Jitter, whats it sound like?

[radiob]

Earlier and now I am listening to Lynard Skynards greatest hits. The the songs in the middle of the disc appear to have a slight delay in the 2 channels of music, sounds almost like a quiet, perfect little "echo" in the vocals, particurally the song "working for the MCA" is it the disc? Or, because it is worse at louder volumes, is it Jitter? Rob, Mark if you read this, I read your PM, and I will be contacting you shortly.Thanks for your response, greatest customer service out there.

[sunnysal]

nothing to do with jitter, however you oculd be hearing the effects of "print through" on the master tape...when a tape is stored the magnetism can pass from one layer to the next on the tape roll creating an echo...tony

[Ray Garrison]

yep, like Tony said.

Here's a PDF file from the AES.org website that talks about it.

As to "What does Jitter sound like?" We have this from the Apogee Digital website:

"What does jitter sound like? Amongst other things, jitter interferes with the brains ability to experience a stereo soundstage to gain an impression of the relative positions of instruments when a recording is played back. Jitter smears the audio soundstage. The sense of width and depth is skewed and narrowed or even lost altogether because the arrival time of individual samples is being smeared across time. Jitter can be produced in a variety of ways. It may be the result of bad analog design, electromagnetic interference getting into digital audio interconnects, the wrong kind of digital audio cable (see Apogee WydeEye literature), and other culprits. The problem is that once a jittery signal is recorded, thats it its thatway forever."

Note that they're talking about the recording chain. CD Players, digital receivers and anything else with a digital filter / DAC chain can introduce jitter on the playback end.

[Audio Flynn]

I really have to get to bed, it is openeing day of pheasant season in Michigan tommorrow.

Great subject!

Jitter causes fatigue.

Wow and flutter in anlaog is less fatigueing to me.

I gage it by how many CDs in a row uninterrupted i can listen to at one time.

Jitter is bad if I can only get to 2.

Jitter is very low if I can listen to 6-8 Cds in a row.

You cannot measure jitter at home.

You can develop a frame of reference for fatigue.

Not being able to identify where the instrument in the soundstage is a great point also.

My soundstage has alwys been better since I went back to tubes 2 years ago.

[D-MAN]

I remember hearing that tape "print-through" on alot of old vinyl albums back in the old days, especially at the very beginning of the cut...

Howver, isn't digital jitter at such a high frequency that the human ear cannot detect it? What is the difference between jitter and dithering? Is is just that one is intentional and the other is not?

DM

[sfogg]

" Howver, isn't digital jitter at such a high frequency that the human ear cannot detect it?"

Jitter can cause noise in the audible frequency range. The problem is jitter is something of a red herring. If a person finds fault with something in the digital world many claim that is because of 'jitter' without really having a clue if this is the case or not.

"What is the difference between jitter and dithering? Is is just that one is intentional and the other is not?"

Jitter is timing errors on the digital bitstream(s) and/or clock signals which cause bits to be misinterpretted.

Dither is basically a low level noise that is added to the source to try to improve DACs linearity/reduce quantization noise down near the noise floor of the DACs or the source.

Shawn

[D-MAN]

Thanks, guys.

DM

[Shock-Late]

Jitter is the main cause of digital "harshness". You can't really hear it by itself, but when it's there, the sound is unnatural and fatiguing.

[sunnysal]

Ray, thanks for the article. I remember "back in the day" at electric lady and later in ole hollywood we used to store tapes "backwards", that is to say tail out, so that print through would not result in echo (which is much eaier to identify while listening), it just resulted in spurious background noises. warm regards, tony

[3dzapper]

Thanks Mark.

And to think I always thought jitter was Don Knotts hands when he was a "man on the street" on the steve Allen Show.

Rick

[greg928gts]

So what does an anti-jitter box accomplish? I have one, but I have no idea what it really does! I bought my used CD front end together as a package and I've been very happy with the performance. The guy I bought it from seemed to know all about these things.

When I first got this package, I tried it with and without the anti-jitter box and I couldn't notice any difference.

Greg

[artto]

Jitter sounds like frequency intermodulation distortion (doppler distortion), causes a loss of timbre and inner detail, loss of resolution.

[maxg]

I have lost count of the number of explanations of jitter that I have read. There seems to be some variation in the explanations of what causes it but general consensus on the effect - which is timing issues.

From the little I understand of the technology involved we have (in the case of a CD player) a disk turning at constant angular velocity that is read by a laser, passed through various functions (like error correction) and out to a digital to Analogue converter.

I am not sure where the clock comes in but I believe it is placed before the conversion. As I understand it there is no buffering in the system and so the timing of the sound ouput is very much dependent on the disk turning at the correct speed. Unless I missed my mark that means that potentially the transport system is actually the greatest source of possible jitter, or timing errors.

The confusing thing is that although the greatest source of jitter might be the transport this effect should, logically, be applied equally to both channels and therefore not necessarily the most audible source.

My own sneaking suspicion (based on nothing other than gut instinct) is that the error correction system might actually be a major source of jitter in and of itself - one that in theory at least - would be very difficult to get around. Of course one would expect that this would result in Jitter being more of an issue of some disks than on others - something I have never read about.

All of the above has been culled (rather dangerously) from a range of articles I have perused on the subject. It seems that jitter is only a problem for audio and not for data (I dont know if there is a jitter issue for video) as presumably a computer will wait as long as it takes for the data to arrive.

As to what jitter actually sounds like - if it is occuring unequally in each channel then the descriptions in this thread of smearing of the soundstage seem logical. If it is happening equally in both channels, however, I personally doubt it would have a major impact on the listening experience at the timescales we are talking about.

I wonder if jitter applies more, less or at all to MP3?

[sfogg]

"If the clock pulses are exactly on time and equal, the waveform on scope will be "rock solid and unmoving." When slight deviations of the clock pulse width happen, the leading edge becomes "fuzzy" as it is actually moving (quickly) back and forth, thus exhibiting a 'jittery' quality. Hence the name."

Further to what Mark is saying the signal feeding to a DAC is typically composed of basically four seperate signal lines. Three of them are basically clocks with the last being the actual digital signal data plus additional bits for pre-amble and such.

The first clock is a master clock that runs at some multiple of the sampling rate. 128x,256x,384x and 512x Fs are common but there are others too. And there are some DACs which don't have a master clock feeding them.

The second clock is the bit clock. This transitions from high to low for every bit 'field' in the data stream.

The third clock is the left/right data clock. This transitions every 32 bits to signal if the data is for the left or the right channel.

The digital data stream is where the music is actually stored alone with additional information. This is 32bits long per channel per sample. The digital data portion of this section is typically between 16 to 24 bits long. How the digital audio data is fit into the 32 bits can vary as there are 3 different formats that can be used. On the DACs I've tested 16bit audio still uses the full 24bits of digital audio data space, just that the 8 LSBs are zeros all the time.

Jitter is any sort of timing problem between these signal lines. If the bit clock was perfect square waves for example but the digital data stream was getting shifted back and forth around somewhat in time that is still jitter as it is a timing issues between the clock and the data.

Picture can be worth a thousand words on understanding how this fits together so here is one I took a few years back...

Top trace is the L/R clock.

Middle trace is the digital data itself.

Bottom trace is the bit clock. I don't show the master clock in this picture.

For more info on how digital data is fed to a DAC read this:

http://www.cirrus.com/en/pubs/appNote/an22.pdf

and this also has some good info too:

http://www.cirrus.com/en/pubs/proBulletin/cs8406pb.pdf

Shawn

[maxg]

I cant quite see how all of this is independent of the transport as Mark said. If the transport is running either fast or slow then the data is out of sync with the 3 clocks. Of course it might be that this issue is not referred to as jitter but as something else - I dont know.

Judging by the importance attached by many manufacturers to the transport mechanisms (especially on the more expensive players) I guess it must come into the picture somewhere....

[fini]

Cripes, I'm just happy it's not some clock in my CD player that keeps blinking "12:00".

[richinlr]

----------------

On 10/19/2004 10:51:43 PM Audio Flynn wrote:

I really have to get to bed, it is openeing day of pheasant season in Michigan tommorrow.

Great subject!

Jitter causes fatigue.

Wow and flutter in anlaog is less fatigueing to me.

I gage it by how many CDs in a row uninterrupted i can listen to at one time.

Jitter is bad if I can only get to 2.

Jitter is very low if I can listen to 6-8 Cds in a row.

You cannot measure jitter at home.

You can develop a frame of reference for fatigue.

Not being able to identify where the instrument in the soundstage is a great point also.

My soundstage has alwys been better since I went back to tubes 2 years ago.

----------------

Interesting! I always tell people at my house that you can start a listening session with CD's and then switch to LP's but, if you start with LP's, you can't later switch to CD's later because you've become accustomed to the LP sound. Fatigue may well be one of the issues here.