Why in the hell would anyone use a test designed for testing the weaknesses of SPDIF back in the 90’s for modern digital audio? Yes a great test for SPDIF and redbook. But not enough for today’s high end.
If you want to do some tests that hold weight. Take some bitclock phase noise tests at the closest point to the analog conversion. But unfortunally you need the proper equipment for this.
No, that’s why it is so hard to get performing this well… Not fan of USB at all and certainly advocate of using networking as underlying layer instead. As you already know form the NAA stuff.
Do you think it is good that it takes so much time and money to get it working well? At least it makes me worried. If it’s that hard, I’m not going to even try to implement it (not that I would have intended anyway, after looking at the specs).
NAA is very simple on purpose and doesn’t transfer any clock over network. And still you could use it to play music to a DAC on the other side of the globe over internet without problems, as long as you have sufficient average bandwidth. Same goes for audio between Roon and HQPlayer.
I’ve sold dozens of Superstream streamers using your NAA protocol. I’m vividly aware of what the system can offer. But now I’ve moved onto a system that not only functions smoother, but sounds better than is possible via USB. That system is Ravenna. And the bonus is none of my clients are bound and gagged to use 1 media player for eternity if they want to benefit from the very best audio over IP can offer.
Because everybody still uses it and because it is not designed for S/PDIF, but specifically for I2S which is still used a lot. And it works great because it shows very nicely all kinds problems.
That doesn’t matter, only thing that matters is what comes out as audio. Your bitclock phase noise tests don’t tell anything because things usually go wrong precisely at the conversion point and nowhere else. And you DAC architecture to large extent defines jitter sensitivity.
What is wrong with my equipment? I haven’t seen any measurements from you and your great equipment. Usually people who complain about my equipment fail to produce any standard J-test measurement results themselves.
Why are you using USB in your NAA system? NAA is not tied to USB in any shape or form, and it doesn’t know or understand anything about USB for that matter. Just build NAA straight into the DAC. I know many people who run DSD512 straight out of NAA without USB or I2S or stuff like that.
Jussi you’re fighting a losing battle. Dozens of manufacturers are implementing Ravenna as we speak. If you want their clients who purchase their products to use your software, it might be wise to stop bashing Ravenna. You don’t see Merging bashing NAA.
I’m not bashing Ravenna, I’m bashing you for bashing everything else, including NAA and my measurements. You started bashing my stuff, so just for the fun of it I bash back. You asked for it. 
For HQPlayer there’s a significant difference, Ravenna goes through the audio driver layer while NAA doesn’t.
P.S. What am I losing? Sleepless nights? 
I’m not bashing anything. I’m just embracing better technology. The Jtest isn’t your test. It’s a test developed in 1992 by a fellow at prism sound for testing SPDIF. It’s still used today because it’s the only standard jitter test that works from the analog outputs on APx machines. It’s mainly just for marketing purposes to prove a DAC isn’t horribly engineered. In reality all it really achieves is misleading folks to think a $500 Ifi is just as good as a $90000 MSB Select. Because it simply isn’t good enough to pickup what matters.
I know who designed it, etc. It is designed to test both clock stability and data leakage from I2S data lines to the I2S clock lines. It also nicely shows any other interference leaked to the clock lines on the PCB. Plus it tells nicely about jitter sensitivity of the conversion process itself, which in turn depends on the conversion architecture.
Works very well even today.
Analog outputs? How do you use it from analog outputs!?
I have the data in a WAV file I just play through the gear, end-to-end (popular term these days). It provides good end-to-end result of the overall timing performance.
Interesting that everybody and friends also in audio press still uses it and find a lot of problems using it.
So the MSB performs worse on it than $500 iFi? Are you trying to discredit the test because you have something that is expensive yet gives poor results on the test? If it is not good enough to pick up what matters, then the results should be always perfect? But that’s not what I’m seeing. I’m yet to see two DACs that would measure the same in that test.
It is a very good test because it shows exactly what matters, both phase noise and all other clock interference on one simple measurement. And I certainly will continue to use it.
Remember that jitter is relevant for audio data only after the D/A or A/D conversion boundary, until that it doesn’t have any meaning.
Guys, could you not just head down to the pub and continue your discussion over a few pints?
Imbibing a nice Weissbier al fresco myself at the moment.
Chill.
.sib
Like I said it’s widely used because it’s easy to use and built into the equipment most use. The audio press uses it because it achieves the objective of making the masses believe the DAC tested achieves great jitter performance. If they can prove the DAC passes both subjective and objective testing, they successfully achieved preforming the job the manufacturer sent the unit to them to achieve. Done deal. However for folks who really want to know why the subjective results of the test were how they were, time for some better gear. But unfortunally no test points on the exterior of the cases for these tests.
The Ifi/MSB comparison was just examples of 2 DAC’s. Almost any DAC today will pass a jtest with flying colours. Does this mean all DAC’s are equal?
That’s not how the results look like.
MSB Analog DAC seems to have some spurious sidebands:
So does the Diamond DAC IV too:
But I’m personally very happy to see that both T+A DAC8 DSD and Holo Audio Spring DAC give cleaner output at DSD512.
That’s not what I’m seeing.
OK, please post those measurements for most of the popular DACs on the market today. I’m eagerly waiting since you seem to have answer to an important question!
Without several references of comparisons with many DAC’s, showing the proper test results would be useless. We are working on new measurement standards that will close a lot of grey areas between what is heard, and what is measured. It won’t be long before they’re widely adopted. 1992 was a long time ago. Time for some innovation in this area.
Just reading that those MSB analog DAC Jtest’s were taken while measuring via the toslink inputs. How many MSB owners use the toslink input? Why not do the test using the I2S inputs from their transport, or the USB like how the 99% will be actually using it?
Then with the diamond, the Jtest only proved that it isn’t good enough to tell apart the differences clearly heard between the inputs from diamond owners. According to that test, toslink from an Apple TV should sound identical to SGM server via USB.
Could we please mark this thread as advertisment. Some kids could be misguided otherwise.
Until such is available, it is vaporware, IOW, doesn’t exist and we stick to the existing standards. Things that don’t exist don’t matter.
HiFi-News does, but I cannot give you URL to those results. But the DAC has many other problems in plain IM measurement, so the Jtest is just one area. If someone wants to loan me one MSB DAC, I will gladly measure it. But I’m not going to buy one just for measuring.
So they have some work to do! Of course all inputs should sound the same. HQPlayer sounds the same from Toslink and FLAC file.
If you prefer embrace the pleasure of listening to music maybe you wont have to get obsessed too much about the technology.
If I owned Ravenna… I wouldn’t like you to talk about it because you sound like a repetitive commercial
I would like to send you a t-shirt with a teaser
“Ravenna is the best, bash the rest”
Just because something doesn’t exist in your mind, doesn’t mean it doesn’t exist. Before they discovered the earth was round, was it really flat? Did it only become round at the point of discovery?
Anyways this will likely be technology for OEM internal use. Laymen end users are happy with 1992 tech, so no point in confusing anyone with test results that are meaningful in today’s audio.
When you build modern audio gear, you must choose a digital interface technology to use. It’s an essential part of the process of building audio gear. And unfortunately if you have specific end goals, you need to use technology that meets your requirements.