Originally posted by: her34
that leads me to a couple of questions then:
1) how can you skip within a track? if a song is 4 minutes long, how can you skip to 2 minute without scanning through half of the song?
Presumably the same way as on CD. The data stream is packaged into 'frames' (basically another word for sectors). Each frame contains a packet of raw audio data, various checksums and ECC, and an ID so that the player knows which frame it's reading. When you press forward or back, the pickup just gets nudged to one side, so that it picks up the next lap of the spiral 'groove'. Every now and again, the pickup can read a frame, so that you can see/hear roughly where in the track you are.
As for tracks - well, there's a directory right at the beginning of the disc, that the player reads., that tells it at which frame each track starts. When you skip to another track, the pickup guesses roughly where on the disc to go to. It then reads the frame, works out whether its gone too far, or not far enough, and repeatedly refines its guess until it hits the right spot.
2) what if there's no change? since only 1 (up) or 0 (down) is allowed, i imagine that for no change it has to 10101010 sort of thing. isn't that creating false information about the sound?
Well, that's correct. However, the frequency of the bitstream is very high (something like 2.2 MHz for SACD). This means that a 'silent' signal will in fact be rendered as a 1 MHz tone. This will be filtered out by analogue filters, and anyway, it won't get through the any of the amplifier, speaker or your ears.
Remember, pretty much any CD player, sound card, etc. using a modern 'audio DAC' will use a delta-sigma type (1-bit) DAC, producing the bitstream internally. The advantage of this is that they produce their noise at very high frequencies, way away from the frequencies in the original sound. The advantage is that cheap low-performance analogue filters can do an excellent, high-fidelity job at removing the unwanted noise. Conventional DACs produce noise very much closer in frequency to the actual signal, this is difficult to filter and requires high quality analogue circuitry to do so.
3) aren't immediate sound changes inaccurate? with pcm you can go from nothing to 100 with a single frame. with sacd, you would have to take 100 frames because the sound is recorded in one step increments.
That's an important point. However, this is compensated for by having a very high sample rate (2.2 MHz for SACD). As a result, you get 64 samples in SACD in the time you get one sample in CD.
The catch is that not all frequencies get the same resolution. In CD any frequency can be reproduced with 16-bit resolution. In SACD the dynamic range depends on the frequency - if you want to reproduce a 22 kHz tone - the bitstream will be 64 1s then 64 0s, etc. This only gives 6 bits of resolution. Conversely, at 10 Hz (sub-bass), you get 200k samples - which is equivalent to 18 bits of resolution (equivalent to an extra 12 dB of dynamic range).
The result is that SACD has a different noise profile to CD, better dynamic range and better frequency response (up to 100 kHz).
Again, remember that pretty much all CD players/DVD player/sound cards use the same technique - they just convert the conventional samples to a bitstream internally. (You may see labels like x64 or x128 oversampling - x64 oversampling means that 1 sample is converted to a 64 bit stream). The advantage is lower cost, and the better noise profile (the dynamic range and frequency response are limited by the source data).
Facebook
Twitter