Quoted by fireworks:
Just to avoid any confusion, the author is using an analogy here - a PLL does not store energy in the way a flywheel does. Both are examples of low pass filters.
Another analogy: PLLs do not use 'inertia'.
We can. Look for the dominant pole in the feedback loop.
Almost like an LP on a turntable, only much better than LP. Interesting that some turntables used quartz crystal locking, and so would be subject to exactly the same effect as digital (but made worse by mechanical issues too). Low frequency jitter comes from the crystal oscillator, and is unaffected by the PLL.Max Headroom said:
Musical Fidelity V90-DAC D/A processor Measurements | Stereophile.com
$299 buys something with no obvious jitter related issues. Guessing the SRC just reclocks everything. Nice linearity as well. What more could anyone want?
$299 buys something with no obvious jitter related issues. Guessing the SRC just reclocks everything. Nice linearity as well. What more could anyone want?
I don't hear digitisus its just another scare story that seems to disappear with blind listening tests same with sub nS jitter figures... Of course esoteric audio has its own agenda (making money) so has a vested interest in twisting reality and physics to suite its needs as well as promoting its own versions of a Grimm reality, I am very cynical about it all and get more so every day... because it does detract from what is achievable and would yield the best solutions for most into good sound reproduction. A prime example is 20ps jitter (or even 200ps) very hard to achieve and not really diyable so its almost an unachievable target, where as aiming for 20ns would sonically give the same results (unless you are one of the GEB) and is infinitely more achievable....
The chart from the 1992 AES paper by J. Dunn is also shown in that reference. it shows jitter of 20 pS is an issue at 20 kHz. It increases to 100 pS in the critical midband frequencies.
So You can read what ever you like from that cite.
It is a good discussion of the types of jitter and the methods for dealing with it.
The original paper has been mentioned here before it is at http://www.nanophon.com/audio/jitter92.pdf
For a brief on crystal phase noise http://www.rakon.com/component/docm...ise-and-jitter-in-crystal-oscillators?Itemid=
Phase noise to jitter is discussed here http://www.analog.com/media/en/training-seminars/tutorials/MT-008.pdf
Last edited:
Worth quoting it more fully
But that was 1992. Is this really a problem any more other than the NOS crowd who hang on to 1980s multibit DAC technology? My Marantz CD80 was great in its day, but its day has come and gone.
Yes read the original paper. The jitter sensitivity curve doesn't look like 6 db/ octave to me. You are welcome to your own interpretation.
Yes the original import 2 LP's bought it in 1980 or so, one of my most pristine recordings unbelievably few tick and pops at this age. Lots of nice film related inserts, in Italian of course.
EDIT - The Good, the Bad, etc. is available on import CD with the original film ensemble.
Last edited:
Is it audible in proper testing.... the research sais NO for most people... is it achievable in an Audio based design! Is it necessary....
If we set a low pass filter to the voltage signal witch drive the VCO, it is adding some kind of inertia, or are-we just play with words ?
Well let us see. The jitter spec in the old AES-3 for data transfer was 20 nS. So if you start with that in your test system do you expect to measure a lower result?
Last I looked that spec was being tightened up quite a bit based on current assessments.
There are low jitter oscillators that certainly will easily provide what is needed to meet that. But if you look at the data on many of the low cost units they have trouble with even the current 20 nS on data transfer.
Basically there are two or three outboard D/A units that do accept the data transfers and then reclock to the D/A with a high precision master clock. The Benchmark that Dick uses is one of them.
So to be clear Jan found 50 nS in data transfer jitter caused outright failure of his semi-pro test unit. Most outboard D/A converters get their clock signal from a PLL tied to the data transfer rate. A D/A clock jitter should be under 100 nS for a current design and under 20 nS according to Demian who actually is in that design business.
Now when you are using the CD players internal D/A in most cases this is considered a budget solution. So the best you would expect is 100 nS of clock jitter and due to cost and other issues probably worse.
Is this the case? Just picking one medium cost outboard from stereophile shows that it reclocks by dint of the SRC and anything that uses an ESS sabre DAC get jitter removal on chip. Or was I just lucky on a random sample and Musical Fidelity is one of a handful of properly designed converters?
I would confine my tests to freqs in the most sensitive hearing range....3-5KHz.
I do not think it will insert jitter.... havent seen that in the manual (yet?).
Master CD/HD Recorder arrived.
Do you have a recommended jitter gen or CD jitter source/recording for me to add to this for comparison tests?
Anyone with real knowledge about the jitter on (used) commercial Cesium time standards (with a 10MHz output)?
THx-RNMarsh
Last edited:
useless for audio actually - they excel in long term stability but have poor audio frequency phase noise - look at the time nutz sites for good info
for a while Rb frequency standards from Cell Towers were cheap in surplus - for low phase noise they all use ovenized Quartz oscillator, "disciplined" usually once a second by the atomic clock
you want to use the better performing Quartz oscillator if you really want low noise for digital audio
and once again the Dunn numbers do not come from listening tests - they are all calculations - at best based on human hearing threshold in quiet
for a while Rb frequency standards from Cell Towers were cheap in surplus - for low phase noise they all use ovenized Quartz oscillator, "disciplined" usually once a second by the atomic clock
you want to use the better performing Quartz oscillator if you really want low noise for digital audio
and once again the Dunn numbers do not come from listening tests - they are all calculations - at best based on human hearing threshold in quiet
Last edited:
- Status
- Not open for further replies.