Ji-Ji-Ji-Jitter

[Blumlein 88]

Okay, here's what I did:

1. First of all, I noticed what JustJones said above when I went to change the DPLL setting to zero. It actually says on the Loxjie D30 menu that lowering the value lowers jitter. So, I guess, what I have been doing is raising the jitter (or the setting that is supposed to address it) rather than lowering it (unless Loxjie implemented this feature backwards!)

2. Next, I listened to about 5 minutes each of the first and second movements of Beethoven's 9th symphony on my Denon 600NE CD player: first with the DPLL setting on the D30 at zero (minimum or no jitter), then with the DPLL setting on Maximum. There was less distortion on both than with the Wi-fi signal streaming from Amazon music -- but, that said, the Maximum setting was better. Setting it to zero made it sound a bit muddy -- not to the point of distortion (which is what the streaming signal sounded like to me) -- but I enough to make me prefer to listen to it at Maximum. It sounded a bit smoother, too (if I can use that word without setting off an argument).

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.

Which brings me back to a post that I made earlier, to the effect that a wider variety of music should be used to evaluate DACs and other audiophile equipment. Specifically, I feel that classical and contemporary music scored for symphonic instruments is under-represented (perhaps because they are harder to record or render?) I mean, really, distortion and noise are often used intentionally in a lot of genres and it seems to me that some tracks would sound just as good regardless of the hardware and settings used to reproduce it.

Seems like a good method to you I'm sure. But it isn't. Do some measurements or you'll continue in a circle of confusion just guessing about what is going on.

Doing more uncontrolled sighted listening longer with more music will get you nowhere

 

[kemmler3D]

Okay, here's what I did:

1. First of all, I noticed what JustJones said above when I went to change the DPLL setting to zero. It actually says on the Loxjie D30 menu that lowering the value lowers jitter. So, I guess, what I have been doing is raising the jitter (or the setting that is supposed to address it) rather than lowering it (unless Loxjie implemented this feature backwards!)

2. Next, I listened to about 5 minutes each of the first and second movements of Beethoven's 9th symphony on my Denon 600NE CD player: first with the DPLL setting on the D30 at zero (minimum or no jitter), then with the DPLL setting on Maximum. There was less distortion on both than with the Wi-fi signal streaming from Amazon music -- but, that said, the Maximum setting was better. Setting it to zero made it sound a bit muddy -- not to the point of distortion (which is what the streaming signal sounded like to me) -- but I enough to make me prefer to listen to it at Maximum. It sounded a bit smoother, too (if I can use that word without setting off an argument).

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.

Which brings me back to a post that I made earlier, to the effect that a wider variety of music should be used to evaluate DACs and other audiophile equipment. Specifically, I feel that classical and contemporary music scored for symphonic instruments is under-represented (perhaps because they are harder to record or render?) I mean, really, distortion and noise are often used intentionally in a lot of genres and it seems to me that some tracks would sound just as good regardless of the hardware and settings used to reproduce it.

I don't want to discount or dispute anything you heard, but there is nothing I can do with or take away from this description, personally.

I understand you changed a few variables here (source, DPLL, music) but I (we) could comment a lot more intelligently if we had the actual output of the DAC to look at in each usage scenario. Short of that all I can say is it sounds like maybe a weird interaction between pieces of equipment, but I can't speculate as to what it means.

 

[DVDdoug]

The best way to describe it is that the sounds suddenly seem very low-fi (like an old transistor radio with a single speaker back in the 1960s cranked up too high,

If you try to get 20 Watts out of a 10W amplifier, that's clipping. It's the most common type of distortion and as you'd expect, it happens only on the peaks or loudest parts.

Clipping can happen if the digital signal or data) "tries" to go over 0dBDS. There are few ways that can happen (digitally). Most commonly it happens when using digital EQ and you boost the bass (or boost something) without compensating by lowering the overall digital level. Or a few software players (like VLC) allow you to adjust the volume over 100%. It's OK to set the volume over 100% if the recording is "quiet" and the peaks aren't hitting 0dB already. And, some "loudness war" recordings are clipped intenentionally to make them louder. Of course when the distortion is baked-into the recording you're stuck with it.

Which brings me back to a post that I made earlier, to the effect that a wider variety of music should be used to evaluate DACs and other audiophile equipment.

And it usually best to use a "known good" recording that you are familiar with. Depending on what you're listening for complex music may be best. Maybe simple acoustic recordings are best for distortion. Or, if you are trying to hear noise, silence. Audible noise is far more common than audible distortion, assuming no clipping.

And... it should also be done blind, and if you are comparing two different DACs, or comparing to a reference, it should be a proper, scientific, level-matched, repeatable and statistically valid ABX test. Amir (our host here) doesn't do blind listening as part of his reviews, but he backs-up what he hears (or doesn't hear) with measurements.

 

[Doodski]

distortion and noise are often used intentionally in a lot of genres and it seems to me that some tracks would sound just as good regardless of the hardware and settings used to reproduce it.

Good point.

 

[Ropeburn]

audiophile digital neurosis

I'm so totally going to steal that term and the whole concept. And use it everywhere whenever appropriate, because it's just that good at describing the "digital is harsh" notion that has been plagueing the audio world since the late 80s.

 

[markstein]

Digital also can be done wrong. Even today. See the discussions about intersample clipping , especially when using SRC for upsampling.

Digital does not automatically mean "good". It needs to be implemented well to be good.

 

[Blumlein 88]

To the OP, is the effect audible enough it would be heard over a phone recording? IF so you could record both versions and upload the file. Maybe some of the people here could tell what it likely is listening to it. As someone suggested it might be easier to detect using a 100 hz sine to detect clipping or dropouts.

 

[Roland68]

I've mentioned, in more than one thread, that I can hear something which I think is what others have called a digital "glare" (especially as it relates to DACs that use an ESS chip).
What I hear, specifically, is a little too much high frequency that appears to be added to, or over-emphasized in, the highs of any given track. I also notice that individual instruments seem to be more isolated than what you would hear if the same piece were recorded in a studio or hall where the sound of one instrument bleeds a bit into that of the others (especially those closest). I don't consider that latter "glare", but I do feel that allowing some of that bleed-through makes a recording feel less realistic to my ears.

I was also sometimes hearing a distortion in the loudest instruments (or those that have been bumped up to stand out in the mix, like dialog over the music in a movie soundtrack). The best way to describe it is that the sounds suddenly seem very low-fi (like an old transistor radio with a single speaker back in the 1960s cranked up too high, pushing out on top of everything else). That's what bugged me the most. Then, one day, I saw a DAC review where the reviewer advised that all DACs be set to the lowest DPLL value to "improve fidelity". I did so and the distortion becaome worse. So, I tried the opposite (setting it to the highest DPLL value on both my DACs) and that removed most of the disortion I was hearing.

Apparently the thing that most bothered me about my DACs (both of which are Loxjie products that use ESS chips) is the jitter. Removing as much of that as possible has had more effect than changing the filters or sound colors (and there seems to me much less difference between them now).

It would be very helpful if you could precisely name corresponding pieces of music from different musical areas, ideally with a description of when and where what can be heard.

 

[antcollinet]

It's your imagination on all things above.

Allow me to re-state:

It's perceptive biases (a construct of your subconscious or unconscious brain) on all things above.

 

[antcollinet]

Okay, here's what I did:

1. First of all, I noticed what JustJones said above when I went to change the DPLL setting to zero. It actually says on the Loxjie D30 menu that lowering the value lowers jitter. So, I guess, what I have been doing is raising the jitter (or the setting that is supposed to address it) rather than lowering it (unless Loxjie implemented this feature backwards!)

2. Next, I listened to about 5 minutes each of the first and second movements of Beethoven's 9th symphony on my Denon 600NE CD player: first with the DPLL setting on the D30 at zero (minimum or no jitter), then with the DPLL setting on Maximum. There was less distortion on both than with the Wi-fi signal streaming from Amazon music -- but, that said, the Maximum setting was better. Setting it to zero made it sound a bit muddy -- not to the point of distortion (which is what the streaming signal sounded like to me) -- but I enough to make me prefer to listen to it at Maximum. It sounded a bit smoother, too (if I can use that word without setting off an argument).

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.

Which brings me back to a post that I made earlier, to the effect that a wider variety of music should be used to evaluate DACs and other audiophile equipment. Specifically, I feel that classical and contemporary music scored for symphonic instruments is under-represented (perhaps because they are harder to record or render?) I mean, really, distortion and noise are often used intentionally in a lot of genres and it seems to me that some tracks would sound just as good regardless of the hardware and settings used to reproduce it.

The problem here, is that all those perceptions you are describing could be (and likely are) a result of the perceptive biases I mentioned above. (We are all subject to these if we are human - it is how our senses work, and we can't turn them off)

You are changing settings. You are expecting those changes to change the sound, so that is what you hear.

Without either measurements showing an audible difference, or a well controlled blind test to check if you can genuinely hear a difference, it is not possible even to conclude there is an audible difference to be heard.

 

[ejr]

AFAIK it applies to SPDIF input.

The post you replied to says ethernet and wifi : "the signal source coming through an ethernet cable on one (to the D40) and wi-fi on the other (to the D30)."

On my balanced system, the signal originates from an ethernet connection, going to a Roku box, which is connected to the DAC with an optical cable.

On my single-ended system, the signal originates from a wi-fi connection to my iPad, which is connected to the DAC with a USB cable (using Apple's adapater) -- or -- it originates from my Denon CD player and is connected to the DAC by an optical cable.

 

[solderdude]

Why not do what @Blumlein 88 suggested ? Record the output (preferably in 24/192) and you will be taken more seriously and can prove what you are hearing to yourself and others.
It really isn't hard to do but requires a recording device.

 

[ejr]

Sorry, but I don't have the equipment or the expertise to measure and give the data that some are asking for. Just about all my professional life was spent in front of the camera, rather than behind it (although I did work as a cameraman on local cable TV for a bit and an assistant editor for two emmy-winning documentaries while in college back in the 1970s and later developed database software for clients in the entertainment industry as a sideline, so some may have considered me to be more tech savvy than most performers).

I did take almost enough psychology courses in college to qualify for a second minor on my BA (my actual minor was music), so I tred carefully with suppositions that anything is a perceptual problem when there are other explanations. Specifically, in this case, my perception of the issue has been going on for a number of years. I listened to different pieces repeatedly, at different times of the day, when I was in different moods and the objections to the sound I was hearing was always the same. I changed one setting and the problem went away. It's been about a week now and it still sounds better, even when I filp back and forth between other DPLL values. So that makes me think it isn't a perceptual problem.

All I was attempting to do here was share a solution to a problem I had been having with my DACs for several years that no one had been able to solve in case someone else is having the same issue. It's something else that you can try that only one reviewer suggested (and, yes, I'm sorry that I can't remember which review, but if I do I will post it here).

For those who are interested, to perform the tests I did yesterday, I used the following CD (because I considered it a bad recording of a challenging classical piece and therefore likey to show the most objectional problems that I have been trying to address). It's a "Laserlight Digital" budget release of Beethoven's ninth performed by the Hungarian Philharmonic Orchestra conducted by Janos Ferencsik in1988 that I picked up from Best Buy back in the 1990s.

I think any classical piece recorded like this (apparently with just a pair of mics at the back of the hall or studio) is a good test piece when using speakers because it potentially has a lot of ambiance and the individual instruments are not easily separated by the ear. I think the worst test piece would be something like Grace Mahya's "Girl from Ipanema" streaming from Amazon Unlimited and using a good pair of headphones -- because it sounds absolutely great to my ears and I suspect it was engineered to sound terrific regardless of how you play it. Ideally, I'd like to see reviewers using more older or not-so-great recordings, or at least using them in addition to those that are designed to test well. Because if everything you test was designed to measure well on your equipment, IMHO it doesn't really tell you anything.

 

[Doodski]

For those who are interested, to perform the tests I did yesterday, I used the following CD (because I considered it a bad recording of a challenging classical piece and therefore likey to show the most objectional problems that I have been trying to address): It's a "Laserlight Digital" budget release of Beethoven's ninth performed by the Hungarian Philharmonic Orchestra conducted by Janos Ferencsik in1988 that I picked up from Best Buy back in the 1990s.

I think any classical piece recorded like this (apparently with just a pair of mics at the back of the hall or in a studio) is a good test piece when using speakers because it potentially has a lot of ambiance and the individual instruments are not easily separated by the ear. I think the worst test piece would be something like Grace Mahya's "Girl from Ipanema" streaming from Amazon Unlimited and using a good pair of headphones -- because it sounds absolutely great to my ears and I suspect it was engineered to sound terrific regardless of how you play it. Ideally, I'd like to see reviewers using more older or not-so-great recordings -- or at least using them in addition to those that are designed to test well.

You are not making sense. Use the best recording that you can get rather than old distorted recordings.... and yes even you can have expectation bias.

 

[solderdude]

Sorry, but I don't have the equipment or the expertise to measure and give the data that some are asking for.

Oh well... if setting the DPLL to another value helped in your case the problem is solved.
As you say... if that option is there it is there for a reason.
Most DACs do not have it, probably because the receiver used in the DAC does not need it.

Seems like a synchronization issue in combination with your source/connection.

 

[Blumlein 88]

Sorry, but I don't have the equipment or the expertise to measure and give the data that some are asking for. Just about all my professional life was spent in front of the camera, rather than behind it (although I did work as a cameraman on local cable TV for a bit and an assistant editor for two emmy-winning documentaries while in college back in the 1970s and later developed database software for clients in the entertainment industry as a sideline, so some may have considered me to be more tech savvy than most performers).

I did take almost enough psychology courses in college to qualify for a second minor on my BA (my actual minor was music), so I tred carefully with suppositions that anything is a perceptual problem when there are other explanations. Specifically, in this case, my perception of the issue has been going on for a number of years. I listened to different pieces repeatedly, at different times of the day, when I was in different moods and the objections to the sound I was hearing was always the same. I changed one setting and the problem went away. It's been about a week now and it still sounds better, even when I filp back and forth between other DPLL values. So that makes me think it isn't a perceptual problem.

All I was attempting to do here was share a solution to a problem I had been having with my DACs for several years that no one had been able to solve in case someone else is having the same issue. It's something else that you can try that only one reviewer suggested (and, yes, I'm sorry that I can't remember which review, but if I do I will post it here).

For those who are interested, to perform the tests I did yesterday, I used the following CD (because I considered it a bad recording of a challenging classical piece and therefore likey to show the most objectional problems that I have been trying to address): It's a-that I picked up from Best Buy back in the 1990s.

I think any classical piece recorded like this (apparently with just a pair of mics at the back of the hall or in a studio) is a good test piece when using speakers because it potentially has a lot of ambiance and the individual instruments are not easily separated by the ear. I think the worst test piece would be something like Grace Mahya's "Girl from Ipanema" streaming from Amazon Unlimited and using a good pair of headphones -- because it sounds absolutely great to my ears and I suspect it was engineered to sound terrific regardless of how you play it. Ideally, I'd like to see reviewers using more older or not-so-great recordings -- or at least using them in addition to those that are designed to test well.

I am going to be blunt. You are being stubborn.

Ideally, I'd like to see reviewers using more older or not-so-great recordings -- or at least using them in addition to those that are designed to test well.

Reviewers who do reviews listening to recordings and describing sound quality are close to useless. Listening done sighted, not level matched over long periods of time is if not completely useless close to it. You would be surprised how an opinion can last for years from different biasing factors. Get a chance to compare blinded and it all goes away. Disappears completely. I don't know what psychology you are using, but this is an expected kind of result regarding listening.

If you don't have the gear or experience to record, then that is not an option. However, by far the most likely thing going on here is the bias of knowing you are changing settings on something. Your description of what is wrong is vague and non-specific. Exactly like we'd expect.

I understand your sharing a possible solution with us. But it isn't clear there is really anything going on. Especially jitter with more than one source for the DAC. Jitter has to be very high, like a 100 times higher than very poor DACs to be audible.

I listened to the LaserLight CD which I think is the same one. It simply isn't a great recording. A bit shouty, sounds like the low end was overloading the space or the microphones were poorly placed. I don't know if that is a two mic recording or not. Two mic recordings can sound much better than this. They are not automatically going to sound confused. I was using a UAD Apollo Arrow to listen.

Please don't take this the wrong way, but maybe get someone to do a blind test between the two settings with you and see if the difference is there or not. You can confirm it is an audible difference for all the rest of us or learn it is not. No shame either way it turns out.

 

[Blumlein 88]

Okay so looking at the manual about the DPLL setting on the Loxjie it is about what I would have thought. It is only for the SPDIF inputs (optical and coax). It would not have an effect on USB.

PLL, phase locked loops, can filter and reduce the amount of jitter for SPDIF inputs as the clock is embedded with the digital stream. If the stream is stable with lower variance a tight filter can reduce jitter. If the speed of the clock over the SPDIF stream wanders or is not stable then you may get dropouts or other issues when the clock lock falters. In those cases a PLL of looser tolerance allows it to keep lock with a lesser SPDIF stream such as sometimes happen with TVs over Toslink. So those DPLL adjustments are adjusting how tight or loose the PLL is. If you aren't getting dropouts and the SPDIF source is very jittery, then yes one setting might have more jitter than another, but it won't make an audible difference. We just aren't that sensitive to jitter.

This from the manual:

There's a setting called DP with a range of 0-9 to "adjust the internal
DPLL digital phase locked loop circuit Bandwidth, so that the chip achieves a balance
between anti clock jitter and input tolerance."

"When the clock stability of the input signal is good, this value can be reduced,
so that the clock performance of the system is better;
When the clock stability of the input signal is not good, the sound may be interrupted.
Increase this value can avoid the sound interruption!"

From this when you moved the setting to a higher value it would have increased jitter. Unless you were getting dropouts or loss of lock sporadically the lower setting would have been fine. As I think AnalogSteph suggested try this with a 100 hz tone and see if there are dropouts or glitches. You can do this by simply listening to the tone. This setting has no effect, zero if you are using USB, ethernet or wifi for the input as when you do this, the DAC uses its own crystal clock and is not synching to one from another piece of gear.

 

[Digital_Thor]

I thought for quite a few years, that I should chase all sorts of little details, to get the best sound. That turned out to be true, just not those details I was chasing.
Many times when I heard little glares or hints of what I thought was caused by something like jitter - it was actually just a bad recording, bad conversions from one file type to another or poor design of some of my equipment.

Jitter - as I understand it - manifest itself as a tiny amount of background noise - hiss if you will. Like higher frequency mistakes. But not mistakes that alter the sound all together. It's just a slight background noise from behind the technique of digital signal transmissions.

There are so many other things that need to be very well tuned and setup, before I think you could ever hear any sign of jitter, which altogether has to be pretty severe to be even audible.

I recently found "Wings of pegasus" on YouTube, a guy who analyzes voices, so you can hear whether they use autotune or pitch correction. Those two techniques create little sudden shifts in timbre and pace, which sounds unnatural, often also leaving a "digital mark" that sounds like something is wrong, and then making you think you hear digital mistakes like jitter, when in fact, they are intentional techniques with bad side effects.

So, always consider what you think you hear. That is why this forum tries to help us discern between tricking our mind and reality - IMO

 

[Old_School_Brad]

For your benefit ....

what does jitter ACTUALLY sound like?

In a previous post we had an encounter with jitter in audio. The talk about jitter is hard to miss these days. There are plenty of products on the mar...

www.sereneaudio.com

Jim

Additionally, most DACs, I believe, typically have jitter levels below 100 picoseconds. This helps to put the tests on the site into perspective.

 

[antcollinet]

I listened to different pieces repeatedly, at different times of the day, when I was in different moods and the objections to the sound I was hearing was always the same. I changed one setting and the problem went away. It's been about a week now and it still sounds better, even when I filp back and forth between other DPLL values. So that makes me think it isn't a perceptual problem.

Let me take you up on this. Here is an example of a perceptive bias. It is specifically speech related, but it fundamentally works in the same way as many other biases. Some biases may be altered by how you are feeling. Some might change over time - but not all...

You'll note it is consistent. It doesn't matter when or how long since you last listened, the difference in sound you percieve will still alter based on the mouth shape you see. Even when you know how and why it happens, you can't stop it happening.