SO that's what you meant by "fancy oscilloscope"? OK, fine.
I'm sorry if you didn't get the implication of what I said. Yes, specialized measurement devices exist. Fine.
What's your point?
Audibility of close-in phase noise in dacs remains controversial in audio forums. The textbooks say one thing about jitter, practical experience with a variety of dacs suggests some listeners are able to discriminate lower levels of close-in phase noise. If the forum guys are willing to keep in mind that limits of audibility are estimates of population averages (not hard limits), and that different perceptual testing experiments may produce different valid results given the particular test conditions (e.g. test tones verses music; different dac topologies; etc.), then maybe there shouldn't be so much controversy.
The thing about audio scene textures is provided for anyone who didn't like me using a new term that was undefined for them. Particularly so since I used it in the context of describing discrimination of low-level low-order distortion. Again, some of the not-liking may have to do with issues related to beliefs about published limits of audibility.
The thing about audio scene textures is provided for anyone who didn't like me using a new term that was undefined for them. Particularly so since I used it in the context of describing discrimination of low-level low-order distortion. Again, some of the not-liking may have to do with issues related to beliefs about published limits of audibility.
Last edited:
yes, I have explained that twice. the modulation signal (AM demodulated signal ) is a sine with some harmonics.
no, it shows the AM signal as is. How can you see some low level harmonics by eye sight?
and what is your point? we have two signals with the same amplitude spectra and only differ in the phases yet sound clearly different. This is hard to explain by masking theory. One of the signals has a constant envelope and we name it FM and the other has a modulated envelope and we call it AM.
close-in phase noise should not be audible according to masking theory. however, close-in phase noise is the same as wow and flutter and our ears are quite sensitive to that. It seems that our auditory system has a built-in FM and AM demodulation capability
I'm not aware of any "text books" that couple jitter to any particular audibility threshold. And by "practical experience", you of course mean a controlled ABX/DBT, right?
Yes, at some point jitter would become audible. That would be a function of many factors, including the degree of jitter, the jitter modulation spectrum, the type and spectrum of the desired signal, and of course listener training. I really don't find any conclusive data, but it's not helpful to look at jitter as a binary result: "Audible/Not Audible".
Until a LOT of controlled testing is done, there's no data to argue about.
Again, the audibility of PN has not been quantified, barely studied. If you're going to keep diving down that rabbit hole, at least post a study with controlled testing. I'm sorry, but I cannot accept sighted testing here.
High frequency flutter (tape scrape flutter) yes, wow, no. The two are different and do not manefest themselves in similar ways at all. You can confine the comparison to high frequency "scrape" flutter, the longitudinal speed modulation caused by pulling unsupported tape across a stationary path element. It's tested for using a 5kHz carrier and wide-band demodulator. See the Dale Manquen T2DS transport analysis system.
The fact that the auditory system has the ability to detect AM and FM is well known , and not a new concept. But people are still talking about this, as applied to jitter and PN, as if it were a binary quantity. All we know is, lower is better. We also know that it's no big task to build a digital system with extremely low jitter and PN, even the most basic and inexpensive are excellent today.
Perhaps we can get of this side-track now?
It's very clear that in your time graph there is at least one major modulating signal, approximating a sine wave. If there are harmonics, they are not visible.
So, you synthesized an AM signal with the the same spectrum as an FM signal? And that FM signal is deliberately band-limited?
I have no idea why such signals would be compared, or what your point is.
I guess we don't understand each other.
Could you provide an audio file?
Why insist on ABX protocol, so long as a discrimination test is done DBT? No other possible protocol can work?
as mentioned, hard to see harmonics using the naked eye
same amplitude spectrum yes.
the FM signal is the carrier FM modulated by a sine. The signal is discrete-time but otherwise not band limited in any way
identical amplitude spectrum but sound very different. that is the point
I would not rule out that possibility 🙂
the wave files can be downloaded from https://purifi-audio.com/2019/12/07/amfm/
I might even be able to dig the Matlab script generating them out
Last edited:
That researchgate paper says several 100ns jitter is needed to become audible. And that, in 2005 consumer gear was already below the audible threshold.
...with expectation bias close on its heels. And another thing; I can't quite believe why any hobbyist would be that masochistic to want to 'train their ears' (as has been touched upon elsewhere) to better hear distortion and artefacts - therein lies lifetime dissatisfaction leading to bankruptcy! I like a good (but affordable) system but clearly don't have Golden Ears as I always enjoy the music more than the system on which it is played...
ABX is designed to determine difference detection. That must be established first.
Here's my version of difference detection in the case of sources if you don't have proper equipment...
Sample the output of both sources, one connected to left, the other to right. Invert one channel, go mono, and voilà - the difference between the two sources 🙂
Sample the output of both sources, one connected to left, the other to right. Invert one channel, go mono, and voilà - the difference between the two sources 🙂
FRom that paper: "The threshold of audibility for pure tones was found to be about 10 ns rms at 20 kHz and higher at lower frequencies. For nearly all program material no audible degradation was heard for any amount of jitter added below the level at which the DIR lost lock. Certain program material was found in which an audible degradation due to jitter was heard. The threshold of audibility for these programs was generally found to be in the range of 30 ns rms to 300 ns rms for sinusoidal jitter."
So...since the range covers a maximum that is 10X the minimum, we don't have a clearly defined threshold. And, they tested with sineusoidal jitter. As I said before, the jitter spectrum is also a factor. They didn't test for that.
"Finally, the audible degradation was found to correspond to measurable changes in the spectrum of the program material. The influence of jitter in causing audible distortion was found to be less than anticipated by the authors, and less than that predicted by both the technical and consumer audio press. Jitter induced by the digital audio interface was not found to be an audible problem for any of the program material auditioned.
In other words, we don't have sufficient data or testing in that paper, which is from 1998
FRom that paper: "It can be concluded that detection threshold for random jitter added to program materials is several hundreds ns even for well-trained listeners under their preferable listening conditions. According to Benjamin and Gannon, sinusoidal jitter as small as 30ns (r.m.s.) might be detectable under a certain condition. Considering these results, the maximum acceptable size of jitter would be the order of ns.
In other words, "We don't really have a threshold".
From that paper, "Jitter can only be considered totally inaudible if the worst case jitter induced sidebands are at least 23 dB below the A-weighted system noise. Above this level jitter may be audible or it may be masked by the program audio."
They're conservative, so they won't be wrong. But there's no published evidence or testing there. It's a fine design goal, Benchmark makes great stuff.
As I said, there is no well defined theshold of audibility. There are various opinions and ranges given specific jitter modulation signals. It's literally all over the map.
Clearly, given the published ranges and methods. But that also means we don't have anything clearly quantified.
right click and view the HTML source, search for 'wav'and click the blue links - I got the download that way (data mining hack)