You are here

Log in or register to post comments
bgoodjny
bgoodjny's picture
Offline
Last seen: Never ago
Joined: Jul 26 2007 - 4:22pm
jitter

does wow and flutter mean jitter? is wow and flutter the specification to measure jitter? thank you, john

bobedaone
bobedaone's picture
Offline
Last seen: 10 months 4 weeks ago
Joined: Feb 1 2007 - 12:27am
Re: jitter

Wow and flutter are terms used to describe pitch variation in analog playback, commonly caused by a warped record or a mis-centered hole. When the stylus isn't tracking at the proper speed, the pitch varies.

Jitter is a digital phenomenon that describes timing inaccuracies (ie when the data are not played back in a way that is chronologically identical to the recording).

cyclebrain
cyclebrain's picture
Offline
Last seen: 1 month 3 weeks ago
Joined: Jun 16 2006 - 11:40pm
Re: jitter

Never thought about it that way before.
Erics reply said that all of the terms refer to timing variations.
While I do understand analog wow effect on sound, I don't understand the audible effect of jitter. It seems to me that the jitter values are so small relative to the data rate as to be insignificant. As I said, I don't understand and may be way off base on this. Any input?

Jan Vigne
Jan Vigne's picture
Offline
Last seen: Never ago
Joined: Mar 18 2006 - 12:57pm
Re: jitter

Wow and flutter are largely analog problems which have been rendered virtually insignificant in CD playback. To my knowledge no one has linked broadband speed inaccuracies to CD distortions.

However, from almost the very beginning of CD playback, the systems responsible for jitter creation have been discussed as prime candidates for modification in order to improve digital playback. In general it would be fair to say less jitter equals better sound. One problem would be reducing "jitter" is about as effective as reducing "distortion" and equally specific.

http://en.wikipedia.org/wiki/Jitter

http://www.stereophile.com/reference/193jitter/index.html

http://www.stereophile.com/reference/1093jitter/

http://www.geocities.com/jonrisch/jitter.htm

mb01b
mb01b's picture
Offline
Last seen: Never ago
Joined: Aug 8 2007 - 8:14pm
Re: jitter

I have been wondering about jitter lately too. What effect does transferring a CD to a hard-drive, storing in lossless compression, then the player software uncompressing it and sending the signal to an external DAC have on jitter. Is there jitter that would get captured in the original rip from CD? Or in some stage of the playback?

How about when we throw in an Airport Express or similar wireless transmission between the computer and the DAC?

jkalman
jkalman's picture
Offline
Last seen: Never ago
Joined: Sep 20 2005 - 7:04am
Re: jitter


Quote:
I have been wondering about jitter lately too. What effect does transferring a CD to a hard-drive, storing in lossless compression, then the player software uncompressing it and sending the signal to an external DAC have on jitter. Is there jitter that would get captured in the original rip from CD? Or in some stage of the playback?

How about when we throw in an Airport Express or similar wireless transmission between the computer and the DAC?

If the DAC has a memory buffer and reclocks the signal right before the conversion process, then it will eliminate any jitter accrued up until that point. DACs don't usually correct for bit errors, as error correction usually takes place when the music is being read off some type of media, like a CD. Though, you can always test to see if what you plan on using introduces bit errors and likely correct the issue (most likely it would be faulty RAM module in the computer, in which case you would be getting random system crashes anyway, though it is possible to have a corrupted hard drive ruin a file, both are rare nowadays).

Jitter takes place during playback. Until that point it is stored in a memory buffer of some sort (most likely RAM) until it is sent to an Input/Output channel. Before that it is on the HD (another type of storage buffer), but it is reclocked in system RAM before it is sent out somewhere. Nowadays, with many RAM controllers interacting directly with the I/O bus (DMA - Direct Memory Access), there is direct transfer from the RAM along the I/O bus to a soundcard. The soundcard will likely rebuffer the signal again and reclock it as well before sending it out along whatever transmission line you will be using. Any time the signal goes into a buffer it is completely reclocked and all jitter is eliminated again. Keeping that in mind, a DAC with a buffer and reclock immediately preceeding the conversion is a good thing for low jitter.

As far as wifi or ethernet based networks. Access is based on exponential back-off algorithms of different types (binary or truncated respectively). This means that a ethernet or wireless device will check to see if the line is being used and if it is back-off in increasing amounts of time until a ceiling is reached. So you definitely would want to have a buffer on the other side of the network if it is busy network, or just in case you have have a pesky program bothering your network with extra traffic you are unaware of, in order to correct for any jitter the algorthm introduces. Add to this the routing issues and/or handshaking requirements of the different layers in the TCP/IP stack and you definitely should have a buffer before that DAC on the other side IMO.

BTW, I recommend a RAID 1 setup plus one extra HD of the same size as the RAID drives to do a complete backup on as well. This way you are covered for all possible HD problems that can occur internally. 1) RAID 1 (and backup) covers for a HD failure, 2) HD backup covers for virus or critical system file failure issues taking your RAID setup down. This way you can restore your HD music server after foreseeable failures (of course, no one can restore your computer if it spontaneously combusts for some reason... Ha ha ha).

Jan Vigne
Jan Vigne's picture
Offline
Last seen: Never ago
Joined: Mar 18 2006 - 12:57pm
Re: jitter


Quote:
If the DAC has a memory buffer and reclocks the signal right before the conversion process, then it will eliminate any jitter accrued up until that point.

I would disagree with the idea any and all jitter will be eliminated by buffer stages. Read the linked articles and you'll find the many forms of jitter which can occur. While there are many steps in digital playback which can reclock the signal, they can only work with the incoming signal. If the data has been corrupted, you may have the signal reclocked and not repair the damage done by the effects of jiter. Remember, garbage in = garbage out.

jkalman
jkalman's picture
Offline
Last seen: Never ago
Joined: Sep 20 2005 - 7:04am
Re: jitter


Quote:
*** You are ignoring this user ***

I have to remind you again, that I can't read what you are saying.

My webpage contains all of my background information with this topic (computer organization and architecture and networking specialization) as well as my accolades in those endeavors. Unfortunately, I don't have any of the work I did at Columbia University listed on it though.

Jan Vigne
Jan Vigne's picture
Offline
Last seen: Never ago
Joined: Mar 18 2006 - 12:57pm
Re: jitter

I KNOW THAT!!!

Why don't you stop being a little baby and join the rest of the forum. At present you are playing the role of a portion of the male anatomy that seldom sees much daylight.

jkalman
jkalman's picture
Offline
Last seen: Never ago
Joined: Sep 20 2005 - 7:04am
Re: jitter


Quote:
I have been wondering about jitter lately too. What effect does transferring a CD to a hard-drive, storing in lossless compression, then the player software uncompressing it and sending the signal to an external DAC have on jitter. Is there jitter that would get captured in the original rip from CD? Or in some stage of the playback?

How about when we throw in an Airport Express or similar wireless transmission between the computer and the DAC?

mubba, if you want to see an audio company that employs the kind of ideas I am talking about, check out Meridian Audio. A prestigious company considered to be a "founding father" of digital high-end. They tend to be at the forefront of new digital audio technologies (another reason I bought the Meridian 861 for my own system, and use it as the DAC for my own computer setup!).

Their CD playback systems use a CD-ROM drive, because they are more thorough with error correction than regular CD drives. Their internal circuitry buffers the RAM right before Digital to Analog Conversion (any time you buffer the signal the bit waveform - squarewave - is completely refreshed, or "reconstituted," and reclocked by their system word clock, i.e. it is exactly as if it had never experienced any attenuation or other degradation factors such as pit discrepencies, etc, etc). Any system on the market today that is close to the lowest "jitter" numbers employs a similarly designed system (such as Classe as well).

The only thing I would recommend is that with whatever you build, you make sure the DAC has a RAM buffer close to it, an excellent wordclock (so it reclocks the signal accurately when it leaves the RAM), and a quality DAC. You also should check, as I mentioned in my first response, to make sure no bit errors are being introduced along the way (this is very easy to do using MD5 checksums and sending a file through your entire network to the DAC output, or input if it has no output). As long as all of these things measure correctly, and the DAC is of high quality, everything is good.

The nice thing about computer hardware is it is extremely redundant in error correction design. Computer software OTOH, is another issue completely. That is why it is important to do some file checks at different stages to make sure a program (such as your lossy compressor, or such, isn't going to introduce errors). You can check this by comparing files on the disc to the final products on your computer. Make a lossy file, turn it back into the regular file again and compare it to the original. Take a file off the CD and store it on your HD, then compare it to the original.

If anyone needs help setting up and validating the integrity of their setup, let me know.

  • X
    Enter your Stereophile.com username.
    Enter the password that accompanies your username.
    Loading