Ji-Ji-Ji-Jitter

[antcollinet]

We're not talking about subtle differences here that only trained ears could distinguish.

I asked this question elsewhere yesterday.

Why do you think perceptive biases cannot create a difference that is un-subtle?

Again - the fact you are hearing unsubtle differences where what is changing should only create subtle differences below the point of audibility is just a pointer that the cause not in the sound waves reaching your ears, but is being created in the wetware between them.

Of course - there is a small chance that something else is going on. But we'll never know without a controlled blind test, or measurements.

 

[boxerfan88]

“wetware between ears” … good one ….

 

[Pareto Pragmatic]

It will only be applied to inputs where the clock is transmitted from the source (SPDIF, Toslink), and the internal clock must be adapted to that.

USB - the internal clock of the DAC is used so no PLL is needed or used.

And I shouldn't need to say it, but your personal subjective impressions are no more meaningful than OP's in the absense of measurements or controlled blind listening.

Thanks for the lesson on DPLL, I learned something today.

No more meaningful, true. Would you say that the difference between min and max settings on DPLL will have no audible effects?

That's a fine null hypothesis for some testing. It's easy to set up, if you have another person who can switch DPLL settings for you. If not, enough up/down button presses to parts of the music will randomize the starting point pretty well. I would expect the change from max to min (or min to max) to be audible.

Yeah, I did random presses to lose my place and then clicked through in sequence up or down. Because I don't trust my brain.

I don't trust my measurements all the time either, but that's because my laptop is literally falling apart.

 

[BDWoody]

Can't do step 3. Don't have any means to record anymore. But the simpler test -- and what worked for me -- was to just do what made the objectionable charactersitics of both DACs that I had been listening to for a couple of years go away

Which will help no one get to the bottom of this.

You certainly may be hearing something 'real,' but it looks like the effort required is just too much for you to be bothered with.

It sounds more like clipping to me than anything, but it looks like we'll never know.

If you want to know why this is happening, take a step up the learning curve to figure this out with the people here who are certainly trying to help you do exactly that. More words are unlikely to get us much further.

 

[antcollinet]

No more meaningful, true. Would you say that the difference between min and max settings on DPLL will have no audible effects?

Almost certainly - unless the jitter performance of the DAC is broken.

I've yet to see one that is.

 

[Pareto Pragmatic]

Why do you think perceptive biases cannot create a difference that is un-subtle?

The placebo effect in testing new medicines is very powerful, that's for sure. In fact, the placebo effect has gotten stronger over time, making it harder to show an effect above placebo. And lots of people rave about great improvements in gear that have no real effect on sound, and that's not subtle in many cases.

Almost certainly - unless the jitter performance of the DAC is broken.

Having now looked at some j-test graphs online of min-mid-max DPLL settings, I can see why you would say "not". And if my ears were not gunked up at the moment, I would retest myself and see if I could replicate my not very formal results from when I got my current DAC.

But maybe you should give it a try, since you said "almost certainly"? Isn't certainty worth 2 minutes?

 

[antcollinet]

The placebo effect in testing new medicines is very powerful, that's for sure. In fact, the placebo effect has gotten stronger over time, making it harder to show an effect above placebo. And lots of people rave about great improvements in gear that have no real effect on sound, and that's not subtle in many cases.



Having now looked at some j-test graphs online of min-mid-max DPLL settings, I can see why you would say "not". And if my ears were not gunked up at the moment, I would retest myself and see if I could replicate my not very formal results from when I got my current DAC.

But maybe you should give it a try, since you said "almost certainly"? Isn't certainty worth 2 minutes?

Not really - since i know the jitter performance of my DAC is not broken.

(the "almost" comes from not having seen measurements of your specific DAC)

But I'd be interested to see those j-test results you have with different DPLL settings if you could post a link.

 

[tjjrx7]

IDK why this is even a conversation these days. It's 2024. Silicon is damn near free. Stick a buffer on the front of the dac so that you can cross clock domains and ignore jitter problems altogether.

But what can I expect. Companies are just now broadly adopting the idea of paralleling dac chips together.

 

[sq225917]

So to summarise. Dpll only affects electrical spdif or toslink inputs. If you think you hear an effect with usb, you're wrong, Imagining it, based on this, without measured samples we can help no further.

Nothing more to be said really. We'd like to help, but ate at the limit of what you've presented, anything further is pointless conjecture.

That said, jitter doesn't present as frequency specific distortions. So the only way it could be jitter is if error correction is badly concealing lost samples through disastrous interpolation. And dacs haven't been that bad for 30 years.

 

[ejr]

Which will help no one get to the bottom of this.

You certainly may be hearing something 'real,' but it looks like the effort required is just too much for you to be bothered with.

It sounds more like clipping to me than anything, but it looks like we'll never know.

If you want to know why this is happening, take a step up the learning curve to figure this out with the people here who are certainly trying to help you do exactly that. More words are unlikely to get us much further.

The only thing I "can't be bothered with" is buying equipment and/or hiring someone to conduct a double blind experiment (i.e. throwing even more time and money to diagnose a problem that no longer exists since I changed the settings). I also marvel at those who, for all I know, have no background in psychology attempting to diagnose confirmation bias or the placebo effect from a forum post. As far as I can tell, no one has the exact same setup as me (though at least one person was able to find a copy of my worst test CD) and no one is listening with my ears. I used to see this kind of thing all the time in my IT sideline, where techs were quick to blame the end user and, nine times out of ten, there turned out to be a genuine problem (although, as often as not, the suspected cause was not what any of us expected).

I am not sure what people are getting at when citing SPDIF connections and USB. I use both DACs with USB and optical connections as follows:

- An ethernet cable goes to my Roku Ultra, which is connected to my TV via HDMI. The TV is connected to the D40 by an optical cable. The output goes to my Genelec studio monitors and subwoofer. No USB connection anywhere in that chain. This is where I was hearing the high frequency distortion (especially with dialog over music in film soundtracks). Changing the DPLL setting to the maximum dramatically improved it. NOTE: I am using the D40 in SP2 mode. I'm not sure if that has any effect on this. (I asked for info about this mode on this forum and never got a response. It seems like I only get responses when someone disagrees with me.)

- The D40 is also connected to one of my computers by USB. I use this connection to listen to music that I created myself on my DAW (all electronically composed and orchestrated with digital sample libraries). This connection never has any type of distortion. I use the same headphones on both computers (for compostition, orchestration, and playback) and what I hear with the D40 is always exactly what I heard when creating the music. Changing the DPLL setting had no effect on this.

- The D30 is connected to my iPad via USB and to my CD player via the optical input. The output goes to my Emotiva preamp and Genelec home theater speakers. This was where I was hearing the worst distortion (though I have to wonder if I was simply hearing the differences in quality between the D30 and the D40). It was there when playing CDs, but not as bad as the USB connection (which, like the Roku on my other system, streams from Amazon Unlimited). The connection to the Internet is via wi-fi.

 

[antcollinet]

I also marvel at those who, for all I know, have no background in psychology attempting to diagnose confirmation bias or the placebo effect from a forum post.

No-one is trying to diagnose it.

But most people here accept that perceptive bias is an issue in sighted listening - and further understand it is not an exception - it is the norm (it is how our senses work). While no-one can be certain what you are hearing is bias, we can know:

1 - if you are hearing changes where non exists (eg jitter adjustment on USB) then it is not the sound waves you are hearing a difference in.
2 - if you are hearing changes where measurements show us such changes are below the level of audibility - then it is vanishingly unlikely that it is coming from the sound waves.

In both cases the most likely explanation is perceptive bias. Ruling that out would require measurements or blind test.

 

[Koeitje]

3. Then, I listened to the same selections, using the built-in DAC on the CD player. It was very hard to distinguish between the output of the D30 with the DPLL value at Maximum from the Denon DAC when I flipped between them. I couldn't say whether I liked one better than the other. IMHO the Denon DAC does well with recordings like this (symphonic, where all the instruments are in one big studio or hall and tend to blend together) and less well with contemporary music or jazz combos with fewer instruments, recorded track by track, with potentially a lot of separation between them.

This is pointless, because chances are the output levels between your CD player and the D30 differ.

I also marvel at those who, for all I know, have no background in psychology attempting to diagnose confirmation bias or the placebo effect from a forum post. As far as I can tell, no one has the exact same setup as me (though at least one person was able to find a copy of my worst test CD) and no one is listening with my ears. I used to see this kind of thing all the time in my IT sideline, where techs were quick to blame the end user and, nine times out of ten, there turned out to be a genuine problem (although, as often as not, the suspected cause was not what any of us expected).

Ok good, because I dabbled a little in psychology when at university. Anyway. the issue is not that no one is listening with your ears. Because I fully trust your ears to pick up sound. What I don't trust is your brain processing that information, and please don't feel insulted by that because I also don't trust my own brain. The brain is a fickle thing and very easily fooled.

Can I also take a quick guess at how those IT problems went? Someone complains, IT doesn't believe it, then the end-user actually demonstrates the problem and IT believes them. We are waiting for you to demonstrate the problem. But you aren't. So what now?

 

[Sokel]

So to summarise. Dpll only affects electrical spdif or toslink inputs.

No,it also affects I2S so USB too,depending how the clocks are made.
I described above a way to test it if the DAC allows it and has all options unlocked,just set it to the tighter setting for DSD and play DSD512 for example,or the same for PCM and play 768kHz something (ESS allows separate settings for these) ,that with USB input.
If the conditions are not ideal they just go broken.

 

[Blumlein 88]

No,it also affects I2S so USB too,depending how the clocks are made.
I described above a way to test it if the DAC allows it and has all options unlocked,just set it to the tighter setting for DSD and play DSD512 for example,or the same for PCM and play 768kHz something (ESS allows separate settings for these) ,that with USB input.
If the conditions are not ideal they just go broken.

According to the manual for the DAC it only effects SPDIF.

- The D30 is connected to my iPad via USB and to my CD player via the optical input. The output goes to my Emotiva preamp and Genelec home theater speakers. This was where I was hearing the worst distortion (though I have to wonder if I was simply hearing the differences in quality between the D30 and the D40). It was there when playing CDs, but not as bad as the USB connection (which, like the Roku on my other system, streams from Amazon Unlimited). The connection to the Internet is via wi-fi.

All of which you posted might work the way you are saying, but not the part I quote above. The USB and wifi path should be uneffected by the DPLL setting at all. You could do the test Sokel describes to see if the manual was in error, but it seems unlikely.

In the end, if setting the DPLL makes you happy and not hearing distortion then do that and don't worry about it further.

 

[voodooless]

No,it also affects I2S so USB too,depending how the clocks are made.

Sure about that? This is typically a setting of the SPDIF receiver chip. With USB, this is not a factor, it will have audio clocked by a crystal.

No-one is trying to diagnose it.

We don’t have to. The OP is human, bias is inevitable. Any experiment not accounting for it is unreliable. It’s as simple as that.

 

[Sokel]

Sure about that? This is typically a setting of the SPDIF receiver chip. With USB, this is not a factor, it will have audio clocked by a crystal.

Yes,I have it tested 100 times.
Here's the notes of the ESS controller I use:

Of course not all DACs have all the options unlocked and it depends on how the clocks are made.

 

[Pareto Pragmatic]

Not really - since i know the jitter performance of my DAC is not broken.

(the "almost" comes from not having seen measurements of your specific DAC)

But I'd be interested to see those j-test results you have with different DPLL settings if you could post a link.

So, you take theory over an easily done experiment? So noted, but I would spend 2 minutes on a lark in a similar situation, were the situation reversed. I would not spend more than that on generating a disconfirming data point in such a situation though.

The j-test I don't have. Google image search of something like... jitter DPLL settings..., quick scan (once I found an example), no attention to methods just the output. So sorry not to have those handy. Generally, from what I remember, -140db and graphs that looked identical at as glance. But again, I did not look at any methods, or text, just graphs.

(google google google)

"As expected, the jitter is higher in level than asynchronous USB. Notice again that changing the "dP" level made no difference to the J-Test spectrum." http://archimago.blogspot.com/2021/01/reviewmeasurements-smsl-m100-mk-ii-hi.html Evaluate as you see fit.

If I could run REW into spdif I would do some tests to see if what I heard in the past shows up as distortion, but I am limited to USB at the moment.

 

[Sokel]

I wonder why I didn't think of it earlier,someone should have told me into it you guys

So,here we go,USB input,Khadas TB+ESS controller running PCM,J-test24:



Mu usual setting,Default Low (look at the note above,that's a 5)




That's with the lowest setting (NOT 0 though)




That's with the highest (15 of the above note)




...and that's all together for easy view.

 

[ejr]

This is pointless, because chances are the output levels between your CD player and the D30 differ.


Ok good, because I dabbled a little in psychology when at university. Anyway. the issue is not that no one is listening with your ears. Because I fully trust your ears to pick up sound. What I don't trust is your brain processing that information, and please don't feel insulted by that because I also don't trust my own brain. The brain is a fickle thing and very easily fooled.

Can I also take a quick guess at how those IT problems went? Someone complains, IT doesn't believe it, then the end-user actually demonstrates the problem and IT believes them. We are waiting for you to demonstrate the problem. But you aren't. So what now?

I'm not because I can't. I don't have anything to measure with or record the output.

I did not post here because I wanted help or an explanation. I posted because, after three years of enduring this issue, I found a solution and wanted to pass it along for anyone who is having a similar problem with eithe of these two DACs. (I suppose if others misinterpret that, it could be considered a similar psychological issue by those who are inclined to see things that way.)

 

[Koeitje]

I'm not because I can't. I don't have anything to measure with or record the output.

I did not post here because I wanted help or an explanation. I posted because, after three years of enduring this issue, I found a solution and wanted to pass it along for anyone who is having a similar problem with eithe of these two DACs. (I suppose if others misinterpret that, it could be considered a similar psychological issue by those who are inclined to see things that way.)

You don't have proof of that at all.