Skip to main content

I was made aware of "jitter" a while back, and was curious if anyone could tell me how my setup faired in regards to jitter, and also if there's an advantage to using something like Apogee's Big Ben.

Currently I'm going out of my mic preamp's, into an Apogee Rosetta 96k, then out the S/PDIF of the Rosetta and into the S/PDIF input of a Digital Audio Labs CardDeluxe, then recording into Sonar at 24bit 96khz, then back out the CardDeluxe through its balanced analog outs, into a Mackie 1402VLZ, then to my monitors.

Is there a weak spot here in regards to jitter? Aside from that, can the Rosetta's inherent jitter be improved using a clock source such as the Big Ben, and would there be a noticeable difference?

I was also contemplating purchasing a Benchmark DAC-1 and just going out of the CardDeluxe S/PDIF to the Benchmark, then to the monitors. Would this be significantly better than my current setup?

I would appreciate your advice

Thanks --Bri

Topic Tags

Comments

Opus2000 Tue, 08/12/2003 - 15:18

Hi

First off let me introduce myself..I am Opus..moderator of the DAW world section as well as the Apogee tech for Apogee....

I will explain jitter as best as I can for you..

Jitter is a change in the way that the waveform is resampled within the translation period in the A/D section or even the D/A section of any converter. Jitter is a jump in the timeframe of the pulses within this process and not in the actual frequency of the samples. So you could have a series of 9 pulses per second which then turns to 10 pulses second very quickly and not have jitter...BUT if between the 9 pulses per second it varies to 9.5 or 9.002(etc etc etc) that is where jitter comes into play.

If you were to take a word clock output into a scope and measure it you would see the square wave of the frequency being used and if there was mass amounts of jitter you would see the square wave start to look distorted..this is basically a way to measure jitter...but on the digital output you would put it to either a scope or an AP(Audio Precision) machine of some sort.

Obviously Apogee is known for it's very ultra low jitter clocking so you are safe in that regards.
As the A/D is the most crucial stage in your chain it is important to have the best possible signal at that point! If you were to have a weak A/D converter with lots of jitter, even a good D/A converter could not fix it as it is printed into the waveform.

This is not something that can be seen, just measured with the appropriate gear from the digital output of the device in question.

It is also very possible that a D/A converter will have jitter as well and the only way to solve that is to get a better D/A device!

I'm not sure if the Benchmark DAC-1 has a dual clocking stage in it but it's a good thing to look into as it will help buffer and clean the digital signal.

The Apogee Mini-Dac has this and is actually a great D/A unit to get in that regards.

Now, the Big Ben will improve any device as it's jitter is pretty much unmeasurable! It's too damn low! Seriously! You would benefit greatly BUT with the Rosetta you would have to have the Big Ben clock from it as the Rosetta is a master clock only!

This is not such a bad thing due to the fact that the Big Ben actually has what is called ALF(Adaptive Loop Filtering). This is where any clock being fed into the Big Ben is reshaped and cleaned up. This ultimately makes the clock as if it was Big Ben's own internal clock making it the clock being sent out of the Big Ben completely jitter free!

Jitter can indeed cause your sound to be degraded as there are some converters that will happily degrade your audio without care!

I do hope this helps explain or educate you on this matter and if you have any more questions please let us know!

Cheers

Opus :D

Bri Tue, 08/12/2003 - 23:44

Hi Opus, I was hoping I'd hear from you. Thank you very much for your informative reply. :)

So then, would there be a noticeable difference if I went from the Rosetta into the Big Ben, or is the Rosetta fine on it's own?

Also, I was curious. Given your experience with Apogee's products, can you give me an honest opinion on the quality of the Rosetta's converters/how they sound, compared to Apogee's other converters?

Thanks Opus
--Brian

Ethan Winer Wed, 08/13/2003 - 10:12

Bri,

Opus gave you a good explanation, but I'm going to weigh in too with my opinion.

I believe that jitter is a complete non-problem with modern digital gear. An answer to a question nobody asked. A solution for which no problem exists.

Jitter is expressed similar to noise - as some number of decibels below the program material. Typical jitter figures are 110 or more dB below the program. And since it's present only when the signal is present, it is not only way softer than you'll ever hear but also masked by the main signal. Adding even 1/4 dB of EQ boost at any frequency will have more effect on the music than anything caused by jitter.

--Ethan

vinniesrs Wed, 08/13/2003 - 13:33

I would have to agree with ethan. Have you ever heard anyone complain about a noisy recording while music was blasting away at the same time?(I am talking about commercial recordings here :roll: ). I think that this, and other debates like it are important, but more so for the development of better stuff. I think that from one device to the next, marginal differences in one area dim in importance when compared to other more important things, such as placement of mics.
The bottom line is that millions of records have been sold on recordings from black face adats in the early 90's, and no one is refusing to buy them today.
I am not suggesting we ignore this stuff, as it's great to know, but the most important thing to remember here is that the ends justify the means.

Just a thought.

Bri Wed, 08/13/2003 - 17:04

Hey guys, thanks for your input....encouraging.

I guess what it comes down to is that I'm trying to make sure my electronic equipment choices are capable of producing a truly professional quality of sound...relative to the electronics used...converters for example. Especially considering I mostly record acoustic instruments.

Considering my budget/finacial status, I'm a little neurotic about making sure I've made the best decision/compromise. So, do you think the Rosetta was a good choice in that respect?

I'm just not completely sure about the difference between jitter and converters in regards to their negative effects on the incoming signal when hearing things like, "this AD converter sounds clearer, tighter, better dynamic range as opposed to this one, etc., etc." For example, what exactly is the difference between a $1000 2-channel AD converter, and $5000 2-channel AD converter in regards to sound quality?

Thanks --Brian

Opus2000 Wed, 08/13/2003 - 17:43

Well, you have to look at the functionality offered by the device over the other devices.

Some may have more analog functionality combined with some special circuitry(such as Apogee Soft Limit or the like there of!)

When comparing converters you really have to look at the overall operation of the unit more so than the coverters to be honest.

Remember that the converters are only one stage and not entirely the most crucial. It's the design around the converters that make the difference. Depending on what Op Amp you may use or what form of capacitor inline with some resistors to truly open things up.

That coupled with the power supply in terms of filtering out the bad current or noise induced in the line will help make a product work better for you..

Now, in terms of sound over the other Apogee converters vs. the Rosetta?

Well, the AD-800SE, Trak2 and the PSX-100SE are all the top of our lines.

After those comes the regular AD-800, regular PSX-100, Rosetta 2 CH A/D and the Mini-Me.

Now I did not put them in any specific order here..just listed them.

With the new products coming out you can be almost gauranteed that they will sound as good as the other products or even better. Newer technology and newer ways of designing converters help in this fashion!

In the terms of what Ethan and Steve said...absolutely! I'll babble all the technical terms and they'll put it in laymens terms for you! :D

HTH

Opus :D

Ethan Winer Thu, 08/14/2003 - 08:53

Brian,

> I'm trying to make sure my electronic equipment choices are capable of producing a truly professional quality ... Especially considering I mostly record acoustic instruments. <

I assure you the room you record in has 100 times more impact than any amount of jitter. Getting your room acoustics in order should be a far higher priority than just about anything else in your studio. What is more important:

1. jitter noise 120 dB below the music

2. 0.0001% versus 0.001% slew rate distortion

3. a frequency response that varies continually by 20 dB throughout the entire bass range

> when hearing things like, "this AD converter sounds clearer, tighter, better dynamic range <

That sounds like a lot of jaw-boning by people who would likely hear no difference at all in a true double-blind test.

> what exactly is the difference between a $1000 2-channel AD converter, and $5000 2-channel AD converter in regards to sound quality? <

$4,000 less in your bank account.

--Ethan

vinniesrs Sun, 08/17/2003 - 06:57

Well put Ethan!

I have often heard the statement "your studio is only as good as it's weakest link."
In the case of most places I've encountered, the room is by far the weakest link. My own place included. You can get around this to some extent. I mean it is possible to make a decent recording with sloppy acoustics.
In most cases however, I think there are too many gear heads complaining about marginal technical advantages, when it is literally impossible to appreciate the differences from within your environment.

I did a recording a while ago for a band who had a limited budget. I don't have clip on mic's at the studio, and I wanted to use them for this demo because of time constraints.
I rented the mics from a music store in kitchener, I checked the first one to make sure that the clip was present, and returned to the studio. When I opened the rest of the mics I found that no clips were included, except for the one I had checked. I called the store to see if they had clips, and the rentals manager there confirmed that he did.
When I asked why they were not included, he proceded to lecture me on subsonic transmissions from the clip.

My point is simply that many techs like this one, are so strung up on the technical aspects of this field, they can quite easily lose perspective of what is practical in a given situation.
I was doing a four song demo for $400. I didn't need to spend an hour and a half trying to squeeze booms into a super tight drum setup, but thanks to genious boy, I did.

I'm not an acoustics expert as you are Ethan, but I know and understand the benefits of acoustics, and I've found that the majority of those in my area find it to be less crucial.
It must be because flashing lights are too much fun! :D

Ethan Winer Sun, 08/17/2003 - 08:12

Steve,

> In the case of most places I've encountered, the room is by far the weakest link. <

This is for certain, and far too many engineers have no clue how bad their rooms are or how much the room influences everything they do. Arguably much more than which gear they use.

I've been accused of turning every thread into a pitch for room treatment, and that's not entirely wrong. :) But I truly believe this is the biggest problem for most folks.

> I mean it is possible to make a decent recording with sloppy acoustics. <

Sure, just as you can make a decent recording with mediocre mikes, pres, and loudspeakers.

> When I asked why they were not included, he proceded to lecture me on subsonic transmissions from the clip. <

Yeah, much easier than admitting he screwed up!

> It must be because flashing lights are too much fun! <

No question that acoustic panels and foam are not nearly as sexy as electronic gear. Just like motor oil and gasoline are not as sexy as a new Corvette. But without oil and gas, the coolest and most expensive car in the world won't get out of your driveway.

--Ethan

KurtFoster Sun, 08/17/2003 - 10:34

To put the thread back on topic,

Originally posted by Bri:
I was made aware of "jitter" a while back, and was curious if anyone could tell me how my setup faired in regards to jitter, and also if there's an advantage to using something like Apogee's Big Ben.

Currently I'm going out of my mic pre's, into an Apogee Rosetta 96k, then out the spdif of the Rosetta and into the spdif input of a Digital Audio Labs CardDeluxe, then recording into Sonar at 24bit 96khz, then back out the CardDeluxe through its balanced analog outs, into a mackie 1402VLZ, then to my monitors.

Is there a weak spot here in regards to jitter? Aside from that, can the Rosetta's inherent jitter be improved using a clock source such as the Big Ben, and would there be a noticeable difference?

I was also contemplating purchasing a Benchmark DAC-1 and just going out of the CardDeluxe spdif to the Benchmark, then to the monitors. Would this be significantly better than my current setup?

I would appreciate your advice

Thanks --Bri

Bri,
I myself have heard the difference good clocking can make with Blackface ADATs. Once by clocking them to a Yamaha 02R console, and once by clocking them off an Apogee PSX 100.. good clocking does make a difference IMO. Almost any experienced recording engineer will agree. It is only the occasional "heretic" that will dispute this... Kurt

anonymous Sun, 08/17/2003 - 12:39

Opus, maybe you can tell me what I was hearing when I a/b'd my midiman audiophile converters against the Korg D1200 tabletop recorder.

I kept the chain identical, recorded to both devices simultaneously on the same take. Same bit depth, sampling rate, etc.

Then auditioned both files from the same D/A on the audiophile.

The files created by the Korg just sounded so much clearer, mostly on vox. But even the guitar tracks were better.

Can jitter account for muddiness? Or is it some other part of the converter?

A buddy and I also did a more informal test with his HD24 and the Korg, and I found the HD24's tracks even cleaner than the korg. We attributed it to the converters, since we tried to keep everything else constant.

What part of the converters could account for this? Am I just hearing things?

thanks

ted

KurtFoster Sun, 08/17/2003 - 13:07

Originally posted by Ethan Winer:
No question that acoustic panels and foam are not nearly as sexy as electronic gear. Just like motor oil and gasoline are not as sexy as a new Corvette. But without oil and gas, the coolest and most expensive car in the world won't get out of your driveway.

--Ethan

Actually those ads you are running in the magazines (Mix, Pro Audio Review, EQ ) look pretty damn sexey to me. (sh-wing!) I have to admit that the mini traps certianly do look much better than foam products which are for the most part butt freakin' ugly.. Foam (usually) charcoal gray, darkens the room and after a few years degrades, crumbles and just get uglier. MiniTraps, sleek, trim, fits into the decore of almost any room.. the good looks last and gives your room that "professional" look..
Yep! Old dogs can learn new tricks..

But I still think good clocking makes a big difference.. ! :D heee heee ... Kurt

Ethan Winer Mon, 08/18/2003 - 09:16

Kurt,

> Yep! Old dogs can learn new tricks.. <

Aw gee, thanks. But now how can I counter with the importance of double-blind testing for jitter? :D

Ted: as far as I know, the only thing jitter contributes is noise, and very low level noise at that. I know some people say they hear all kinds of stuff change when they switch clocks: imaging, clarity, low frequency detail, yada yada. I'll reserve judgement until I've personally sat in on a double-blind test. But I can think of no logical reason why normal amounts of jitter - typically 110 dB or more below the program - could have any audible affect.

--Ethan

KurtFoster Mon, 08/18/2003 - 11:09

Originally posted by Ethan Winer:
Kurt,

But now how can I counter with the importance of double-blind testing for jitter? :D
--Ethan

Ethan, If you have any data from double blind tests that supports the contention that jitter is in audible, I would be happy to consider it and perhaps change my position. I acknowledge that the "placebo effect" can influence ones impressions. But until I am shown that my ears are fooling me on something, my tendency is to believe what I hear.. :tu: Kurt

anonymous Mon, 08/18/2003 - 11:15

Ethan, Opus or anyone:

Is there anything else beside jitter that can cause a converter to sound clearer?

In other words, what are all the factors/elements contributing to a converter's quality? Is jitter the only one? How about aliasing and all that fancy stuff?

In my test, the difference is significant to my ears, even though it wasn't a truly double blind test. I'm looking to understand why, if it's not jitter...

When I get the time, perhaps I'll post the two files...

Ethan Winer Mon, 08/18/2003 - 12:19

Ted,

> what are all the factors/elements contributing to a converter's quality? <

All of the usual [audio] suspects, plus maybe one or two more specific to digital audio. So there's frequency response, THD and IM distortion, slew rate limiting which is just another type of distortion, noise - several types - and [digital only] ringing caused by filters, and jitter. There's also phase shift, but I've never seen or heard evidence that phase shift alone is audible. I probably missed a few things.

The real problem is that all of this stuff can be measured easily enough, yet some people claim to hear degradation even when the measurements show that should not be the case. Which is why double blind tests are so important. Not that you aren't hearing something wrong. Maybe I should add to the list above that sometimes gear breaks, so even if the stated specs are fine, that particular gear may not be meeting its specs.

--Ethan

[ August 18, 2003, 02:20 PM: Message edited by: Ethan Winer ]

Opus2000 Mon, 08/18/2003 - 16:03

Indeed jitter does not attribute to sound differences in any way. It's all in the way that the clocking is shaped and how it's response is in the long run.

As Ethan stated there are differences of the circuitry around the converters such as Op Amps and resistors.

Certain resistors will indeed cause some coloration or lack there of. Also some Op Amps may have a higher DC offset than others causing some problems in the audio.

The cheaper the converter the lower the quality will be. There are factors in the designs of converters which do indeed change the characteristic of the sound.

Now, we were blown away here at Apogee when people said that the Big Ben clearly changed the sound of everything. We attribute this to the clocking having a "bell" curve, thus coming up with the idea of the name Big Ben as it is what the Big Ben is! A big bell for the clock itself!

This bell curve is very smooth and does not spike or randomly change at variations.

This can attribute to a very clean sound.

So in the long run the clock has a crucial stage, which helps in the jitter level, but again as we stated here, jitter has no change in the audio quality of the way we hear it.

HTH

Opus :D

anonymous Mon, 12/08/2003 - 22:07

as far as I know, the only thing jitter contributes is noise, and very low level noise at that. I know some people say they hear all kinds of stuff change when they switch clocks: imaging, clarity, low frequency detail, yada yada. I'll reserve judgement until I've personally sat in on a double-blind test. But I can think of no logical reason why normal amounts of jitter - typically 110 dB or more below the program - could have any audible affect.

I've sat in on double-blind tests and heard the difference a good clock (specifically, an Aardvark) can make (I also think I heard an improvement in the converters in my Tascam DA38 when I added an Apogee Rosetta to my own setup, although I never bothered to test that). I didn't hear the difference as noise as much as the things you described others as having heard...mainly in terms of a sense of space, both front-back and left-right. As I understand it, what I was hearing wasn't jitter noise per se, but the smearing of high frequencies due to "slop" in terms of samples not happening precisely every forty-four thousand, one hundredth of a second. It makes sense, since that's where we derive many of our spatial cues from, and it makes sense why people hear "tighter" bass as well since high-frequency overtones play a big part in our perception of low-frequency "detail". Regardless of why I heard this difference, though, I herad it, and the only variable was the clock.

Is there anything else beside jitter that can cause a converter to sound clearer?

Sure, just about anything...filters, power supply, cheap or broken cables, line noise...

Now, we were blown away here at Apogee when people said that the Big Ben clearly changed the sound of everything. We attribute this to the clocking having a "bell" curve, thus coming up with the idea of the name Big Ben as it is what the Big Ben is! A big bell for the clock itself!
This bell curve is very smooth and does not spike or randomly change at variations.
This can attribute to a very clean sound.
So in the long run the clock has a crucial stage, which helps in the jitter level, but again as we stated here, jitter has no change in the audio quality of the way we hear it.

I'm a little confused here...are you saying that Big Ben does or does not improve the quality of converters that are clocked to it?

-Duardo

Ethan Winer Tue, 12/09/2003 - 06:47

Duardo,

I'm surprised to see you weigh in with that opinion, but I know you're not easily fooled by pseudoscience so I'll give you more slack than I'd give most others. :D

> As I understand it, what I was hearing wasn't jitter noise per se, but the smearing of high frequencies due to "slop" in terms of samples not happening precisely every forty-four thousand, one hundredth of a second. <

Any deviation that small/fast is by definition supersonic, so how could it be audible? I wish I had been present at that test. You say you have a high-quality clock but haven't tested it yet. You're near me, right? Let me know if you're willing to test it because I'd like to be there.

--Ethan

Guest Tue, 12/09/2003 - 07:36

Jitter is real...it is measureable, and the distortion artifacts it creates are audible. I understand the argument that while compared with anomalies in room frequency response, worrying about jitter at the picosecond level may seem like fussing over an insignificant detail. But, room modes are primarily a low frequency, and low order effect...they don't generate any significant high order harmonic distortion, wheras jitter can. The fact that people can hear (and obsess over) the effects of changing clocks, sync sources, or converters demonstrates that the human ear is sensitive to the type of distortion artifacts that jitter introduces.

Also, remember that the THD spec by itself is pretty meaningless...you need to know the spectrum of the distortion to be able to come to any conclusions. If the distortion is largely 2nd harmonic, then odds are we won't complian if THD is high, but if it's 7th harmonic listeners will complain even at very low levels.

Anyone doubting the effects of jitter should check out the following links.

Link removed

This one is good as they show the effect of jitter on antialias filter stopband attenuation (at 26kHz and 100kHz...note that instruments like cymbals still have much energy up there even though we can't really hear it, if it aliases that will be audible).

Link removed

This one shows the distortion vs freq for an unnamed converter with jitter vs, a Benchmark converter with jitter reduction. There's something like 60dB difference in distortion at 1kHz. Pretty significant IMO!

http://www.nanophon.com/audio/jitter92.pdf

Cheers,

Kris

Ethan Winer Wed, 12/10/2003 - 10:37

Kris,

> Jitter is real...it is measureable, and the distortion artifacts it creates are audible. <

I'm sure jitter is measurable. And I can't really dispute that it's audible until I've heard a blind A/B comparison. But the notion that jitter is an important facet of digital audio makes little sense to me. All the graphs in the first two links shows jitter contributing content more than 100 dB below the music. That hardly seems important to me. The last link showed what I assume is contrived jitter. That is, they proved you can have some distortion if you force a very high level of jitter. If I misread that text, please direct me specifically to where it shows meaningful distortion levels with typical semi-pro gear.

None of those references convinced me that jitter is a real problem when using decent quality A/D converters or a semi-pro or better sound card. And that's the real issue. Every day someone posts in an audio forum that they're not happy with their amateur sounding recordings. They say they read somewhere that jitter is a big factor, so they ask if they need to buy an external word clock. This is all I'm addressing, not whether it's theoretically possible for extreme jitter to be detrimental.

If jitter really were a big problem, nobody would like analog tape recorders. :D I would love it if someone within an hour or two drive of my place would invite me over to do a direct A/B test comparing a standard semi-pro sound card and the same sound card driven by an expensive outboard clock. I've mentioned previously that I once attended a manufacturer's demo of exactly that comparison. But the demo was so lame, on so many counts, that it showed nothing to me or anyone else present.

--Ethan

anonymous Wed, 12/10/2003 - 20:42

I'm surprised to see you weigh in with that opinion, but I know you're not easily fooled by pseudoscience so I'll give you more slack than I'd give most others.

Thanks! I'm just repeating here how it's been explained to me, and it makes sense to me...and I figure since I heard a difference, and that was the only variable involved, it was either jitter or somethign else clock-related, and I don't know what else it could be.

Any deviation that small/fast is by definition supersonic, so how could it be audible?

I suppose that with enough deviation for every sample...plus or minus however many picoseconds or nanoseconds or whatever...the cumulative effect could be enough to alter the sound perceptibly, maybe? And it is those higher frequencies that affect localization, is it not? All I know is that I heard a difference, and it was not subtle.

You say you have a high-quality clock but haven't tested it yet. You're near me, right? Let me know if you're willing to test it because I'd like to be there.

Unfortunately, it was about a ten hour drive to New York for the AES Convention, so that probably wouldn't be doable. I'll probably be heading out there to visit my sisters next Thanksgiving, and I'd be more than happy to drop by at that time if you're still interested (and if I can convince my wife to let me go for a half a day or whatever..."Why are you packing that Apogee thing for our trip? Where are you going?"...). Anyhow, you know where to find me if you want me to bring it by next year.

I don't have a dedicated clock, I have an Apogee Rosetta converter, which is supposed to have one of the better clocks out there in it. I certainly do hear a difference between the converter itself and my others, but I just haven't bothered to A/B my other converters clocked to it to those same converters clocked internally.

None of those references convinced me that jitter is a real problem when using decent quality A/D converters or a semi-pro or better sound card. And that's the real issue. Every day someone posts in an audio forum that they're not happy with their amateur sounding recordings. They say they read somewhere that jitter is a big factor, so they ask if they need to buy an external word clock.

I don't know if I'd consider it a "problem" per se, but it does seem to me to be one of the factors that makes one converter sound better than another. The difference between having my system clocked to the Apogee and clocked to its own converters (those in a DA38 and a MOTU 2408mkIII) is certainly more subtle than the differences between just about anything else...again, I think I've heard it but I've never tested it, and it's hard to separate those differences from just the clocks and the converters themselves, since I track everything through the Apogee unless I'm doing more than two tracks at once. As far as converters go, I'm sure things like noise, harmonic distortion, and frequency response make a much bigger difference than jitter, but I do think that jitter makes a difference as well, and that it can be audible, depending on the converter and the clock. Just one small component of the whole.

-Duardo

Ethan Winer Thu, 12/11/2003 - 11:15

Duardo,

> I figure since I heard a difference, and that was the only variable involved, it was either jitter or something else clock-related <

I wasn't there so I can only repeat that it makes no sense to me. I really do want to get to the bottom of this, which is why I attended the clock demo last year that I mentioned.

> I suppose that with enough deviation ... the cumulative effect could be enough to alter the sound perceptibly, maybe? <

I really don't see how. Those deviations are up in the RF frequency range!

> it is those higher frequencies that affect localization, is it not? <

Probably not that high!

> it was about a ten hour drive to New York for the AES Convention <

Ah, I assumed you were nearby.

--Ethan