Skip to main content

Topic: NEC Read/Write Offset (Read 4918 times) previous topic - next topic

0 Members and 1 Guest are viewing this topic.
  • Rain
  • [*][*][*]
NEC Read/Write Offset
Hi there,

I've been a regular reader of Hydrogenaudio before I decided to register.

Firstly, I currently have a NEC DVD-RW ND-2510A (which is unable to overread into Lead-in and Lead-out). However, I am unable to determine its read or write offset - Since many of the CDs I own are not included in EAC's Reference CD (this also applies to AccurateRip's key discs).

I have checked the AccurateRip database and found:
  • Read Offset is +48, as reported by 95 users.
  • I have searched through many websites and found out that NEC drives which have a read offset of +48 also usually has a write offset of +18 - I have yet to find the write offset of my NEC model (Is it safe to assume this? Or am I wrong?).
Secondly, if I rip an original CD to my PC as an image file (assuming it was 100% perfect), and mount it on a virtual drive (DAEMON Tools in this case), and use EAC to rip the tracks, will it be necessary to specify read/write offsets for virtual drives? And will the audio be identical to the original (i.e. not missing, for example 30 samples at the end)?

EDIT: D'oh, wrong forum
  • Last Edit: 27 January, 2006, 06:03:40 PM by Rain

  • edwardar
  • [*][*][*]
NEC Read/Write Offset
Reply #1
I can confirm that my drive, an NEC 2500A (flashed with 2510A firmware) has a read offset of +48 and a write offset of +18.

I confirmed this by ripping a CD with EAC, then burning it back to CD and re-ripping the copy.  The file sizes and CRCs of the 2 files were identical.

You could try the same with DAEMON tools,

Ed

  • Rain
  • [*][*][*]
NEC Read/Write Offset
Reply #2
Thank you for the reply. I tried it out and yes, the read and write offsets are correct.

I have another question, I have another drive - I used EAC to detect whether it overread to lead-in and lead-out and it turns out the drive can read "Only Lead-in" so should I check EAC's option that it can overread into lead-in and lead-out just in case?

NEC Read/Write Offset
Reply #3
Quote
I can confirm that my drive, an NEC 2500A (flashed with 2510A firmware) has a read offset of +48 and a write offset of +18.
[a href="index.php?act=findpost&pid=360230"][{POST_SNAPBACK}][/a]


Is +48 the read offset or the read offset correction? Also, is the write offset really +18? Most drives I saw have negative write offsets.

  • edwardar
  • [*][*][*]
NEC Read/Write Offset
Reply #4
Yes, +48 is the "read sample offset correction" and +18 is the "write samples offset".

Ed

  • Martin H
  • [*][*][*][*][*]
NEC Read/Write Offset
Reply #5
Quote
I used EAC to detect whether it overread to lead-in and lead-out and it turns out the drive can read "Only Lead-in" so should I check EAC's option that it can overread into lead-in and lead-out just in case?

Overreading should only be enabled if the drives read offset correction is possitive and it can overread into lead-out, or if the drives read offset correction is negative and it can overread into the pre-gap of track 1(called lead-in by EAC).
  • Last Edit: 30 January, 2006, 10:59:34 PM by Martin H

  • Rain
  • [*][*][*]
NEC Read/Write Offset
Reply #6
So, should I be worried? I'm kind of paranoid. Is the audio that I ripped using EAC perfect (i.e. there are no samples missing in the beginning or the end)? I'm using a cuesheet + lossless audio file. I've read the recommended settings on this site.

My NEC's drive has a read offset correction of +48 (48/44100 = 0.00117 sec) and write offset correction of  +18 which means I'll lose the last 18 samples (18/44100 = 0.000408 sec) since it can't overread into lead-out. Wouldn't the first and last few samples be silence anyway and hence be filled with 0s? Which in turn means nothing is lost (i.e. my backup is perfect) since most CDs first and last tracks start with silence?
  • Last Edit: 31 January, 2006, 12:54:23 PM by Rain

NEC Read/Write Offset
Reply #7
Quote
My NEC's drive has a read offset correction of +48 (48/44100 = 0.00117 sec) and write offset correction of  +18 which means I'll lose the last 18 samples (18/44100 = 0.000408 sec) since it can't overread into lead-out. Wouldn't the first and last few samples be silence anyway and hence be filled with 0s? Which in turn means nothing is lost (i.e. my backup is perfect) since most CDs first and last tracks start with silence?
[a href="index.php?act=findpost&pid=361020"][{POST_SNAPBACK}][/a]


Exactly - but some people are really anal about it.

  • Martin H
  • [*][*][*][*][*]
NEC Read/Write Offset
Reply #8
Quote
So, should I be worried? I'm kind of paranoid. Is the audio that I ripped using EAC perfect (i.e. there are no samples missing in the beginning or the end)?

To make you understand this subject better, i have copy/pasted something i wrote on the EAC forum recently on the same subject :

[span style='font-size:8pt;line-height:100%']"The offset correction feature is about having a reference to adjust the drives after, which makes it possible to compare CRCs of rips done with different drives, to use AccurateRip and for avoiding generation-loss when making several generations of copies of copies... Basically it makes the ranges ripped from the CDs the same, allthough different drives with different offsets is used... But what is important to understand, is that the rips aren't any more bit-perfect or exact. This is because CDs have varying starting/ending possitions and hence, adjusting a drive to read from a specific possition of a CD won't make it any more exact when the starting/ending possitions of the different CDs is highly varying... Andre once looked at some old CDs where it was possible to determine their exact starting/ending possitions. He looked at many different CDs but found 6 CDs that had the same starting/ending possition. He then made that possition the reference, and hence, drives extracting from this precise place of the CDs would have an offset correction value of exactly 0, and if a drive reads 10 samples before this possition, then the offset correction value of that drive would be +10. Pio2001 has also once looked at some CDs to determine the offset of one of his drives... He looked at different CDs, and found two that he could use for the test and which had the same starting/ending possitions and used them to calculate his drives offset. The value he found was 12 samples away from Andre's reference, meaning that those CDs Pio2001 used for determining his drives offset value, had a different starting/ending possition than the 6 CDs that Andre used for his reference..."[/span]

Links -

How to find a reference offset (Pio2001's offset test described here) :
http://pageperso.aol.fr/Lyonpio2001/offset.htm

Arguments against offset correction (Listed on HA.org's FAQ) :
http://www.digital-inn.de/showthread.php?threadid=4193