Help - Search - Members - Calendar
Full Version: Does incorrect read offset affect sound quality ?
Hydrogenaudio Forums > CD-R and Audio Hardware > CD Hardware/Software
goodsound
I believe my CD-RW (LiteOn LTR-48246S, Firmware ver SS04) has a read offset of +6. All I know about read offset is that its the under/overshoot that the drive does from the actual starting point of the requested sector.

So everytime a sector is requested it will miss that much part of the sector ??
Wouldn't that affect the sound quality in a big way when playing a CD off the drive ?

Also, if I were to rip a track to a wav file using secure mode in EAC, but without specifying the correct read offset, then wouldn't the resulting wav file be incorrect anyway ? i.e. some initial part of ALL the sectors in that track would be missing ?! no matter how many times eac reads that sector ?

Can someone please help clear this up for me ? unsure.gif
Thanks!
TBO
No. You will only miss a miniscule, unnoticable duration of audio. 6 samples is equivalent to 1/7350 of a second.

Wave files don't have sectors. For all intents and purposes, read offsets will not cause any audible or dangerous (for the non-perfectionist user) effects.
Jan S.
QUOTE(goodsound @ Feb 25 2006, 06:21 PM)
So everytime a sector is requested it will miss that much part of the sector ??
Wouldn't that affect the sound quality in a big way when playing a CD off the drive ?

Also, if I were to rip a track to a wav file using secure mode in EAC, but without specifying the correct read offset, then wouldn't the resulting wav file be incorrect anyway ? i.e. some initial part of ALL the sectors in that track would be missing ?! no matter how many times eac reads that sector ?
*


The offset is only relevant at the very beginning/end, so only stuff at the beginning/end of the cd/track can be lost.
Duble0Syx
As already said, read offset in no way affect the sound quality. It's only purpose is to garauntee a "perfect rip" although the rest of EAC's setting need to be right also. Basically using the correct read offset means you can rip the same disc on any drive and get bit-identical results.
Cosmo
"read offset of +6"

6 samples / 44,100 samples per second = 0.000136 seconds
goodsound
ok thanks. just a little unclear about whether that 0.000136 or 1/7350 seconds of loss is only for the first sector of the whole cd ? or the first sector of each track ? or every sector ?
Cosmo
QUOTE(goodsound)
ok thanks. just a little unclear about whether that 0.000136 or 1/7350 seconds of loss is only for the first sector of the whole cd ? or the first sector of each track ? or every sector ?

Like Jan S. said, only at the beginning or end of the CD/Track (depending upon whether your offset is negative or positive).

In your case, I believe only the last 6 samples of the last track will be lost. Tracks prior to the last one will be shifted by 6 samples, but nothing is lost. (unless you consider things on an individual track basis...)

(edit: IIRC, +6 is your offset correction value, because your Lite-On is reading 6 samples too early...)

http://users.pandora.be/satcp/eacoffsets00.htm
Fandango
Doing read-/write-offset corrections while copying Audio-CDs matters most likely only for bootleggers anyway: when a CD gets copied many times by people who trade live recordings an incorrect offset can indeed cut off audio information after some generations.

But for normal use: it's fine to know your drives offsets and it is recommended, but it's not a must have. Having set "Secure with NO C2, accurate stream, disable cache" is more important for a good quality rip.
Duble0Syx
QUOTE(Cosmo @ Feb 25 2006, 10:38 AM)
QUOTE(goodsound)
ok thanks. just a little unclear about whether that 0.000136 or 1/7350 seconds of loss is only for the first sector of the whole cd ? or the first sector of each track ? or every sector ?

Like Jan S. said, only at the beginning or end of the CD/Track (depending upon whether your offset is negative or positive).

In your case, I believe only the last 6 samples of the last track will be lost. Tracks prior to the last one will be shifted by 6 samples, but nothing is lost. (unless you consider things on an individual track basis...)

(edit: IIRC, +6 is your offset correction value, because your Lite-On is reading 6 samples too early...)

http://users.pandora.be/satcp/eacoffsets00.htm
*


Obviously with no offset correction you'll still be missing samples. So why not use it? With some drives, like most Plextor's, you can overread and get all of the extra samples. My Lite-On drive though also has a read offset of +6 and 99% of the time the rips will match what my plextor does. In the small cases where it doesn't it's only the last track where it can't overread and the plextor can. Fact is, using read offset is easy, so why not? Especially when you use accuraterip. Nice thing is, most of the time the missing samples at the end of the disc in your case will be digital silence, there they won't be missed. EAC still includes that 0.000136 at the end, and can optionally fill it up with silence or do whatever it does when that option is off. smile.gif And some might think read offset is used mostly by bootlegers, but any bootleg rips I've come across that even have EAC log files didn't have offsets used nor any of the other correct setting for making proper rips. It's a shame too.
Eli
you can always use AccurateRip with EAC to get a correct read offset and more importantly know if your rip is accurate or not.
This is a "lo-fi" version of our main content. To view the full version with more information, formatting and images, please click here.
Invision Power Board © 2001-2008 Invision Power Services, Inc.