|
|
楼主 |
发表于 2011-10-22 16:31:09
|
显示全部楼层
A Whole New Level of Clock Performance Digital audio is all about timing. The further we press the limits of clock precision the better sonic results we experience. We delved into space age technologies and produced a clock of unprecedented accuracy and the sonic results are simply stunning.
It all starts with a crystalline material. Crystals were found to resonate at a precise frequency a hundred years ago. We have since pushed our understanding of crystal oscillation to an exact science. When the right crystalline material is cut exactly the right way, and allowed to resonate at the right temperature, a clock is created with unprecedented accuracy. (97 femtoseconds of jitter is less than a tenth of a picosecond - a new class of clock.)
Conventional specifications and industry buzz words just do not begin to quantify the startling results of our clock research as we move toward a whole new level of performance. Our novel clock design has never before been applied to an audio product. Unique proprietary materials are precisely fabricated in an equally proprietary clock circuit and absolutely temperature controlled in an environment made from an actual space shuttle tile.
 The FemtoSecond Galaxy Clock plugs into the clock upgrade header on any DAC IV with an immediate sonic benefit. An excellent DAC clock tends to reveal the true nature of the original music. The DAC IV with the FemtoSecond Galaxy Clock reveals fine, delicate, very dense harmonics that are accurately organized and therefore correctly reveal the true sound of the instruments and voices. The harmonic congestion that translates to harshness and the familiar digital "annoyance factor" is just absent. The music and instruments sound like they are coming from live mic feeds on the best recordings.
The ability to recover the soundstage is closely related to the correct recovery of dense harmonic information. Because of this ability to reveal the finest resolution without any trace of harshness, the soundstage is defined to a degree never heard before. Each recording moves from venue to venue, each instrument takes its focused place in the soundstage. The combination of the dense harmonic recovery and the exceptional soundstage presentation results in a level of "believability" that has to be heard.
A personal note:
I do NOT want to push this clock on anyone. It is an expensive clock and I do not want a single customer feeling they made a bad investment. I cannot control what a distributor or dealer will do with a customer but I publicly pledge to happily give a full refund to any of my distributors who feel the clock did not live up to their expectation. I encourage my dealers to do the same for you customers. This is an extravagant luxury for those who want to hear something new. Let me explain.
The problem in quantifying the sound is that the sonic impact is on a different vector than for example DAC level. There are two sonic vectors that are 90 degrees apart. One has to do with timing and one with accuracy. They are completely unrelated. Digital filters, upsamplers and clocks all have to do with timing. Platinum - Signature - Diamond mostly have to do with resistor accuracy. The sonic impact of each vector are not related. The clock impact all three levels of DAC on the timing vector only. So the clock upgrade does not make a Platinum DAC into a Signature or a Signature into a Diamond. The clock impacts the time alignment of the sound. The more accurate the DAC, the more stunning the clocks impact.
So the impact of the clock depends on what a listener cares about. Vince (our US sales manager) wrote the description of what the clock did for the sound above. He has a pretty narrow (and super critical) focus. For him, it came down to one of his favorite recordings. When he sits off axis a particular cymbal sounds so real it sounds live. That hooked him and made him a fan. Every time he hears the recording without the clock he is sadly disappointed. He is a true audiophile. For me, I heard a high-res violin solo. The violin image went from an unmoving meter wide to actual size and I could see the violinist sway and turn as he played. Wow. Can't put a price on that. At that moment I changed my mind and determined to sell this clock even though the price is high. I want to hear that sound again. Irrational and reckless, but what did I work so hard for all these years for if not to indulge myself from time-to-time in something unique, new and wonderful.
The FemtoSecond Galaxy Clock MeasurementsClock measurements are a tricky thing. Comparing clock measurements is nearly impossible. Everyone quotes the lowest jitter numbers they can, which are the time interval error show in this textbook illustration below. But this error is very frequency dependent so usually this number is quoted for a very high frequency, where the clock is very good. They do not mention the jitter at lower frequencies where clocks are not so good, and where jitter is more relevant to audio sound quality.
A more useful measurement is phase noise. Phase noise shows clock jitter in the frequency domain. We look at the undesired power of the clock at various frequency ranges away from the primary clock frequency. The red lines in the textbook chart below shows the noise 1 kHz away from the clock frequency. This specification is more relevant to audio performance.
The MSB Galaxy Clock has amazing noise performance with a guaranteed jitter measurement at the worse case of just 1 Hz away from the ideal clock frequency of less than 97 femtoseconds (0.097 picoseconds). OK, so here is another pretty unbelievable jitter specification. We better stop right here and share our ideas about specifications.
Just as we broke new ground with our Ladder DAC design, 32x Digital Filters, Upsampling to 32 bit 384 kHz, and making the first 384 kHz USB connection, we have done something totally different and groundbreaking with this new clock. Jitter this low is very hard to measure and I am sure the mathematicians can argue about the exact test method that should be used and the exact numbers for years, but the fact of the matter is that this clock makes a huge sonic difference, and thats what counts. Our goal was great sound, not great measurements. We published what we measured with our own test system and method and I believe them to be accurate but there is a lot of variation in test methods, such as signal level and measurement frequency (we are around 25 MHz where most clocks are at 10 MHz) that skew the results dramatically so it will always be hard to make detailed comparisons. I think all our engineering critics will agree that far more important in a DAC design is where you put the clock, and how you transmit and receive the clock at the DACs themselves. The potential jitter added in these areas far exceed the clock jitter anyway. A poor implementation of this clock may add 1000 times the clock jitter to the DAC output! The DAC IV was designed from the ground up with this clock in mind, so the implementation is very clean.
 Here are our measurements at 24.576 MHz**:
- Phase Noise at 1 Hz -85 db
- Phase Noise at 10 Hz -110 db
- Phase Noise at 1 kHz -120 db
- Phase Noise at 10 kHz* -150 db
- Guaranteed TIE (jitter down to 1 Hz) .097 psec***.
* at CMOS output levels (For reference, at normal +13 dB levels the noise would be -170 dB).
** these measurements were at our measurement system noise floor. Actual values could be better but certainly not worse.
*** Calculated with an algorithm supplied by Agilent for low noise jitter calculation (3% accuracy predicted by author).
How long is a femtosecond? It is one millionth of one billionth of a second. Visible light oscillates with a period of about 2 femtoseconds. So forget about measuring clock jitter in picoseconds. MSB has now brought a new reference to the industry!
[ 本帖最后由 feifazhuce 于 2011-10-22 16:37 编辑 ] |
|
发表于 2011-10-22 16:29:12
提升卡
变色卡
