Aquarium

I have a good ear for music and play an instrument so I think I can gauge when something is right rather than obviously incorrect!(and Lo-Fi is a myth).I have issues with anyone that does not use proper AC line conditioning for their digital components and computers,because bluntly it makes a hell of a difference.I know how poor CD's sound compared to one burnt in my system and I am reporting here I find further improvement with the new offset.

greynol

JeanLuc is right, offsets have nothing to do with sound quality (nor does the use of null samples in CRC calculations!).

I'm not going to discuss this with you any longer other than to say that you shouldn't be surprised when people don't take you seriously.

Aquarium

I heard the differences I heard is all I can say to this.I do not claim to be any sort of technical expert and would concede that to you.I tested "That's All Right Mama" from Elvis 2 just to make sure again and the improvements I noted before were all there.The new offset was better than EAC's.There is no reason why it shouldn't be so as I was getting better CD's using just line conditioning before I discovered EAC.Everything is coming together better and I hear it.In a blind test it really would not be an issue because a veil is lifted quite noticeably and more detail is present.Needless to say I will use the new offset in future.

Hunter

If you are so certain that you could head the difference in an ABX blind test, could you please try and share your results? I would be very interested to see the outcome.

Aquarium

There is a major issue here and that is one I can't do much about unfortunately.The fact that I use a dedicated AC line filter and conditioner that gives me the ability to run all my digital and computer components,AT THE ABSOLUTELY CORRECT VOLTAGE without issues that effect hardware performance.Now this means that I can't really give you a sample that would be meaningful unless your system was at least as sensitive as mine.Sound issues I know,occur just because of the lack of regulated power and it is not pretty.Now I would only advise that if you care about sound,invest in line conditioning at the audiophile level anyway-you have not heard what EAC can do until you have.Yes my CD's are the best in the world.

Andre Wiethoff

Please calm down, and please stop being off-topic!
All of you were satisfied when nobody knew what the correct offset really was, it always was an approximation (some of you knew, the other were not really interested...)
So what did change? Basically nothing. If you don´t want to use AccurateRip, just take the "proposed" offset and enter that value -30 into the drives option field. Voila, everything is fine. If you prefer to have AccurateRip or just don´t care (as I do for 30 samples!!!), then just keep the old offset. These 30 samples will make a difference only at the very beginning and the very end of a CD!
Hope that clarifies the issue!

cu, Andre

TheLegioneer

I'm not an expert by any means, but I was under the impression that the offset affected the whole disc. So, as an experiment, say if the offset was shifted 600 samples sooner, then Track 1 would start 600 samples too early, 600 samples from Track 2 would be at the end of Track 1, 600 samples from Track 3 would be at the end of Track 2, and so on, until the last track of the disc would be missing it's last 600 samples. Is that the way it works, or no?

greynol

Yes, that's the way it works.

Now, does this new and improved offset hold true for all CDs, regardless of the manufacturing process?

TheLegioneer

AFIAK, no it doesn't. Because there is no standard for how far or in which direction manufactures skew the "start" of their disc, it will only offset the drive to where the "start" should be. Though in practice, Andre's offset may be more common. If he found that 6/10 discs all shared the offset he derived, it could just be that the deviation on those CDs is a common skew many manufacturers use, even if it's not at "absolute zero".

calestyo

Ok does that mean now that the real read offset is +30 from the value PlexTools etc. reports?
And thus the read offset correction value (that what EAC uses) must be corrected by -30.
e.g. if my PX760 has:
read offset -30
write offset -30
read offset correction +30

Then I must use
read offset 0
write offset 0
read offset correction 0
combined offset: 0

Right? Or doesn't the whole thing affect the write offsets?


But in general. I have understood the offset issue with the drives pretty well, but do CDs itself have an offset too?
If so,... what does that mean exactly? That the drive does not find the beginning of data on the CD where it should be? And that this even varies from studio/manufracturer to studio/manufracturer? If so the whole offset issue may should be useless at all,.. because no one can tell what the combination of the drives offset + CDs offset is.



But I've always thought that there is a standard that defines where to find what on a CD,.. and so I've always thought that it wouldn't matter if there is an offset on the CD itself?!


And what should one do now to get the most exact/best results? Use EAC's values or the possibly real offset values? Becaus if CDs have their own, variing offset,.. the "real" offset found by IpseDixit may be still not as good as EAC... becaus EAC's could correct the offsets of CDs better.


Wouldn't it be possible to make a test where a CDDA starts? If one reads the whole CD in bytes as they come,... there should be some fixed points (lead in/out for example) where on know how to find the beginning of audio or so?


Is it possible to test for the offset myself and how would such a test look like?
I mean if I use EAC with a write offset of 0 it should burn the data just like the drive takes it.
Now if I burn some "raw" audio where the first sample is 0x1... and everything else 0x0
and afterwards rip the whole thing with different read offsets I should be able to get the correct offset as soon as the ripped wav file contains th 0x1 as first sample.
Would that be possible or do I miss some details?


Now for my PX760A...
When it has a read offset of -30 (correction +30) .. as EAC says,.. and it is able to overread into lead in/out that would mean it begins 30 samples to early with reading (at the first thread,.. thus I get some unwanted data from lead in) and stops 30 samples to early at the last track, thus I'd miss data that I want.
But thanks to overread lead in/out I can get those data when I use read offset correction.
But now,.. if it's true what IpseDixit claims... it would mean that I loose the first 30 samples and get 30 unwanted from the lead out.
Right?

So would you suggest to use EACs values or IpseDixit's?
Or is the whole thing unrealistic because of the offset of the CDs itself?

Thanks in advance,
Chris.

calestyo

That would be interesting,... I've written a mail to Plextor asking them about the issue.

Chris.

calestyo

If the read offset correction value by EAC would have to be corrected by -30 (as IpseDixit claims)....
Wouldn't it be easyly possible check that?
If the real value would be -30 away,.. that would mean I loose 30 samples in the beginning and get 30 null samples in the end.
No just make a CD where the first track begins immediately with non-null samples and the last one that ends immediately with non null samples.
No one could check if those samples appear after ripping and could tell if -30 is true.

Of course that wouldn't solve the problem that CDs may have their own offset.

greynol

If it's the real value you lose nothing. With overreading, non-null samples were never replaced with null samples to begin with and this will still hold true. The difference lies in the fact that these Plextor drives cannot write to the lead-out, which means that with a write samples offset of -30, the last 30 samples will always be null. I covered this already in post #12

EDIT:But to clarify further, according to the new and improved offset correction, rips with the old offset correction will be missing 30 samples from the beginning and will instead have an extra 30 samples at the end. It was commonly believed that these 30 samples weren't really comming from the lead-out, just that the drive thought they were. Now it turns out that they actually were coming from the lead-out.

calestyo

I wrote a mail to Plextor,.. this is what came out.... any ideas?

Chris.

calestyo

What? I meant that if real correction value would be -30 off EACs... and in this case you would loose the first 30 samples and get 30 unwanted at the end.


That was what I meant.

Well I'm still not sure if it's better to use the "new" offset values (which would be 0 for my px760A as far as I understood) or the old ones,.. because as CDs have their own offset,.. the "old" values might still be better or at least not worse than the new ones.

Please have a look at my replies from Plextor, too.

Anyway,... that doesn't answer my question for a simple efficient check for the whole issue.