Slashdot Mirror
The Rise of "Hybrid" Vinyl-MP3s
Khyber writes to let us know that First Word Records, a U.K.-based record label, is now selling vinyl records that come with codes that allow you to download a 320-kbit MP3 of that record's content. The article mentions another independent label, Saddle Creek, that also offers DRM-free downloads with some vinyl records. The co-founder of First Word is quoted on why they didn't DRM the download: "Making a legal, paid-for version of the file less useful than a copied or pirated one doesn't make sense."
I'm not an audio engineer but from a telecom course I took the basic idea is that you sample at twice the highest frequency (IE. 20kHz frequency would require 40k samples a second).
For the most part humans focus on the 300Hz-3.3kHz range which is why the phone companies only give you about 3k Bandwidth and sample at about 8k samples a second over POTS.
09F911029D74E35BD84156C5635688C0
+2 Troll is Slashdot's way of saying groupthink is confused
Yes. Nyquist. There's a Theorem, a Limit, and a Guy that discovered these, all by the same name.
A 44.100kHz sample rate will theoretically get you up to a 22.050kHz max frequency in the audio signal. Humans can focus on any part of the audible spectrum, but voices won't typically fall outside the 300-3300 Hz range. Thus aLaw (US) and mu-Law (outside the US, a.k.a. "uLaw", since the Greek mu looks like a u with a tail) are typically 8000 Hz sample rate, 8-bit-sample, monophonic (who has a stereo telephone?) signal when digitized.
The GP was worried that the bit depth is "coarse". This is not the case. Bit depth "distance" is constant for a given depth.
CD's are 44.1kHz, 16-bit, stereo. Always. So there are always 44100 samples per second per channel. There are always two channels (stereo, one left, one right). And each sample in each channel is always 16 bits. A 16-bit integer can represent numbers from 0-65535 (2^0-1 through 2^16-1), and since there's no need for negative numbers (this is Pulse Code Modulation, or PCM, so no, you don't need to represent a +/- of a waveform) you get the full 0-65535 swing. From there, the value is directly translatable into a DC voltage that goes to the speakers. (Most of the heavy lifting is done in the A/D phase, D/A phase is a simple value-to-DC conversion.) The change in DC voltage over time is what causes the magnets to move, which moves the speaker cones, which moves air, which moves your tympanic membrane, which blah-blah-blah... eventually you hear sound.
So there's no need to worry. Nothing gets coarse. Nothing loses fidelity. Nothing loses audible quality. This is why vinyl fanatics get laughed at by people who know how and why digital audio works. The limits of even now-mundane CD audio are far above the possible limits of even hypothetically perfect human hearing. Nobody can hear 22kHz. Nothing below 22kHz is misrepresented in CD-quality audio. For mastering work, where effects will be applied later, higher quality recordings are wonderful, since you can guarantee that it will stay high-quality when downsampled to CD-quality, but other than that (and "economies of scale" where better parts are just as cheap to produce), there's no need for anything better.
here is an illustration
------ hi mom
http://en.wikipedia.org/wiki/CD#Audio_format
"The sampling rate of 44.1 kHz is inherited from a method of converting digital audio into an analog video signal for storage on video tape, which was the most affordable way to get the data from the recording studio to the CD manufacturer at the time the CD specification was being developed. A device that turns an analog audio signal into PCM audio, which in turn is changed into an analog video signal is called a PCM adaptor. This technology could store six samples (three samples per each stereo channel) in a single horizontal line. A standard NTSC video signal has 245 usable lines per field, and 59.94 fields/s, which works out at 44,056 samples/s/stereo channel. Similarly, PAL has 294 lines and 50 fields, which gives 44,100 samples/s/stereo channel. This system could either store 14-bit samples with some error correction, or 16-bit samples with almost no error correction."