From my job experience with digital electronics I have assumed that the data read in from the disc would go into a buffer memory, and be "clocked out" with precise timing. Parts-per-million would be easy (inexpensive)to do. It would not matter how irregular was the timing of data input from the disc.
I gather that most, if not all, disc players fail to mechanize this approach, and are therefore subject to "jitter" problems. Perhaps this is understandable for a $100 item of equipment, but at high end prices it is inexcusable.
I've been listening to several DACs and CD players over the last year. I enjoyed listening to those with 24/192 upsamlers the most. I attributed this preference to the upsampling.
Well, I tried a Benchmark DAC1, which has no upsampling at all and it had the same characteristics that I liked in the upsampling DACs. After a bit of review, I realized that all the DACs and players I tried (including the Benchmark) had special low-jitter circuits or third party add on clocks to reduce the jitter (in addition to the upsampling).
At this point, it appears that the low jitter was providing a benefit that I completely mis-identified as the "upsampler" effect. I don't know that I can "hear" jitter, but I can identify when it is not there.
According to Benchmark Media several months ago, the DAC 1 upsamples to 24/110 via Analog Devices AD1896. They planned to increase the frequency to 192kHz once they were satisfied with the sound. The unit I received two months ago was marked as a 24/192 DAC.
Of course, their claim to fame is immunity to jitter. I'm using the coaxial input with DH Lab's D-75 RCA to BNC interconnect. In any event, the sound is absolutely transparent.
Jitter is a problem with virtually all Transports, particularly stock. Jitter is also increased with lossy S/PDIF cabling. It is aggravated by systems that are not properly impedance matched as well.
I can only speak for myself - I can plainly hear the effects of jitter - they are like echoes or halos around each instrument or like high-frequency sibilance. Adding a Superclock2 to a transport or DAC significantly reduces jitter and this change should be obvious to anyone that is not deaf.
Audioengr- have you measured the before and after jitter levels on the units you mod?
I agree that the effects of reducing jitter are obvious. The system will sound much smoother and more dynamic. I find reducing/eliminating power supply noise to be just as great an improvement as add-on clocks, all of which recommend using their own power supply.
Eldartford: You keep having the same misconception about the jitter problem in CD playback over and over again. All your background in digital electronic does not help because your background is stricly only in digital domain. In CD playback, there is an interaction between digial and analog and it is completely different from what you think.
Kana813 - no I have not measured the jitter. I would like to, but I need some equipment that I cannot currently afford in order to do this....
Eldartford is correct. Solving the jitter problem only requires FIFO buffering at the desination. However, it is non-trivial to still maintain all of the real-time functions of a typical DVD or CD player with a FIFO buffer queuing the data, such as fast forward, skip etc, without incurring large latency penalties.
Audioengr: All cd have some kind of buffer, large or small. In fact, they would not function without buffering their data after reading from the disk. The problem is that the clock itself has jitter. Anyway, I hope people would understand more before voicing their opinions.
Andy2 - I am referring to a large FIFO buffer at the DAC, not in the transport. If this is in place, then the data can be clocked out of this buffer independent of the speed of the drive, virtually eliminating jitter, assuming the read clock is very stable.
I know a little bit about this, having designed digital harware as a EE for almost 30 years. I was a design team leader on the Pentium 2.
Audioengr- thanks for your response. I know the measurement equipment is very expensive, but it would be interesting to find out what level of jitter reduction is required for better sound quality. It also appears, that the input receiver on some DACs is more sensitive to jitter than others.
Andy2- "The problem is that the clock itself has jitter." This is why I still use the discontinued Genesis Digital Lens, and a number of audiophiles are now using a Apogee Big Ben to deal with clock issues.
But, as I said in my eariler post, if you feed your transport or CDP's clock cleaner power, you can improve it's performance and reduce jitter. Upgrading cheap polarized caps in the digital circuit path also helps.