I've read your arguments in that thread and I'm not convinced.

The guide is exactly that, a guide, if people are too lazy to research what the settings do then thats their problem. Some of the stuff in the guide is outdated, like the recommended FLAC settings, but that doesn't mean it's a bad guide.

I can't say I'd be tempted to use Burst mode even with Accuraterip, I trust a secure T&C rip over Burst mode anyday - to each their own.

Your statement about Secure T&C vs. Burst w/ AccurateRip leads me to believe that you don't fully understand the differences. Consider whether you know what consistent errors are and what EAC is able to do about them.

Who are you willing to believe, Andre Wiethoff when he tells you that he uses Burst T&C and Spoon who says AR is the best assurance in getting an accurate rip or some fool who doesn't know how to do much more than copy the bad parts of what was otherwise a good guide (The Coaster Factory, RIP)?

EDIT: Didn't fully proof-read before submitting (a common mistake of mine).

I'm aware that EAC can read the same error(s) twice and come out with the same CRC, but I still believe that it's more accurate and safer than Burst T&C.

I'm not just following "some fool", he could have said "use Burst Mode" and I still would stick to T&C for my own peace of mind.

Which guide should people follow, if any?

The HA wiki where it's a group effort, of course but it still needs some work since it probably still has the same BS about using secure in order to ensure the most accurate rip.

Considering the chances are usually lower that you get the same consistent error twice in a row when using Burst mode over Secure mode, I find your statement (whats a nice way of putting it?) somewhat odd.

Do you have any evidence to back that up? I've never heard of that before - if it's true then I know a few thousand people that are using not so good settings to rip...

I know how the two modes work. I also know that burst T&C is the foundation of dBpoweramp as of R12, a ripper that is reported to do a more accurate job than EAC by several people on this forum (and I've demonstrated this for myself as well).

What exactly, in Jiggafellz guide, do you strongly disagree with?

The settings he suggests are a pretty good standard for people to use, there's nothing in his guide that would ruin a rip. All tagging etc can be done after the ripping process which won't affect audio quality.

And surely with T&C and Accuraterip the chances of <insert confidence number here> CD's to be reporting the exact same error both on test and copy are VERY low?

The chances of getting an agreement with AR on a consistent error is far lower than getting a consistent error using T&C with EAC. I'm not going to guess at numbers but I would say getting a consistent error from a damaged disc using the same drive in EAC's secure mode is higher than you think.

Coming at it from the other end, have you ever seen EAC not report an error but give CRCs that don't match in secure mode with no C2 and cache flushing?

What do I have against the guide? Well, the insistence that cache flushing and secure mode always be used. Why exactly? Because it spreads FUD. Do you realize that there are people out there who think using the native Windows interface will result in lower quality rips than using an external one?

PS: Here's the discussion where Andre says he uses Burst T&C with caching drives:
http://www.digital-inn.de/exact-audio-copy...ive-answer.html

That's just an example of people not doing their homework, my EAC settings are based off of jiggafellz guide but with a different FLAC command and I use the native W32 interface, still - using the nero ASPI layer wouldn't harm a successful rip.

His post was in 2004, maybe he has changed his ripping methods since then, maybe not.

But I still prefer T&C, I haven't had a single audible error whilst ripping that way.

But in a way, couldn't the same be said of you? Earlier you said you wouldn't want to use burst+AccurateRip over secure+T&C. How could you have more confidence in ripping the exact same disc on the exact same drive using a mode that increases the chance of consistent errors being an issue than comparing and verifying your rip against other rips done not only on different drives but also with different physical media? That doesn't make logical sense to me.

Diversity is crucial in maximizing the chance of getting to the truth.

Some people actually think they've done their homework because they read that guide.

Are you suggesting he may have somehow "seen the light" since then?

Have you ever had an audible error with matching T&C checksums in burst mode?

I don't really understand where you're coming from, I use T&C + AR. If diversity is so crucial in getting to the truth why is greynol shooting people down for recommending jiggafellz guide?



Again, if people fail to read up on things then they risk being wrong. It's the same for anything.



Yes, or maybe he just decided to rip differently. Maybe he has a big collection and doesn't want to wait 20-30 mins for a CD to rip.



No, I have never ripped in burst mode before. I simply don't trust it to be as accurate as secure mode.

Bottom line is, if Burst mode T&C was more accurate why would the program recommend "secure mode"? I'd like some real evidence which shows that burst mode is more accurate. Plus, burst mode doesn't show track quality in the log does it?

Edit: From the HA wiki:





Either the wiki is out of date or it has the wrong infromation on.

Use Secure Mode when Burst Mode fails to provide a consistent result. To me, it makes no sense to use Secure Mode for any other reason. One pass burst mode with an accuraterip match is good enough. Why put your drive through the extra stress?

Does it actually add extra stress to the drive? Even so, I've done hundreds of rips/burns with my current NEC drive and it's fine.

And for £16 who's going to complain? If it does die It's not going to bankrupt me.

It's a mechanical device. More usage = the closer it'll come to death .. It's not just that, it's the wasted time for something that gives you nothing in the way of security (just a warm fuzzy feeling) as opposed to a one pass Burst Mode accuraterip verified rip.

But for £16, I really don't worry about my Optical drive giving up, I'd simply buy another. Using the device what it's meant for shouldn't stress it, they are made to do these things. I agree overtime all mechanical stuff dies, but does that mean you're scared to use your HDD incase it dies?

The warm fuzzy feeling I get is what I want, I also want to know that the ripping program has checked the same sectors multiple times to ensure a proper rip, not just zipped over it without checking.

This thread has REALLY confused me...

A silly analogy. Would you rather buy a car with 10,000 miles on the clock or 100,000 miles? Another silly analogy. If you don't care about the money and time, fair enough.


Accuraterip verification (read mode is irrelevant) is stronger security than any Secure T&C, especially if done with just one drive.
Consistent errors aside, why would a Secure Mode T&C with a matching CRC make you feel better than a Burst Mode CRC or better yet, what is the magic number of passes with consistent results that is acceptible to you? That's all that it boils down to, consistent errors aside ...

But I use accuraterip AND secure T&C. Why are you limiting it to either Secure T&C OR Accuraterip?

Secure mode T&C makes me feel better because as it says in the wiki, it tests the sectors multiple times to check they match, if not it tells you.

Remember how this started, Greynol was against people recommending jiggafellz guide to people - I still don't have any real evidence to show that what jiggafellz suggests is "wrong". It's the most accurate/secure way of ripping a CD.

From the wiki:





That alone says that Secure mode is more accurate. It even goes on to say how Burst copy is worse as no sync is done and no error checking.

To me, Secure mode is the first step in ensuring a "proper" rip, Accuraterip is just a nice bonus make sure that the results you get are accurate.

Hence the suggestion for one pass burst mode in my first post. Having a good idea of what a matching CRC is giving me I'm happy with burst mode T&C as opposed to a secure mode T&C. This doesn't even take into account the increased probability of consistent errors as greynol has pointed out. You've even been told the developer of EAC himself uses this method yet still fail to believe.

I'm predominantly a dBpoweramp user now so I'm bowing out of this.

I'm not failing to believe what you're saying. I personally prefer secure rips, if you really think burst mode is "better" then that's your opinion.

This still doesn't explain why jiggafellz guide is so bad. If you have no clue about ripping, his guide will at least ensure decent, secure rips.

In any case, if matching CRC's is all you are after what makes you dislike secure mode so much? If people are willing to spend extra time ripping a CD, get matching CRC's - why do you care?

Afterall, matching CRCs & verification from AR is pretty solid to me.

Sounds to me like you personally prefer to remain clueless.

Thing is, I know a few thousand people that "swear" by the secure method, whereas you say you know people that "swear" by the burst method.

I prefer to have evidence before switching my entire ripping routine, the fact that "secure" is meant to be more secure just clinches it for me.

Unless an EAC dev is willing to show me evidence that burst mode is more accurate I'll continue using secure mode.

/shrug

I already explained (in part) why I'm shooting the site down. I'm not shooting people down, except when they demonstrate such a high level of stubborn ignorance.

<speechless>

Let us know when you're done jumping up and down with your fingers in your ears continuously yelling "I can't hear you".

http://www.hydrogenaudio.org/forums/index....st&p=512751

Burst T&C with matching checksums is no less accurate. There is a difference. Do you even know what the track quality represents?

That information was copied directly from the official website.

Please tell us what it is that you think is right.

This should be amusing.

Why are you being so aggressive?

The entire reason for this thread was that you advised someone to stay away from jiggafellz guide:



I still haven't recieved an answer as to what you mean by "sub optimal". Burst mode might be good enough for some people, some prefer to use secure mode, I still don't see why his guide should be avoided.

You are acting as if people that rip securely are doing it completely wrong, even though you say that burst mode can give just as accurate results - then I'm failing to understand your problem with the guide. IF someone wants to use the guide, they can, it doesn't give you any problem rips. Just because YOU don't like the guide it certainly doesn't mean that others should avoid it.

The official site even explains the different extraction methods,



So I ask, is the official website wrong/lying?

Sub-optimal:
Telling someone with a non-caching drive to use cache flushing.
Telling someone to always use secure mode.

I question whether someone would make such recommendations if he actually understood what was going on.


No. You were right about information being outdated earlier as that was written before all drives had "Accurate Stream".

If you use burst mode with T&C, the error checking is done by you. If the two CRCs don't match there was an error with at least one of them. If they do match then the rip may either be accurate of may have a consistent error. Generally speaking this is no different than ripping in secure mode without C2 pointers and getting 100% track quality.

Now imagine a situation where there's an error on the disc and extracting this section resulted in the right value 40% of the time, the same wrong value 35% of the time and a finite number of wrong values, each occurring with far less consistency than the other two cases 25% of the time. In secure mode, EAC will read over this area and will likely spot the bad area and begin re-reading. When it is done re-reading there's a good chance that it will return either consistent value without reporting an error(*). Repeating the entire process twice could easily return one of two sets of matching CRCs. How do you know which pair is correct? Are you as likely to see the same thing happen in burst mode?

(*) and the chances go up when you increase EAC's error recovery quality setting, something else I guarantee the author of that site does not understand.

This is off topic, but why couldn't EAC just report that a sector was reread and report that as an "error." To make users feel better, Andre Wiethoff can program the message as a "Warning"- not as scary of a word that way.

Good point. There's the "track quality" figure, but unfortunately EAC often cries wolf at the end of a track, making this a less effective warning.

I see where you're coming from, I agree with what you're saying.

However, Jiggafellz guide is a guide to "secure" ripping with eac, thats why he tells people to use Secure mode, he doesn't make any claims that secure mode is better/more accurate than burst mode.

No, but somehow it seems to reinforce people like you who come along suggesting that Burst T&C is not secure.

Burst T&C is secure and based on the way the two modes work, burst mode is less likely to give you a pair of matching checksums for an error than secure mode. IOW, burst T&C is more likely to reveal a problem with a rip than using secure mode.

Now are you interested in discussing the utter silliness of using cache flushing with non-caching drives?

People like me? What's that supposed to mean? Because I prefer secure mode rips I'm wrong or inferior?

Burst T&C isn't a secure rip. It hasn't been done in secure mode - no matter HOW accurate burst mode may be. It would be silly to write a "secure rip guide" then tell people to use burst mode.

He does assume alot in the guide I'll give you that, but the settings he recommends works on most of the latest systems.

I'm not really interested in continuing this, we obviously aren't going to come to an agreement, this will just continue on and on

Lets just agree to disagree.

From the wiki: "Secure ripping is the process of making sure there were no errors during the extraction of audio from a CD."

Burst T&C is also a secure mode especially if it is verified by accuraterip. The secure mode does not have a monopoly on making sure there were no errors during the extraction of audio from a CD.

@ greynol

I think you are beating a dead horse, here. If Phixion doesn't understand after your fine explanation, then I think that you are wasting your time trying to prove it to him.

Prove what? He thinks people should avoid jiggafellz guide, I don't - it's good for what it is, a secure ripping guide.

Other than that were pretty much agreed. I agree you can have a 'secure' burst mode rip, but I won't switch from my secure mode ripping. I have more faith in secure mode, maybe I'm wrong... but after thousands of CD's ripped this way I haven't had any audible errors from "perfect" secure rips.

Well, if you agree that you can have a "secure" burst mode rip, then there is more than one way to acheive a secure rip. Burst mode is secure with T&C and with Accuraterip.

You will do, what you will do! Though, anyone else who reads jiggafellz will be misinformed of their "secure" options using EAC. That's not good.

Phixion,

Your definition of secure is obviously based solely on hitting that secure mode radio button in EAC without any understanding of what it is, still, after what has been explained to you. I'm completely flabbergasted.
No longer relevant.

Many drives interpolate, so unless the damage is severe you are unlikely to hear pops, your secure rips could have errors and you might not know.

Burst Ripping is secure (if done correctly), we made our whole secure mode around burst passes, I would not change it.

One final point, about drive failure due to ripping too much, drives deteriorate over time (that is their ability to recover from errors is lower), so if you are stressing the drive 2x as much as you could, in comparison to an identical unstressed drive, the unstressed drive would be giving better secure results at disc 1000 than the stressed drive.

Hi,

I personally disagree with everybody ;-) I believe that T&C in secure mode is not secure enough, and T&C in burst mode is not secure enough.

Ok, I will probably sound like a broken record (vinyl, not CD). I personally don't care about the theoretical superiority of one ripping method over another and how much better it is on paper. The only thing that matter to me is the result, and I try to work based on hard data.

I've found consistent errors with EAC in secure mode.
I've found consistent errors with EAC in burst mode.
I've found consistent errors with dBamp reference.
I've found errors in the AccurateRip database.

I personally think that in most case I would not hear the difference introduce by the error. But I don't have the time to audition every single track I rip to check if it has audible error or not (definitely a 1X process). So, I rely on diversity to detect errors, and I believe I've been pretty successful at maximising diversity.

This is what I personally would consider secure, from most secure to least secure :

1) AccurateRip match, with confidence greater than 5 and consistent confidence (less than 20% deviation).

2) T&C, on two drives with different chipsets (different offsets), one pass in burst mode, one pass in secure mode.

3) T&C, on a single drive, one pass in burst mode, one pass in secure mode.

I realise that (2) is probably not possible for everybody (people using laptops), which is why I included (3). I've found that (3) adds more diversity that I would have guessed. Also, I would advocate for the secure pass above to be done with C2, to maximise diversity from the burst pass (i.e. different block size), but again, it's not always possible (get a better drive, dammit !).

I personally don't have advice for dBamp users. From my test, I believe that dBamp does not have enough diversity in the ripping process (all passes are exactly the same) and was not giving me those expected better results. But, other people may have other experiences.

You might wonder why I care what other people do and waste my time writing this. Well, I would hate other people to pollute the AccurateRip database. If every single AR user converge to the same ripping method (i.e. all use dBamp, or all use EAC burst), then we will loose some of the current diversity from the AR results. So, I would prefer for people to continue using different methods/tools (as they do today) or to use a method having enough diversity (dream on).

Have fun...

Jean

Why stop there when switching between the use of C2 pointers and read twice and compare (non-C2 rips) can also reveal errors?

I have encountered multiple instances where Burst and C2 rips delivered one CRC while non-C2 rips delivered another CRC. Also, non-C2 rips will use a different block size from C2 rips so the beginning of each new read will occur at different places and quite probably the overlap in order perform synchronization will begin at different places on the disc as well. For burst rips, the situation is different in that no synchronization is performed.

Oh yeah, I totally agree. The more diversity the better. But you have to start somewhere.

I believe that the secure-C2 and secure-non-C2 are quite similar in the way they operate (read command used, synchronisation, code path), whereas the burst was more different. Spoon claim that C2 work (and I've seen that), which balances the case you describe above.

So, to cover all bases, you would need to try all three, burst, secure-C2 and secure-non-C2. However, there are practical issues.

First, doing 3 different passes add more time, especially that the non-C2 pass tend to be slower than C2 passes. Agreed, in around 60% of cases, you have a AR match and don't do the second and third pass.

Second, the current UI of EAC only allow to compare two passes, the Test and the Copy passes.

Third, at this level of paranoia, you are better off using a second drive, I believe that will give you better diversity than three different modes on the same drive.

T&C is well established with people, and they understand it well. Changing mode between the T and the C is a small change to the procedure, and there is hope that people would start to adopt that (I wish EAC was easier). Asking them to add a third pass is a much more significant change to their workflow.

Note that I've submitted a wishlist to Andre to make mixed-mode T&C easier to do with EAC, maybe other people could convince him...

http://www.digital-inn.de/133042-post201.html

It's all about different level of paranoia, and the probability of an error to slip through. I can not guarantee that the most paranoid procedure would catch all errors. So, in other words, nothing is 100% secure. Also, I believe that people need to make their own mind and balance it with convenience.

Thanks, and have fun...

Jean

I'm not so sure about this claim, especially concerning the read command that is used.

dBpa and EAC are totally different animals when it comes to secure modes and also when it comes to the use of C2 pointers. None of my comments or examples in this thread take dBpa into account unless I've mentioned it directly.

In terms of similarities, when EAC is able to deliver consistent results in burst mode with different results in secure mode with no C2, dBpa results will usually always agree with EAC burst results. So if it's determined that there is a consistent error in EAC using burst mode, it's affecting dBpa as well.

Let me clarify, then. In the first post you said, "I can't say I'd be tempted to use Burst mode even with Accuraterip, I trust a secure T&C rip over Burst mode anyday..." That means you hold secure+T&C above AccurateRip verification. In response I posed a question about how can you have more trust in a rip from the same disc on the same drive in a mode that increases the chance of consistent errors being a problem vs. verifying a rip against other rips done with different media on different drives (which, since it apparently wasn't clear, provides a great deal of diversity). That's where I was coming from - I was addressing something specific you said.

I'm not sure how you connected diversity with jigg's guide within the context of my post. In fact, I never mentioned that guide at all.

It wasn't aimed at you, it was towards greynol

Sounds like you're asking him why I'm shooting people down if you ask me.

After reading Jean's excellent commentary do you now have a better idea as to what diversity means?

FWIW, my recommendation to begin ripping in burst mode is only aimed at people with caching drives. Because I use FUA and C2, I usually don't rip in burst mode.

For people with caching drives that don't or can't use C2 pointers...
http://www.digital-inn.de/129988-post4.html

I'm English, I understand perfectly what diversity means, it seems you don't though.

You are making out the burst mode is good enough so why bother using secure mode, that is not diversity.

I'm sure greynol meant something in the sense of "the importance of diversity in the context of secure ripping" and not just "the meaning of the word diversity".

Hi. I've read this topic a few days ago, and I'm suprised to see there is still an argument here.

The original argument was that greynol said he did not like Jiggafellz's guide and similar internet guides because it recommends a "sub-optimal EAC configuration", and Phixion is defending the guide saying it's "a pretty good standard for people to use, there's nothing in his guide that would ruin a rip."

I do think using Jiggafellz's guide is not going to make your rips less "secure," but there ARE settings in the guide that are recommended without any good reason for them. Guides like this are very misleading. I think they give people the wrong idea which settings are making your guide more secure. I used a similar guide 2 years ago when I first started ripping, and I used the disable cache option even though my drive didn't cache.

IMO, there are two possible reasons why guides like this recommend these settings.

1. The maker of the guide does NOT understand what the setting actually does, and could simply be copying it from another guide.

2. The guide was made for people to use so they can share these rips over bittorrent or something. I mention this because when people trade rips like this, they look at other people's logs. When most people on these are looking at logs, they ASSUME that if someone didn't select the disable cache option, used C2 pointers, or ripped in burst mode, that they are NOT selecting these options because it was the most efficient mode for them, but because they are stupid and simply misconfigured EAC. I think it's likely that Jiggafellz's guide and similar guides recommend these settings only to make people use the same settings, regardless of how efficient it is for them, in order to set some sort of standard across bittorrent sites.

Oh, and one last thing regarding Phixion's last post.

You quoted greynol saying that his recommendation to begin ripping in burst mode is only aimed at people with caching drives, but you reply saying that he is making out the burst mode to be good enough so why bother using secure mode. He clearly doesn't mean that everyone should use burst mode because it's good enough for everyone. The settings you use should depend on the drives you have.

Yes, like "Secure Mode". That's the whole point of this argument. Because I guess you didn't had secure mode in mind when you wrote that.

Speaking of cache flushing, here's a very interesting situation where use of cache flushing has resulted in a bad rip where not using cache flushing resulted in a good rip...
http://www.digital-inn.de/133067-post36.html

This thread evolved over a couple of weeks. If you're into this kind of thing, you might want to read from the beginning.

Rips with errors resulting in repeated samples in multiples of 6 often dog many drives. This is just one example. I can point to others if it interests anyone. The data clearly points to a consistent manufacturing problem with the pressing of the disc in question. Jean recently posted a great deal of data for a disc giving similar problems demonstrating drives giving consistent and wrong results in burst mode as well as in secure mode with both EAC and dBpa.

Wow, that is interesting. I didn't think that setting in particular could actually cause problems. The HA wiki article about drive options didn't mention this. It sounded like the only harm with this setting for people who didn't need to use it was the reduction in ripping speed and the shortening life of the drive.

I was trying to be consistent with what was already said. I just wanted to make a point that guides like this did confuse me about the settings I should be using.

> The HA wiki article about drive options didn't mention this.

No, and I don't believe that it should. I think it's safe to consider this situation an exception. If you're paranoid that you're getting consistent errors, the best thing to do is follow Jean Tourrilhes's advice. Maybe the wiki could use another article discussing paranoid ripping techniques, I don't know.