 : Compare how two waveforms differ, they're completely different sounds, like between dog and fish!

---

------------Below are Results of Test #2 -----------------

---

Fig.9 : Original Test Tones (440Hz sine wave 5sec blasts with 3db decrement each from 0db to -3db,-6db,-9db,-12db,-15db,-18db,-21db,-24db,-27db with Linier Fade with 1sec gap, and then reversed.

---

Fig.10 : How those test tones look like after posted on YT

---

Fig.11 : Overlay of YT waveforms over original waveforms to examine YT compression behavior.

---