Now Hear This
 |
| Paul Wolff (center, in tan coat) at the Los Angeles AES in September 2002. |
The Nyquist theory is correct by stating that any sine wave sampled at twice its frequency can be reproduced accurately. The problem is it does not guarantee that it will be reproduced in the correct location in time as it was originally — with relation to other present waveforms. Since the sample rate is constant, as you slowly increase the frequency that is being sampled, the lower frequencies get more samples than the higher ones — and it slowly ends up at 22.05KHz, which has only two samples.[an error occurred while processing this directive]As the frequencies slide up, like the subtle harmonics of just about anything, they are not all getting the same treatment as far as sampling. Although each individual sine portion is reproduced correctly, its position in time with relation to the other associated frequencies is not always the same. This creates the smear that we have all heard so much about over the years. Technology has definitely minimized this, but it is still there. How much this effects our brain's ability to just listen and enjoy is another topic of debate. I can tell you from experience, that spending an hour in a room with a 44.1 CD or a 48K recording eventually fatigues me. On the other hand, my first experience with Sony's SACD demo at AES was quite enlightening. I listened for about 30 minutes with relatively loud levels, and when I left, I felt like I had been outside in the fresh air. I would have to say that it was 50% hearing and liking it and 50% subconscious. That is the only way I can describe it. One thing I noticed is that I could tell the signature of the console hiss of the different song selections to the point that I picked out the API Legacy as one of them.
There are those who are under the impression that this higher bit rate and sampling rate is for extended high frequency listening that we can't hear anyway. It is true, we can't hear it, but it takes that much to get what we can hear to sound like it did before we sent it to the chopping block. The higher bits, going from 16 to 24 was done not to increase the dynamic range of the listener and get the noise floor down below liquid nitrogen — or the peaks loud enough to blow out all speakers. If that were the case, a little increase to 18 or 20 bits would have been fine. The problem is that with 16 bits, unless you have every single channel on tape maxed out to the limit of the converters to reach the full 16 bits, any sound that is quieter than that level gets a linear reduction of bits that it was sampled with. So a sound that was recorded lower in level ends up with 8 or 10 bits of resolution. Pretty grainy. Move to 24 bits, and now you can record in a more reasonable method, and the same lower level sound now has 16 to 18 bits of resolution. Much nicer. So, Yes, 18 bits would be good enough, and now we can get there, but only with 24 bits available.
To say that even a 24/96 system is "too expensive" is a stretch. I would feel that way with a SACD resolution system, but not 24/96. We have to make that our minimum. These higher rate systems are here to stay, and thank God they are. You've all heard it before, but when our kids find the record player, they tend to listen to it and really like it. With the falling price of memory, processors and hard drives, it all becomes moot. Computers double in speed every 11 months, it seems, but do we want to stop that because it is good enough? They won't stop until they can think as fast as we can, and when that happens, they will wonder why we are so slow. The most depressing part of this whole debate is that ever since the first CDs, anything that was recorded on the 16/44.1 systems will never be able to sound any better because there is nothing else there to work with. Even old tapes can be cleaned up and can sound better than the first time. Even the CD part of the SACD dual layer recordings sound better because they were first recorded at a higher rate and then sampled down to play in the 16/44.1 format.
It is true that better bit rates don't make the band any better, but it just might make them want to be better.
Paul Wolff, Director of Engineering for API (Automated Processes, Inc.), has been involved with API since the early 80's and helped define the bulk of API's products.
[an error occurred while processing this directive]
