jitter

Jitter is created in D/A or A/D conversion by uneven/jittery conversion clock.  Jitter produces added noise and once conversion is completed it becomes permanent.  We can prevent D/A conversion jitter few ways before conversion, but in case of A/D conversion clock has to be perfect to avoid jitter artifacts.  Analog recordings digitized with jittery A/D clock cannot be fixed and the only remedy is to digitize again, if analog tapes still exist.

Jitter during recording is different. It shows up as sample errors and that gets baked into the recording. I imagine it would show up as Harmonic Distortion or, based on a quick reading, reduced signal to noise.

What I mean by sampling error is that the value that is written down won't be correct.  For instance if the ADC would record 2.0 V at T0 with a perfect clock, with ADC jitter it might record 1.99997 or 2.0004 instead.

As far as I know it doesn’t produce the same noise side bands that jitter in playback does.

Still, both should be minimized for the best musical recording experience. :)

Jitter is a non-issue solved decades ago! Present day DAC's and streamers are immune to jitter. No need for re-clocking devices, despite what the neurotically obsessed will claim.

No device is completely immune to jitter, but it did get remarkably better after the year 2000.  Any device can be overwhelmed by upstream jitter, and testing with medicore reclockers have shown this.

In general I will say that a good streamer with multi-second buffers should have vanishingly low jitter to add to a DAC's output.

No current DAC can remove jitter that was embedded in the file by the ADC when the recording was made. This is the same as trying to "fix" a recording where the wrong mics were used or poorly placed, or bad mixing, balance, EQ and other decisions were made during production. The latter items impacted many analog recordings beginning decades before digital recording was used.

The best a DAC can do is not add any more jitter and make the situation worse.

One can use tone controls or EQ to improve things, but you'll never totally fix a poor recording any more than you can unscramble a scrambled egg.

I suspect at some point a sophisticated AI could improve things, but even then, it would still be a guess as to what should have been done in the studio years or decades ago.

Hopefully though, we can enjoy the music itself for what it is rather than obsessing over recording quality and technical what-ifs. Good stereos are nice and I certainly enjoy mine, but I can still have a lot of fun and enjoyment listening to a good tune on the car's factory radio. For those of a certain age, think about al lthe great times you had as a teenage listening to the top-40 tunes on your AM radio.

Jitter results from timing errors both in recording and reproduction. Top studios therefore run superior clock sync. In reproduction clocking (Ocxo or rubidium) can make a major difference if the dac allows for it. This particularly applies if dac and server are linked via asynchronous USB. The assertion that modern dacs don‘t create jitter is non-sensical; stanalone quality master clocks alone cost in excess of $7K, those are not contained in $3-5K Dacs.

"Jitter is a non-issue solved decades ago! Present day DAC's and streamers are immune to jitter. No need for re-clocking devices, despite what the neurotically obsessed will claim."

*sigh*

Your ignorance is absolutely monumental, @jasonbourne71 

Anyway, back to the OP's question.....you can't fix/remove jitter if it is embedded in the original recording, but you can certainly minimize it during reproduction in your system by using quality equipment.

my new DAC exhibits dropout on a few HD tracks that my old DAC does not. The manufacturer of the new DAC says it is from excessive jitter being received from my streamer, and will adjust the buffer....I will see if this is correct when I try a new streamer next week....I also wonder if  the buffer is adjusted will this affect  the overall sound on  tracks that do not have dropout?

I presume the reason the oldDAC does not have dropout is a "better" buffer?

I suspect adjusting the DAC's buffer will fix your dropout problem, but I'm rather skeptical that the issue is "excessive jitter" from the streamer or recording.

I don’t think jitter causes dropouts. Are you using WiFi?  If you are, that’s more likely the issue I think. 

@mlsstl , +1.

Good answer.

Coralkong, if you remember audioengr Agon member, has reclocking box to solve this jitter issue?

I don’t think jitter causes dropouts. Are you using WiFi? If you are, that’s more likely the issue I think.

Networks have jitter, but it’s defined differently than what we think of in audio. It’s variance in packet timings and if bad enough, can certainly overwhelm a DAC’s jitter reduction/buffers. If it's bad enough though it's called "packet loss" instead. I usually have a good Internet connection but I still set my streamer’s buffer relatively high to 10 seconds or more to avoid issues.

If on Wifi, use a free Wifi analyzer for your PC or phone to make sure you have a strong signal that isn’t shared by many other router signals.

BTW, recording studios use Master Clocks to synchronize multiple ADCs or DACs together. Not to reduce jitter.

As I understand it, while external clocks are excellent the distance due to the cable lengths actually can increase jitter vs. an internal clock of the best DACs today.  The internal femtoclocks used today sit right next to the DAC so don't have the same distance issues.  This is similar to the need to keep RAM right next to the CPUs.

It’s quite possible that an external clock can merely change or worsen jitter performance.

I think it's generally agreed that modern A/D converters sound significantly better than the earliest ones. Whether this is due to jitter is open to debate but something changed for the better. I have a few early digital recordings that sound atrocious and it isn't just the EQ. They are shining examples of poor digital transfer. It doesn't matter what kind of DAC I play them through - they still sound gritty and brittle. 

This phenomenon was one of the reasons that the remastering business took off fairly soon after the CD was released. One example is HDCD releases. Pacific Microsonics made the A/D converters that allowed HDCD encoding as an option. These converters were considered a substantial leap forward and were felt my many studios to be the best available at the time (they are still used by some studios). Some HDCD collectors, myself included, feel that one of the main reasons HDCD remastered discs sound better than their original releases is that the converters were so much better. The HDCD encoding is icing on the cake.

@erik_squires 

Your statement on master clocks is somewhat contestable: interestingly leading dac manufacturers like DCS, MSB or Esoteric offer master clocks for their top of the line stacks. See my comment on the cost/benefit aspects of built-in vs. external clocks.

@antigrunge2 Yes they do. And some have published papers explaining the difference in error between a built-in clock and an external unit is different. You may get a better clock in an external unit, but by the time it’s made it inside the DAC the delta in performance may vanish.

Mytek, for example, recommends an external clock only when needed in a studio setting, but still offer the connection.

My personal feeling is, you should listen for yourself and determine if it’s even different or not and whether that difference is actually better for your tastes.

Further, in the last 20 years the performance of internal clocks has improved so much I can’t justify an external clock anymore.  Redbook performance (44.1kHz/16 bit) has gotten so good that if a DAC sounds significantly better with high rez music I blame the DAC and not the bit rate.

That Esoteric and other offer external clocks is not, IMHO, a sign of superiority of the idea. One could argue that they put an inferior clock in to begin with and then make you pay for the difference you’ll undoubtedly hear.

I have heard dcs Vivaldi with and without masterclock and the difference isn‘t small.  Similarly adding Cybershaft or Antelope clocks to a number of top end dacs (where pos) yields significant improvements. The effect is particularly pronounced in soundstaging and separation of instruments in terms of attack and decay. Sigital is all about timing and the Ed Meitner argument on cable loosses isn‘t born out in my experience.

I currently have Gustard x26 pro DAC with built in jitter control. Gustard brags about their external clock which is almost the price of the DAC. Is there any need for that and will it enhance the sound quality in any way, even if Im not experiencing jitter problems ?  Does my question make sense?

Sadly, this Stereophile review of a dCS Rossini did not do comparative tests.

https://www.stereophile.com/content/dcs-rossini-player-rossini-clock-measurements

If we had, then we'd be able to see how the bare unit compares to other bare units and how much the clock improves or changes the results.   It is interesting to compare the jitter performance to the $1,999 Benchmark DAC, which costs 1/4 of the dCS clock alone. 

While personal preferences trump all measurements, I have trouble thinking of an external clock of $8k as being a good value.

@bobbyloans 

There is actually a simple way to find out: buy a cheap $100 clock on alibaba and if that is an improvement, the more expensive clock will be more so. The Gustard from memeory takes a 10mhz master clock, those are broadly available.

@jw944ts

Sampling clock jitter is inherent to the digitization process so it can't be removed by anything in the playback chain. Here's a good overview from Texas Instruments on how sampling clock jitter affects ADC performance - https://www.ti.com/lit/an/slyt338/slyt338.pdf?ts=1698696178966&ref_url=https%253A%252F%252Fwww.google.com%252F

Practically speaking, given today's ADC performance, clock accuracy and the relatively low input frequencies of audio vs. sampling a signal at radio frequencies, this type of jitter can most likely be considered inaudible.

For example in the paper linked above, the 'average' jitter is calculated at 26 ps which equates to spurious frequency components that are ~100 dB below the RF carrier frequency. If we translated this to an example at audio frequencies, even with music playing at rock concert levels (110 dB), the jitter would still be quieter than a whisper (~10 dB). If we also account for the fact that we now have sampling clocks with femtosecond accuracy (as in the case of RME's SteadyClock technology), the jitter components will be below the threshold of perception in virtually all cases.

@yage thank you for  this information and explanation. How then would you likely

explain repeatable "dropout"when streaming only certain HR tracks. The drop out occurs in similar places each time, and on one DAC, but not another?

In the other thread you mentioned that you were going to try a different a streamer. I also suggested trying a different digital input on the DAC. Did you ever test those situations?

Finally, can you tell if the ’glitchy’ DAC actually loses connection due to jitter? For example, on the Benchmark DAC3 B, the input light will flash if sync is lost. Does the other DAC give you some sort of indication that the connection is bad?

@yage,  the only indication would be if power is interupted...we shall see what happens with a "better" streamer....the manufacturer of the DAC in question has created a fix, which I believe increases some sort of buffer to accept higher degrees a jitter

Hopefully that fixes things. Given what you've posted, I think most signs are pointing to something going wrong on the DAC since replacing it with another DAC solves the issue.

+1 @yage....manufacturer is sending an updated model which will have an way to adjust for jitter in the input

>> "manufacturer is sending an updated model which will have an way to adjust for jitter in the input"

That's my story and I'm sticking with it!

No.

Guess one needs to add a "satire" indicator to one's posts....