 
|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
|
|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
78.16.244.230
In Reply to: RE: Jitter & Bit-perfect Data posted by Tony Lauck on August 14, 2009 at 10:32:36
Tony,
When you say " If a signal is closer to 0 than it is to 1 as the circuitry processes it it moves it closer to the ideal value, similarly if it is closer to 1" I take it to mean that the file being read off CD & copied to HDD is going through this process & so when it arrives on HDD all 1 bits will be at the correct signal level (or nearly) irrespective of the whether they arose from a CD reader or a Blue Ray disk reader?
One question - how wide is the allowed tolerance for a signal that represents the ideal? Is this published anywhere I could research? And what happens on playback if this signal slips outside the tolerance - the system corrects it - this is where I see added jitter can come from (ie anything that is extra work during playback can affect timing)
Follow Ups:
"One question - how wide is the allowed tolerance for a signal that represents the ideal?"
I was talking about the operation of all digital logic circuitry in computer systems, such as the AND gates, OR gates or FLIP-FLOPs. The specifics will depend on the particular equipment. Even if the logic gates restore a signal completely, by the time the signal propagates down a PC board trace or a wire to the next gate it will be degraded by rise-time problems, ringing, reflections and cross-talk from other systems. The actual quality of a signal when it reaches its destination will depend on the overall quality of design and implementation. Of course most computers are not under designed. There will be enough margin to keep the 0's and 1's separated nearly all the time. But not always. While many "blue screen of death" crashes in computer systems are caused by software bugs, many others are caused by hardware glitches due to poor signal integrity. As in all other products, you get what you pay for. Standards are higher for equipment used in avionics applications, etc... (hopefully)
However, by the time a signal gets from an optical drive to a HDD it is unlikely there will be a significant difference in the levels caused by variations in the optical drive, assuming the computer is working properly. The usual problem reading CDs is that they are sometimes damaged beyond the ability of the error control hardware (or software in the rip program) to make sense of the data. Normally there are thousands of bad 0's and 1's on an optical disk under good conditions, but the first and second level error correcting codes figure this out, just like you can figure out misspelked werds...
If you read the same CD with a different drive you are likely to get different results (i.e. different 0's and 1's). This is because of an arcane problem called an "off-set error". This is one of the problems caused by the general poor design (by modern standards) of the Red Book CD format. When there is an offset error all the samples in a track will get shifted, and there will be a few zero samples added or removed from the ends. Theoretically these two different files should sound the same, since no one really cares when they pressed the "Play" button, but some people have reported that these different files sound different. This would account for rips from a CD reader or Blue Ray disk reader being different (e.g. having a different MD5 checksum). If you have two files from different rips that sound different, you can send them to me and I will investigate exactly how they differ.
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
Tony said
"misspelked werds." I understand but it takes me just a tiny bit longer to read this than the correctly spelled words & I'm sure different ares of the brain are firing. So even though I can correctly interpret it if I needed to do so in a certain limited time I might find it difficult to hit that time barrier (You might say I'm a bit slow :))
The offset issue is a strange one - what explains the sonic difference in these files? Those files would not have the same hash totals, would they?
I would be happy to upload these files to a site & let you download them. Just have to check some issues first.
Post a Followup:
| FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
||||||||||||||
This post is made possible by the generous support of people like you and our sponsors:
Top of Page ]
[ Contact Us ]
[ Support/Wish List ]
[ Copyright © 2013, Audio Asylum, all rights reserved ]
[ Take5 ]
