Tony discovered he needed to config his offset, and got it right: >-I had some WAV files on my hard drive >-I wrote them to CD >-I used EAC to extract the files >-I compared the first version of the WAV to > the copy extracted by EAC using EAC?s > ?compare WAVs? feature. I got this: >Track 1 of the CD: >Error type: 135 repeated samples (on the 2nd gen file) >Other tracks of the CD: >Error type: 135 missing samples (on the 1st gen file) >I poked around the EAC documentation, which I didn?t think was particularly >easy to access. Only half-grasping what I was doing, I changed the ?Read >sample offset correction value? to 135. Then I re-extracted the files from >CD. This time the ?compare WAV? reported no errors. > >So my questions: > >How important is this?
Debatable. By etree/audiophile standards, it is clearly "required." If you don't archive your shns as data (which you should), even a 100% quality re-EAC'ing of the wavs from an etree shn won't produce identical md5s because of the shuffling of the data 135 samples ahead/behind, whereas with proper offset correction you theoretically can produce shns identical to the originals. ***But you oughta just archive the original shns as data so you don't have to mess around with EAC so often.*** Beyond that issue of replicating the original shn, one's own uncorrected offsets almost certainly will not have any audible consequences in the audio generation you create, since we're dealing with such minute increments of time here. However, there is a concern in trading circles that over generations of uncorrected extraction and burning, bits of signal could be lost at the beginning or end of tracks due to the cumulation of offset-related movement of the data over many gens. The philosophical debate over the probability of this actually occurring (and, if it does occur, the implications for Western Civilization as we know it) could dwarf Tom & Jeff's discourse, and I hope I haven't sparked it just now. I'd invoke H**tler to cut it off before it begins, but PJP's link says deliberate invocation of Godwin's Law bars its implementation. So back to the short answer: How important? Around here, generally considered very important. YMMV. >If it?s important, do most people who trade live music know about this? It >feels like I had to dig a bit to figure it out. Should it be mentioned in >the Etree FAQ? It should, be in the FAQ and it is, tho' perhaps not as clearly as it should be. The key resource on offset correction is Dick's Unofficial EAC Page, http://pages.cthome.net/homepage/eac/setup.htm ,it's listed on etree's EAC page. Etree's CDR FAQ mentions it too, tho' (hey webguys!) not with a hot link, which would help: "For additional information and EAC troubleshouting, check out Dick's Unofficial EAC Page." BUT TONY, YOU SHOULDN'T BE NEEDING TO EXTRACT DOWNLOADED SHOWS. Burn the shns as data to archive discs before you expand them to wav on your hard drive and blow away the shns. Put those away safely and you'll never have to EAC downloaded shows. One of the reasons EAC isn't covered in great detail in the FAQs is that etree is about shn, the trading and downloading takes place in shn, and most of us are only very infrequently having to use EAC to extract the occasional audio show that we couldn't track down in shn. At 01:21 PM 3/25/2002, Adam wrote: >Hello. I have been trying to configure the offsets of >EAC for the last 3 hours to no avail! I have read the >webpages listed at etree.org and Im just stuck. I >followed all of the instructions, and when I extract I >always get either 99.8 or 99.9 track quality. I want >to get 100% . Your offsets may be fine, and almost certainly are not the cause of those track quality readings. If you've run the "compare' wav drill laid out at http://pages.cthome.net/homepage/eac/setup.htm , and it comes up no missing/added samples, your offset is set ok. Those barely sub-100% quality readings are a yellow flag warning that EAC had to back up and re-read some sectors to satisfy itself it was getting the goods right. Andre, creator of EAC, sez in his (relatively new) official FAQ: >Q: What does the Track Quality really mean? A few times I get 99.7% or >97.5%. But there are no suspicious position reported. A: When you get >99.7% and so on, that means that a bad sector was found, but the secure >mode has corrected it - from 16 times of grabbing the sector, there were 8 >or more identical results. So it only indicates read problems. It is the >ratio between the number of minimum reads needed to perform the extraction >and the number of reads that were actually performed. 100% will only occur >when the CD was extracted without any rereads on errors. ONLY when there >are suspicious positions reported, there are really uncorrectable read >errors in the resulting audio file. Far as I know there isn't anything you can tweak in your EAC settings to bring that number up to 100%. It's more than likely a function of CD reader compatibility. Some CD drives read CDR more accurately and quickly than others, and this gets reflected in their EAC performance. My old Teac drive got marginal performance under EAC, and for about $40 I replaced it with an Afree 56x, which gets good reviews here and elsewhere on EAC performance. Goes much faster under EAC now and those 99.9%'ers are much rarer. But apparently 99.x% isn't a big problem so long as EAC does not report suspicious positions. All of this can be largely avoided by archiving your shns to disc as data upon receipt, so you never have to re-extract those shows. Then you only have to EAC shows you unavoidably acquired in audio with no shns. wilbur _______________________________________________ etree.org etree mailing list <[EMAIL PROTECTED]> http://mail.etree.org/mailman/listinfo/etree Need help? Ask <[EMAIL PROTECTED]>
