Some fools perceive cost as meaning quality and buy into the BS, frankly I consider that to be their problem
Just some evidence which backs up what I said above about low close-in phase noise & why the perceived improvement is most notable in bass sounds - this slide shows the listening results from 6 different research papers plotted on one graph. it shows the smallest pitch changes discernible at various frequencies - turns out that @ around 1KHz differences of 2Hz are discernible in sinusoid test signal, @ around 500Hz it drops to 1Hz, around 125Hz it drops to 0.5Hz
What we see is that close-in phase noise can have its greatest perceptual effect at low frequencies where we are most sensitive to frequency changes. Frequency modulations 1Hz to 0.5Hz offset from the fundamental is perceptible in frequencies from 500Hz down to 125Hz
I'm looking for that presentation from which that slide comes & I'll post the link when I find it
What we see is that close-in phase noise can have its greatest perceptual effect at low frequencies where we are most sensitive to frequency changes. Frequency modulations 1Hz to 0.5Hz offset from the fundamental is perceptible in frequencies from 500Hz down to 125Hz
I'm looking for that presentation from which that slide comes & I'll post the link when I find it
Last edited:
Many of SOTA DAC designs behave the same re jitter. Talking repeatedly about jitter issue reminds me giving uA741 as an audio opamp example.
I don't want to overplay the jitter thing as in my experience with Jocko's low LF jitter clocks, the improvement was small but noticeable
The point is that your posted graphs from Stereophile tell us nothing about close-in jitter so your claims are spurious with regard to this topic.
Try to deal with/ enter into the discussion rather than just posting bullish claims along with evidence which doesn't support these claims
Last edited:
A scale is a basic requirement of a graph though. Can we ignoramuses have one so we have some idea of the magnitude of the issue being discussed?
Not an appropriate analogy, IMHO. Nobody uses uA741 for audio any more. However, most people at diyaudio do not appear to be using SOA dacs, although some are. Guys working on tri-amping speakers hardly ever seem to consider use of SOA dacs, because of the excessively high cost. Scott Wurcer also questioned me about trying to get people to use better dacs. He wanted to know if I would then expect people to upgrade their whole systems to take advantage of the better dacs. I replied that I only recommend to people that they mod a low cost LME49600 headphone amp which should be good enough to hear everything coming out of a good dac. Then people can make informed decisions as to whether they want to work towards a better power amp over time, say. In addition, I have one SOA dac. There is no way I would consider biamping, much less triamping due to the very high costs of SOA dacs and power amps.
Moreover, I agree with Crane Song and their choice to design 45fS jitter dacs down to 10Hz offset, and .45pS jitter down to 1Hz offset. They find jitter at 10Hz offset to be critical for optimum sound quality. Only reason other SOA dacs aren't down to jitter that low is due to the slow rate at which audio technology advances. Eventually, jitter in SOA dacs will probably go even lower because the effects are audible to at least some people.
If you want to experience very low jitter, pick up a Crane Song Solaris dac. You might adjust your position some after that. Only one way to find out.
Well, PMA showed FFT plots & claimed that they showed low jitter. As I said the x-axis of these graphs are so gross that no close-in jitter could possibly be seen. This is typical of device measurements & the lack of understanding of what's being shown. I wouldn't mind if people weren't so bullish about their claims.
You have to understand that measuring close-in phase noise is difficult to do & requires time even for oscillators. Not many oscillator vendors show phase noise less than 10Hz for this & for the reason that it would not look good for most.
So measuring close-in phase noise for DACs is even rarer & hence why we don't have many examples
This is a typical type of Jitter FFT plot of a DAC (taken from Joe Rasmussen site)
And here is the same plot done to show the the area inside the blue box on the above plot. Note the different x-axis scale between these plots.
Also note the close-in phase noise now visible in red - it is now seen to rise to -80dB but we're not really able to see the amplitude of the jitter even closer in to the fundamental due to the scale of the x-axis - it needs a finer scale to show this
Thanks, I had a quick scan & see they have a page on jitter with some good references to Allan deviation, close-in phase noise etc.
Only issue I had was they mention the use of an ASRC for jitter reduction - this puts me right off but I need to find out more about their ASRC. My experience with ASRCs is that they can improve bad (jittery) sources but tend to disimprove good, low jitter sources - in other words everything is brought to the same level. Can sound good but not as good as the low jitter source without ASRC processing
ANAP Power Cable is priced "$1,149 up to 2 meters with spade termination." and the features show,
"Excellent solid, milled spade connectors
Very transparent and natural sounding interconnect cable
Engineered venting on the top and airflow system avoiding heat buildup inside the chassis"
Spade connectors, interconnect cable and venting on top of power cord? The quote above clearly show number of typos. Shouldn't the website be updated right away so that it's not a public (web site is accessible to public) laughingstock?
That would be good. IMO there is a gross misunderstanding here of close in jitter vs. pitch, the time scales are being highly distorted. 1Hz shift at 200Hz is 0.5% the grossest cheap ceramic resonator is not that bad. Please find out what the stimulus was in the pitch sensitivities tests were, I suspect it was taking an oscillator and shifting the frequency and holding it at the new frequency. This has no analogy to close in jitter.
Pretty much all decent DACs come under the green line these days. It takes a high end designer to get above it! Whether that test is still valid now we stream rather than play CDs is worth discussion
Yet TIM and SID are still discussed and the seminal paper on that needed to use a 741 to get measurable amounts of either. So it is valid wrt shooting boogeymen.
The problem here is that the gains from a tri-amp setup are way greater that the possible improvements from DAC upgrades. Most of us cannot see a reason to improve something below the level of the brownian motion in the room when our speakers are only +/-2dB in FR. Certainly for me DAC upgrades are the last thing on the list once I have everything else as I want it.
Not saying you are wrong, just I'm looking at gross errors in the system first.
Yes, I suspect you are correct, having looked again at the chart - I still can't find the original but it's really thresholds of FM discrimination that we should be looking at in relation to jitter but I don't think single tone tests would be appropriate, rather dichotic, complex signals seem more analogous to jitter?
Yes, we need a test signal which is more specific for how jitter may be generated in streaming devices rather than the current J-test signal which was designed to stress SPDIf receivers & their early recognized issue of inter symbol interference
Not sure we have a good grasp on possible jitter inducing mechanisms involved in streaming processes? I suspect that there noise mechanisms are at the heart of it?
This is worth some research, I'm very interested in Jocko's stuff but I would like to put it into a technical framework. I still remain unconvinced, the average reproduction system is not stationary to anywhere near the level of the best clocks. There are clocks now that can test relativity from going up and down one floor in an elevator.
Well I am currently in the camp that thinks there is more FUD than issue here. Happy to be proven wrong but so far its the smallest fish in the audio pond.
OK, you seemed to want to discuss the validity of the J-test measurement for streaming audio - what do you want to discuss about it?
- Status
- Not open for further replies.