Some additional thoughts about measuring jitter

[PeterSt]

FYI : Here is another : http://www.bd-design.nl/contents/en-us/d168.html

 

I can't recognize that it sounds better because of less jitter. It sounds very different though than its predecessor (the TwinDAC+) which is basically the same DAC.

 

I'm afraid that DACs whith really less jitter at the output, will have it superimposed on certain frequencies (hence less randomly spread), which may be worse for net result than more jitter, randomly spread.

 

This is by no means based on measuring, but how my ears perceive it, combined with that it should be better ...

 

Peter

 

Edit :

To put this in some kind of perspective : By no means I wish to imply that this CrazyT is a bad DAC (it is top of the bill actually). The person behind it compared it with the Wavelength (I did not) and it is his perception that the Wavelength sounded "less good", FWIW. Again FWIW I trust that observation. The point here is, that the same person is as ready to tell that my own DAC is better again, and by no means this is a low jitter DAC at the output when run in the best sounding mode (which is around the SRC).

Now what ...

But there are so many parameters involved ...

 

[Roseval]

The only question, really, is whether or not it is audible in most gear

As I don't own 'most' gear and being a bit egocentric I’m only interested if it is audible on my gear.

 

 

[Roseval]

Another implementation of asynchronous USB indeed.

But at the expense of using dedicated player software and limiting you to WAV/FLAC at 16/44.1 only

 

 

[Ashley James]

We've built many DACs and nine different CD Players since the early nineties and more recently we've compared SRC, which removes jitter for an extra two quid with digital receivers that reduce it to insignificance and we've not been able to hear or measure significant difference. This is as predicted by the chipset manufacturers and by science.

 

For these reasons, I quoted the maths (which was ignored) to illustrate at what level jitter noise would be worse than 100 dB, a figure that is much better than most preamps and power amps and miles better than the noise in passive speakers. These came from Dr Chris Smith of Roke Manor, which I thought good because he has no commercial interest and great deal of knowledge as Research Engineers tend to.

 

So to sum up; Jitter produces noise at very low levels that can be measured, we know how much produces more noise than -100dB and we know that most people's hi Fi systems already have more noise than this. So the discussion is probably of no value in a practical sense, although it has obviously been enjoyed. Tim is probably in a better position with headphones to hear problems that people with a separates system and so are we because our active systems have much a lower noise floor than the older technologies.

 

I think I ought to stress that there are not significant choices available to someone wanting to make a DAC, all he can do is follow the Chipset manufacturers applications notes. Some provide a PCB layout and all an evaluation board with measurement points etc, so that companies like us can avoid making mistakes and compromising performance.

 

As I've said many times, the problems have always been ultra sonic spuriae and RF. Stuff above 20 kHz that upsets some amplifiers. Years ago this was only about -30 dB but now with a good DAC, it's better than 100 dB, however some are out there with -75dB and there are still amplifiers that are profoundly affected. Therefore DACs can sound different with different systems, more different than anything caused by jitter, but for some reason and despite this being the biggest single factor, no one seems bothered by it.

 

Ash

 

[tfarney]

"Most? Mine!

 

The only question, really, is whether or not it is audible in most gear

As I don't own 'most' gear and being a bit egocentric I’m only interested if it is audible on my gear."

 

Agreed, but I evidently wasn't clear. I was looking for a way for manufacturers to substantiate their claims, something I'd expect them to welcome. Something quite useful to the audio shopper. And the question wasn't audibility ON most gear, but audibility IN most gear. Or any gear, really.

 

This would be a piece of cake for someone manufacturing outboard equipment designed specifically to reduce jitter -- same source, same DAC, same system, same listeners...magic box in/magic box out. Blind test. On the most sophisticated high-end system or a Yamaha AV receiver. Or all of the above. All the discussion of picoseconds would be moot.

 

It would be less straightforward for the DAC manufacturer who dares to build jitter control into the core device instead of building another groovy box for us to buy, power and interconnect, possibly creating much greater opportunities for noise than the box could ever eliminate, but it isn't all that hard: ABX the old model DAC, without the supersonic jitterphonic bat-ear reduction circuit against the new one. Can a difference be identified in a blind test more often than you'd get heads in the toss of a coin? No? Back to the drawing board. Yes? NOW is the time to judge the difference subjectively and reel out the poetry and the statistics.

 

It's really very simple, in concept and execution. It is also very inexpensive. The fact that it is obviously not being done (obvious by evidence of its complete absence from these long jitter discussions) speaks volumes.

 

Tim

 

 

[markr]

... those last 5 posters were either Gordon or Juergen, were they?

 

sheesh :-(,

markr

 

[BobH]

It's a shame. As a couple have already suggested, they'll probably continue the conversation in private and we'll be left with the usual merry-go-round of drivel, until the thread dies or Chris kills it. Ah, well, maybe I can buy a book.

 

IMHO, YMMV, etc, etc

 

[markr]

.. Juergen didn't really want to expand on this here anyway. I don't blame him - the web is the place to just post a paper, not really one to discuss it, UNFORTUNATELY.

 

yeah, IMHO, YMVV, ad nauseam....

 

- markr

 

[PeterSt]

I think the saddest thing is that this thread is just about getting prepared to measure beyond current possibilities. So ...

 

So I think these two guys should really stop nagging about some positives here, and most of all they really should stop interfering with every other thread with their by now long known positions, and stop spoiling the fun for everyone.

 

Trolls ... (which is what it really comes down to after 1046 of such posts, sorry ...)

 

Peter

 

[JR_Audio]

Originally I started this post to just answer one question, where I wanted to clarify a little bit more specific, where the jitter numbers came from, when I compared different software drivers, which leads to different jitter and also different sound.

 

Justifying jitter is a little bit like “fighting against the wind”, because there are so many people out there who just believe in math and the thinking: what can’t be measured can’t be heard. That was also my thinking when I finished my electrical engineering.

 

I am happy that besides being an audio engineer for over 25 years, I also make music for over 30 years and actually do some classical recordings and own a small recording studio (for 5 years) so besides the learned math and the ability to use measurement instruments a have at least a small ability to be sensitive to music.

 

So I am playing around a lot with sound and music and most times it is the hearing of something as the starting point, followed by endless “research” to try to answer the question, why is this audible and what is is the most appropriate measurement tool to find an answer.

 

And also with my classical recordings, I am surprised and also nervous, that when I have the raw recordings, to what extent some points are audible. Really, this sometimes totally pushes away what I was told during my engineering education, what would be audible and what not.

 

So really, I originally do not what to open this long thread, because discussing about jitter is a similar long story as discussion about THD, but in both field I wished, that if a number is raised, one should really specify the conditions.

 

So I wish all of you a nice day and follow your heart, when listening to music or justifying audio quality. I have better results, in doing this, than following my “brain”, when hearing to ABX (ob man, I wish this explanation is clear (because I am not native English)).

 

Juergen

 

 

[markr]

.. Thank you Juergen (AGAIN). This is what I thought we might hear at the end of things. I am really sorry that some here felt it necessary to advance their personal agendas ( you SHOULD know who you are.........) and I have to say that I totally respect your position. I pray that you have total success with your adventures on and on into eternity. Thank you again for taking the time here to delineate some your experience here!

 

[Roseval]

When using playback software with native EMU ASIO support (Wavelab, Cubase, Reaper, JRiver Media Center, Foobar with ASIO Plug-In), I got bit perfect true data. With theses software I got jitter values of random nature with approximately 2.5 ps

 

This type of in formation is rare.

Most people simply claim they hear a difference (big of course, being blown away seems to be an absolute minimum).

I’m glad we have somebody on the forum who is able to establish if the software allows for bit perfect data and can measures its impact on jitter.

What we need are ‘facts’ especially from guys like you who know that a number is a fact but not an explanation.

Keep up the good posting (and stop complaining about your bloody English).

 

[Roseval]

There are a couple of manufactures nuking any tread with promotion of their products.

These are the brands that will never make it on my short list.

 

 

[DavidJPettifor]

PeterSt wrote:

So I think these two guys should really stop nagging about some positives here, and most of all they really should stop interfering with every other thread with their by now long known positions, and stop spoiling the fun for everyone.

 

I totally disagree. I originally joined this forum believing it would be beneficial to a beginner approaching computer based music systems and provide helpful and sensible advice in getting the most out of such a system. tfarney, Ashley James and a few others here, have done just that, providing sensible, well intentioned advice that won't break the bank and provide for a well rounded and well constructed computer based audio playback setup.

 

In these times of hardship for the many and the fact that being frugal is a necessity for the majority, sensible and practical advice is what's required. This, both Tim and Ashley have offered in spades. The positivity found here at CA, IMO, comes from THEIR posts and not that of others, who continue to ramble on about the same old, no doubt driving away any newcomers to the site who already find the whole experience of computer audio daunting.

 

Basically, through all the fog of jitter, Tim and Ashley have been a beacon of light that has allowed me personally, to find my way without too much issue. The amount of threads that have been started in relation to jitter make me question what the actual agenda is here? CA is going to suffer if it continues IMO.

 

PeterSt wrote:

Trolls ... (which is what it really comes down to after 1046 of such posts, sorry ...)

 

THAT comment is TOTALLY out of order.

 

--

djp