Right rip vs wrong rip

[Parisian]

Naim audio theory about ripping CDs

 

Continuing my quest for getting the best out of my newly installed ethernet home network, I found last week an interesting and surprising "white paper" from the naim audio company:

 

http://www.naimaudio.com/userfiles/modules/attachment/naim_cd_ripping_engine.pdf

 

It says that the current extraction softwares do not address some critical difficulties when ripping a CD. I advise you to read the paper as I didn't fully understand all the explanations. I can shortly summarise the different issues:

 

- The burst mode for reading a CD on computer is not adapted to the accuracy needed for reading an audio CD

- More and more CDs are copy protected and current softwares do not handle the problem. The protection may have different implementations : corrupt table of content, corrupt data layer and others...

- offset correction

- capturing lead in and lead out times

- mechanism requirements (??)

 

What is the opinion of the forum participants ? Has anyone made a listening comparison of a CD ripped with a naim (HDX or Unitiserve)and the same ripped with a good software (EAC or dbpoweramp) ? Was there any audible difference ?

 

Gilles

 

[Audio_ELF]

@Gilles

 

Did you read the conclusion? Naim are basically saying that their ripper use "best practice" methodology - the same as EAC and dbPowerAmp use.

 

Naim don't say you can't get good rips using a PC; just that getting a good rip is not as simple as it first appears (it's not copying in the same way as transferring your photos from a CD to your HDD.

 

At the end if the day there's not (to me) anything surprising in that White Paper.

 

Eloise

 

[sandyk]

In the April 2009 issue of HI FI Choice, there was an article called "Ripping Yarns" by Malcolm Steward.The attached is a small part of that article.

"To snap myself out of this near delusory state into which I had fallen.

I ripped one of the CDs I had been using to produce my test rips onto my Naim HDX,

a machine that was designed and purpose-built by audiophiles to rip CDs.

I compared it's output to the computer generated rips.

To say that the difference was night and day would be an understatement of some magnitude."

 

The Naim HDX is also using a CD-ROM and HDD.

 

[Parisian]

@eloise

To me the Naim's conclusion is rather ambiguous. I understood (sorry I am not a native English speaker) that good ripping is not on the reach of average PC (or Mac) users (that I am) and that it needs some good knowledge on software and hardware combination ?

 

@sandyk

I've already seen that comment. It made me investigate a bit further on this issue. I think the best would be to get some testimonies on listening tests.

 

Gilles

 

[Erwin S]

As Eloise pointed out, we, the readers of Computer Audiophile dot com knew this already.

 

My XLD ripper is set to XLDsecureRipper mode. With all the bulk-bought old battered CD's I feed to my Mac mini, it can take ages for some CD's. Most come out with all tracks "acurately ripped" in the end.

 

If you use the internal DVD drive to do the reading, don't worry about the offset correction, either XLD or iTunes will detect and auto-set the correct offset (48 in my mini). However, if you use an external disk drive, such as the excellent LaCie I used before, this will be a different value which you need to look up.

 

I used the LaCie in combination with the internal disk reader of my iMac, to make good progress, as I gained time with loading the CD's. In hindsight, this was totally wrong as the offset value was different for the respective drives... We all learn, don't we?

 

That's what CA is for!

 

(thank you, Chris, my CA membership turns two next week, time to order the birthday cake and have a party..)

 

[Jud]

People who're familiar with my other posts at CA may think demons have temporarily taken over my body when they read this, but isn't it correct that at least in this instance, "bits is bits"? That is, shouldn't one simply be able to compare the resulting files from two rips (very easy to do - matter of a couple of terminal commands), and if they are bit-for-bit identical, we're done?

 

[Parisian]

I am very respectful of CA and I have looked at the archives before posting this new topic, thinking it had already been discussed.

 

It appears that there is no mention of any listening test between a current extraction software and a rip by naim. This is a question since Naim audio claims to have a superior method.

 

Why do I insist on listening tests : this is the final relevant test.

Referring to technology can truely help sometimes but can also lead to some serious mistakes. Remind of so to say obvious truth in the past :

- the sound of CDs is perfect because bits are bits

- the sound of valves is the same as the sound of transistor since they behave the same way

- cables don't matter because the only thing they have to do is bring current to the devices or the speakers...

 

So, who has tried the naim and has some experience ?

 

Gilles

 

[jhwalker]

In the case of audio files ripped from CDs, if the resultant files are identical, there *can be no difference* in sound. This is as fundamental to digital data as the law of gravity. Identical files are identical.

 

That's not to say that a different hardware / software combination might be more successful in making bit-perfect extracts ("rips") from CDs - but given the resultant files are identical (as you point out, this is easily proven), the will sound identical, regardless of their provenance.

 

 

[ldolse]

I think there is a lot of marketing/FUD in that Naim document... What they say is generally true, but they're blowing many of the issues out of proportion. Copy protection isn't that common and circumvention techniques exist when it occurs. Offsets don't matter in terms of the audio quality, but not having offset correction makes it difficult to determine if your rips match the same cd on other systems. Lead in/Lead outs can cause naive rippers to miss audio, but this isn't a problem with the primary contenders.

 

Burst mode does provide for the fastest rip, but is not meant for accuracy, true. On a clean cd this doesn't really matter though, and burst mode works 9 times out of ten. This is where I still think Dbpoweramp's logic is king - rip in burst mode first, then check against AccurrateRip - if there's a match, call it good, if not automatically re-rip in a more accurate mode. The other ripping software require you to manage all that logic manually, but it can be done by reviewing the logs.

 

They don't use AccurateRip, and if you truly care about getting 'perfect' rips I don't think there is any better solution - even a secure ripper can rip something incorrectly and not be aware of it, and AccurateRip is the only way to recognize this.

 

Edit, some good reading from dbPoweramp's creator:

http://forum.dbpoweramp.com/showthread.php?17003-AccurateRip-Sold-(also-defending-AccurateRip-s-good-name)

 

[Mr.C]

"The Naim HDX is also using a CD-ROM and HDD."

 

So the theoretical difference would be quality of hardware rather than category of hardware, or software. I know that Steve Nugent was talking about the differences that he was noticing with different software rippers, but not specifically the Naim system.

 

[bixby]

yes bits are bits and remain so even when you put them into an environment where things affect the sound of those bits. Programs running in the background, memory play versus streaming from a drive, internal vs external drives, other computer system processes, wire yes wire and even the "bit perfect" player software all affect sound.

 

[esldude]

I have posted how I did extensive ripping with multiple reads to compare the rips. Almost never had it come up differently. Eventually ripped several hundred with just one read. Checked quite a few later with a multiple read rips and found no differences in the bits. I really think it simply isn't much of an issue. I have had a few discs that read differently and required a few extra reads. I couldn't tell a difference in the faulty reads and the correct reads on a very good high end system or using headphones.

 

While everyone ripping hopes to get perfect rips as they don't want to store forever more a faulty file, I don't think there is likely to be an audible difference even when those slightly imperfect rips will play properly. Now when I rip a single CD or three I do use software that does 3 reads before pronouncing it good as the time involved is a few extra minutes usually. But doing a few hundred for the first time the difference between single reads and something more elaborate is 300-500% more time to get the job done. Then you likely won't have more than a track or two per thousand tracks differ. You likely cannot hear the difference even in those couple 'faulty' tracks.

 

While no one has mentioned it in this thread, it is funny how some who talk of LP's being better will also insist on highly ritualized ripping for super accurate ripping. Yet I am pretty sure the difference in between any two playback sessions with an LP will differ more than will minor bit variances in a ripped CD due simply to different dust in different places of the LP upon playback. Cracks me up.

 

[thesurfingalien]

I believe this to be more of a marketing influenced article than one that stands to any ground for more computer-minded people you will likely encounter here.

 

But there is a category of people that do not have the knowledge (or are simple not inclined to put in any effort) to "tune" EAC (or other rippers) for perfect rips. For those who have money to spend it could be a nice solution.

 

Having said that I must confess that my faith in Naim as a no-nonsense (as far as that goes in the world of audio) is dropping rapidly. To be honest I find the article to some extent misleading and I believe it is intended to prey upon overly worried non-technical Naim-lovers to make them buy their stuff to get the best rip-results.

 

In reality, I believe that most audiophiles tend to take very good care of their software (CD's etc...), and making perfect rips should not pose much of a problem.

 

Regards,

Peter

 

[Brian A]

I've ripped a few CDs that were in really really bad shape (so bad that not all the tracks would rip).

 

The surprise was for the (really bad) tracks that DID rip, there were spots where there the errors were audible. They sounded like patches of static.

 

All this was done using iTunes with "error correction" turned on. Discovering these bad tracks is one of the reasons I switched to dBpoweramp which has more thorough error checking and which generates a report describing the quality of the rip.

 

[esldude]

Worst CD I ever ripped was quite a surprise. A co-worker had found it after it was lost in his wife's car. Spent a few months in the floorboard and under seats apparently. It was scuffed, scratched, and no player I tried it in would even read it. I had only recently at that time come to know about Exact Audio Copy. I figured it was a good test.

 

Well to my surprise EAC recovered that complete CD. This was several years ago when the machines were slower. Using a low end CDRW drive from HP, it read, and read and read tracks over and over, but after 11 hours had all the tracks on this CD. They sounded fine though I had nothing to compare it to of course. One track had a single momentary dropout, and one had two such dropouts. The others were clean.

 

Years later I compared those tracks to a clean copy of the CD. The two with dropouts didn't match. One other track that sounded clean didn't match. The remaining 9 tracks were bit for bit correct. That really kind of floored me. I would have thought there would have been more errors.

 

This doesn't prove all that much other than a program like EAC and similar multi-read ripping software is a pretty powerful tool to get accurate reads.