|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
67.176.62.39
In Reply to: Interesting EAC results posted by Presto on March 1, 2007 at 13:19:08:
Hi Presto,Why not use "Test and copy selected tracks"? That way, the CRC of a test read is compared with the CRC of the actual read. This avoids having to do manual bit-for-bit compare of two consecutively-ripped files.
Also, are you doing read offset correction? If not, and you compare rips done with two different drives, offset of one file relative to the other will need to be done before comparison.
I've done a number of different experiments with about four or five drives with secure mode and burst mode (but not fast mode). Results of burst mode vary considerably from one drive to another. One drive, the AOpen COM5232AAH Pro, passed level 5 (the highest) of the CD-check audio test disc at cdrinfo.com. Using burst mode with this drive, the only time I ever got mismatched CRCs with "test and copy" was on discs that caused the rip to slow waaaaay down in secure mode. Rip speeds were up around 24x average and higher. From these results, one might conclude that burst mode is almost always safe. However, I also tried another drive in burst mode with very different results. This was a LiteOn model. Using burst mode with this drive, I got at least one mismatched CRC on every disc I ripped.
My current strategy is to use two drives. First, find a drive that rips very fast in secure mode and use it for almost all rips. Second, find a drive that has the best error correction performance possible. If the disc cannot be ripped with the drive that's fast in secure mode, use the drive that has the good error correction. I've never found a drive that meets both these criteria. The ones that rip fast in secure mode are the ones that don't cache audio data when ripping. For an example of a currently-available drive that's very fast in secure mode, see this page. Scroll down for EAC secure mode benchmarks. I use an LG GSA-4167B for most of my rips. It will usually do an average rip speed of 24x across the disc, but its error correction is not so good. I use a Plextor to rip the ones that the LG won't do. I don't think this LG drive is available anymore. The ASUS drive I linked to above is the only secure mode speed demon I know of that's currently in production.
Follow Ups:
Has a really nice ripperInitially rips in burst mode, does an AccurateRip comparison. If a mismatch/error is detected, it then rerips in secure mode (and it can vary the read speed on different passes).
It can disable FUA on Plextor drives.
I also have a GSA-4167B, but I prefer using a Plextor PX-760A for rips these days.
I went with a PC for audio specifically because the best ripper around (3 years ago) was EAC. I'm intrigued by the Benchmark USB Dac but if I went that route I gather that Apple provides a better USB output so I'd like to find out if a good ripper (Apple compatable) is availble.
Gerry
.
...but I am about to give it a shot right now. I just got a Dixie Chicks CD whose last track is flakey, so I'll see how dBPowerAmp handles it.
PS - as of today I have ripped nearly 8000 audio files (over 600 CDs) into my 1.2TB NAS.
Well, that is a nice ripper. I like the UI.When ripping "Taking the Long Way" by Dixie Chicks on the LG drive, I had an AccurateRip mismatch on the last track as expected from my previous experience with EAC. The LG drive made some noises but the rip did continue without giving up. I tried to enable secure mode in dBPowerAmp, but it said I had to install the $18 "Power Pack" to get secure mode. I didn't do that.
Next, I reinstalled AccurateRip into EAC (which I usually don't use). Since this CD failed to rip the last track with the LG, I did it with the Plextor (PX-740). It took a very long time on the last three tracks, but it did eventually give me a clean bill of health with AccurateRip on all tracks.
So I went back to dBPowerAmp and tried it with the Plextor. I ended up getting AccurateRip errors on the last three tracks, the same ones that caused EAC to slow way down.
It would be interesting to see if secure mode in dBPowerAmp would allow the LG and/or the Plextor to get through the last tracks of this CD with a good AccurateRip result. But not having secure mode available in dBPowerAmp freeware version makes this an unfair comparison.
PS - as of today I have ripped nearly 8000 audio files (over 600 CDs) into my 1.2TB NAS.
Unfortunately, that's only about half my collection :-(
Yeah, I've got 935 CDs ripped - which is my entire collection. I sure wish I had the LG when I first started ripping. The drive I started with was very slow in secure mode, and I'm doing "test and copy", which almost doubles the time to rip a CD.
The machine I'm typing this on is also my music server. It's an Antec P180 box with 4 300 GB Seagate drives hooked up to a Promise RAID5 controller. There's also 2 IDE drives - one for the OS and apps (but no data) and a second to hold a backup image of the OS drive.
That should enable you to try out the "Secure" and "Ultra Secure" modes (for up to 30 days I think) to see if they help. After 30 days it reverts to a freeware version.I was using the beta version, and liked it so much I ended up paying for a "Reference" version.
PowerPack doesn't give you "Ultra Secure" - only the rights to create MP3.
The software was originally intended to be "free" - but the MP3 licensing vultures got onto the guy. I believe nearly all the money for the Power Pack goes towards paying for the MP3 licence. Same deal with Reference - a large part of it goes towards paying for the AMG data.
I had a chance to try out dBpoweramp with my Dixie Chicks CD that has the defective last track. As mentioned, I have two drives - the LG-4167B and Plextor PX-740 (which I think is a rebadged BenQ). In EAC secure mode, I could rip the last track successfully with the Plextor but not the LG. I tried this in dBpoweramp secure mode. The Plextor did fine (same as EAC). With the LG, the results were variable. On one attempt, it got into this very long procedure, and when it was done, it ended up with the wrong CRC according to AccurateRip. Then I tried it a second time without removing the disc from the drive or anything else. This time, it ripped the track very fast, and gave the correct CRC. It's worth mentioning that dBpoweramp appears to use AccurateRip as an integral part of its algorithm to determine if it ripped the track correctly. This gives it an advantage over EAC in the case where the CD is in the AccurateRip database. EAC appears to only report the AccurateRip results after the fact.So I tried to modify EAC's settings to get it to do better with the LG drive. I had "skip track extraction after duration longer than 4 times real time" enabled, which was causing it to give up. I disabled this option and tried re-ripping the problem CD using "test and copy". I ended up with the correct CRC, but the test and copy CRCs had a mismatch (the final copy had the correct CRC). I tried it again and got the wrong CRC in the copy.
Based on these tests, I'd conclude that dBpoweramp is the equal of EAC in terms of data integrity. The only concern I have with dBpoweramp is what might happen with a CD that has problems and whose information is not in the AccurateRip database. If you sit and watch the extraction, you'll see that it will slow way down on bad tracks. But if you set it to rip and then walk away, there doesn't seem to be any indication after the fact that a problem occurred. Maybe Christine has found an option that could help with this. With EAC I knew there was a problem because it gave up on the rip. Even when I disabled this option, I still knew there was a problem because of the mismatched CRCs in "test and copy".
So dBpoweramp looks very promising at this stage. But I'm not ready to give up on EAC yet.
I have quite a few CDs not in AccurateRip.What dbpoweramp seems to to do is it will automatically go into Ultra Secure mode and rip the CD at least twice (although this is configurable) and reports okay only if both rips are identical.
It gives an error if the rips are not identical even if no C2 errors are reported (and allow you the option to rerip).
So it seems reasonably paranoid (in a good way).
I've tried ripping a badly scratched CD that I borrowed from a library (no, I did not keep this rip). It had to rerip some of the tracks up to the maximum of 6 times, and marked several tracks as bad. This disc wasn't in AccurateRip. Interestingly, after I ejected the disc and retried again, it was able to rip one of the bad tracks successfully. But it was never able to rip the whole disc successfully.
And speaking of MP3 licensing vultures, does anybody's name ring a bell in this article or this one ? :-)
I think it would be hard for Microsoft to claim that he had no involvement with WMP given he actually posted as a Microsoft rep in AVS Forum and stated his connection to the audio processing modes in the Vista audio stack. :-(Thanks for the links to the article - Ted and I were talking about jj only a few days ago in the Digital forum!
This post is made possible by the generous support of people like you and our sponsors: