I ran into this problem in a big way when I posted this. There are many places where the limiting is creating distortion (especially in the right-hand chords), but the worst is at about 6:46. If you start at about 6:30, you'll get the full effect when you reach 6:46 --- OUCH!!!
Slashdot Mirror
The high quality version of the audio will have the 19 (or up to 22.1) kHz sine wave ...
Actually not; it gets filtered out by YouTube's compression algorithm. Here's a demo video with spectrograms showing the audio before upload and as received from YouTube (after compression); you'll see that the sine tone is completely removed (though there are other distortion artifacts that are not).
Okay, here's a video that will give you a quick sense of what the YouTube compression/clipping distortion can sound like and what the effect of the 20KHz sine-tone workaround is; the audio is accompanied by before/after spectrograms showing (a) the "spectral splatter" of the distortion, (b) the filtering out of the 20KHz sine tone during compression, (c) the reduction of the spectral splatter distortion, and (d) other distortion artifacts that are not removed with this technique (they look like a reflection of low-frequency energy, so they're probably Nyquist-related aliasing).
If you want to hear this music in context, it's from this video at around 6:46.
I ran into this problem in a big way when I posted this. There are many places where the limiting is creating distortion (especially in the right-hand chords), but the worst is at about 6:46. If you start at about 6:30, you'll get the full effect when you reach 6:46 --- OUCH!!!
I made a test video that shows that the technique of mixing in a high-frequency tone (I used 20 KHz, in a 48 KHz file) works. The original distorted version is at 0:04 and the same with the distortion eliminated with the addition of the tone is at 0:23.