On Mon, Feb 11, 2013 at 3:54 PM, Richard L. Hess
<[log in to unmask]>wrote:
> As I understand it, jitter at the A-D conversion can be baked in as
> amplitude nonlinearities in the sampling. If jitter occurs, then the
> sampling window is not looking at the analog waveform evenly and no longer
> accurately following the waveform.
Yes, but it's worse than that. There is jitter imbedded in each
byte/packet that the normal counters don't register, which is why this
hasn't been addressed until recently.
> Jitter in digital transmission should be inaudible UNTIL the bit-error
> rate increases beyond the error correction capabilities of the channel. At
> that point there are either glitches or error concealment.
While true, banging up against the ECC limits should not be a problem in
the future; the whole ECC structure must be rethought.
> Jitter in D-A conversion also introduces amplitude non-linearities unless
> the samples are buffered and de-jittered.
It now appears (although I cannot swear to it) that buffering hardly helps.
> So, Carl's "when" is accurate because if the "when" is not evenly spaced
> across a wave, it will incorrectly digitize (i.e. write down the wrong
> number) or give the right number at the wrong time (on output) which messes
> up the reconstruction filter.
> I propose that jitter is a major issue with unrecoverable digital audio
> tapes and why baking works for both 3348 and DAT tapes (sometimes) as it
> removes the stick-slip which is an annoying form of jitter.
"An annoying form of jitter." Greatest understatement I've ever read on
Jitter ends up closing the eye pattern in the "bit slicer" and you cannot
> accurately read the digital data. I have heard of people adjusting early
> tape machines' bit slicers to try and play problem tapes. I keep saying
> "bake" or "cold play" and neither has happened to the best of my knowledge
> with a couple of tapes that have been floating around unplayable.
> Some of Charles Poynton's books on digital video have good explanations of
> sampling theory.
Thanks for the lead.
> On 2013-02-11 10:43 AM, Carl Pultz wrote:
>> The way I've thought about this matter is that the data representing the
>> audio waveform is the WHAT. The other key aspect, the WHEN, is assumed.
>> is, WHEN is defined by the sample-rate, but is not defined within the data
>> in such a way that it can control the subsequent hardware clocks. Timebase
>> errors at a-d will have an effect because there is no way for d-a to know
>> about those defects. It assumes perfection. For copying or DSP WHAT is all
>> you need to know. For conversion, WHEN becomes critically important and is
>> subject to various approximations, even given the much improved hardware
>> A reasonable interpretation?
>> -----Original Message-----
>> From: Association for Recorded Sound Discussion List
>> [mailto:[log in to unmask]**GOV <[log in to unmask]>] On
>> Behalf Of Tom Fine
>> Sent: Monday, February 11, 2013 9:19 AM
>> To: [log in to unmask]
>> Subject: [ARSCLIST] Jitter (was Re: [ARSCLIST] Audibility of 44/16 ?)
>> Can jitter be introduced on the A-D stage? As I understood Mike Gray's
>> posting, he was saying jitter can be induced from the get-go, in the A-D
>> process. Konrad, do you know that to be untrue?
>> Also, I've been told by one of Sony's senior EE guys that it can be baked
>> into a glass master. As I understand it, jitter can be induced any time
>> bits are clock-aligned for whatever reason. I'm not sure why that occurs
>> making a glass master, but a lot of research was done on this back in the
>> 80s and 90s, at least that's my understanding from what the Sony guy told
>> So, I think (but may have learned this wrong, I'm not an EE) that bits is
>> bits only when the bits are kept absolutely intact and the
>> timing-transmission is rock solid.
>> -- Tom Fine
> Richard L. Hess email: [log in to unmask]
> Aurora, Ontario, Canada 647 479 2800
> Quality tape transfers -- even from hard-to-play tapes.