Skip to main content ↑ ↓

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: Bit-perfect rips (Read 45915 times) previous topic - next topic
0 Members and 2 Guests are viewing this topic.

Bit-perfect rips

Reply #50 –
Quote from: Porcus on
Quote from: spoon on
around 2% of the tracks ripped have errors according to AccurateRip


Estimated how? Singleton checksum with a non-singleton track?


Estimated from the drive accuracies (different pressings and discs not in the database were discounted, so discs/tracks either match or they do not):

http://forum.dbpoweramp.com/showthread.php?t=18088

Bit-perfect rips

Reply #51 –
Quote from: spoon on
You realise that 99.xxxxxx could be 99.000001 or 99.999999 hence why I did not put figures in those xx's, there huge difference between the two, and you still pull it up! I am quite sure that AR does fall within the above range, you agree or disagree?

Your choice of decimal places indicates precision that you simply do not have.  Does AR fall within the above range?  I don't know for sure and regret to say that I don't have much faith that you do either.  Your miscalculations on such things as how robust your original "CRC" was, how often discs end in non-silent samples and (more recently) how fool-proof your gap detection routine is have pretty much shaken my confidence.

Bit-perfect rips

Reply #52 –
Quote from: greynol on
Your choice of decimal places indicates precision that you simply do not have.


No, it indicates range, it is you who were implying precision by highlighting it and pulling it up. If I was indicating 99.999999% I would have written that...

Quote from: greynol on
fool-proof your gap detection routine is have pretty much shaken my confidence.


And this is relevant to AccurateRip how exactly? this is in dBpoweramp, a feature which I agree has not had the time spent on making it robust against all drives and discs, but it is early days.

Bit-perfect rips

Reply #53 –
Simply that it is quite ironic that you should open with a link on statistical significance.  I'm looking forward to your next percentage, maybe it will finally have some. 

Bit-perfect rips

Reply #54 –
Bickering on a forum when it adds nothing to the discussion isn't really necessary is it?

Bit-perfect rips

Reply #55 –
I wish this could be elevated above bickering.  Perhaps spoon can present actual evidence and honest calculation rather than a hand-waving dismissal in response to what are documented problems.

Bit-perfect rips

Reply #56 –
I need to present evidence that the handful of problem drives are statistically insignificant? why would I waste my time? if it is only to satisfying just a singular or limited user count who are not believing this. You are asking me to prove a negative.

The fact is, I said I did not believe these these drives are significant, which you took issue with, yet the reverse is fine...you can go waving the assumption around that these drives are a significant 'problem' all unchallenged, without the evidence you so demand from me...

It is getting so bad, I feel like replying to any discussion about AccurateRip with a link inviting to discuss on forum.dbpoweramp.com, so that discussions can take place in an un-hassled environment (at no time did I try to question your ability to make assumptions, because of something you got wrong in the past - ie a personal attack), and where would that leave HA.org? a place where AR discussions cannot take place because of the actions of one person?

Bit-perfect rips

Reply #57 –
Spoon, do you know how many entries in your database are affected by consistent defects? If yes, how many and how do you calculate it? I agree that the tone of the discussion is a little bit too personal.

Bit-perfect rips

Reply #58 –
Quote from: Soap on
If not using AccurateRip it is very probable that one can achieve an accurate rip, but far less certain.


As far as I can tell the probability of getting an accurate rip is absolutely identical whether one uses AccurateRip or not.  Unless something major has changed it doesn't actually make any difference to the DAE, rather it enables the user to assess the worth of the output in a meaningful and rational manner, which is a rather different thing (if a man speaks his mind while alone in the forest and his wife doesn't hear him, is he still wrong?).

Bit-perfect rips

Reply #59 –
Quote from: Takla on
As far as I can tell the probability of getting an accurate rip is absolutely identical whether one uses AccurateRip or not.


Assuming you only read once, yes. But what if you read N times and get n different answers? Then at least (n-1) of those n results are wrong. You want a/the right one. So: if one of them matches what someone else have obtained, on a different copy of the same CD? For most practical purposes, the matching one is the right one. AccurateRip helps you (A) determine the best candidate from N readings, and (B) know with a fairly high confidence that you have chosen N large enough.

(I have here assumed that all bits are really 0 or 1 out of the press, and disregarded pressing errors where a 0 or a 1 comes out undecidable even after all the error correction. Furthermore, I have disregarded error correction -- it could be that some of these n answers have information indicating they are less likely to be the correct one.)

Bit-perfect rips

Reply #60 –
Quote from: spoon on
You are asking me to prove a negative.
You have that backwards.  You are the one making a claim as to how significant the problem is.  The burden clearly falls on you.

Quote from: spoon on
It is getting so bad, I feel like replying to any discussion about AccurateRip with a link inviting to discuss on forum.dbpoweramp.com, so that discussions can take place in an un-hassled environment (at no time did I try to question your ability to make assumptions, because of something you got wrong in the past - ie a personal attack), and where would that leave HA.org? a place where AR discussions cannot take place because of the actions of one person?
About as un-hassled as some of our users posting at stereophile or head-fi, I'm sure.  I make no apologies telling unsuspecting readers that you have a history of citing probabilities that do not reflect reality, not when you think you can toss a number out there and expect that people should believe it.

BTW, of the problems that have been listed, only one of them is necessarily related to a certain make of drives.  Another issue covers all drives that cannot overread (IOW, most drives).  Another documented issue was recreated by two individuals with two completely different makes of drives (more than one model of Plextor and a drive from NEC, IIRC), indicating also that it was not make/model specific.

A question to those who may be disciplined in science, math or engineering, is it proper to cite a tolerance of something expressed as a percentage to six decimal places when you are unsure where a number falls within a window of one percent?

Bit-perfect rips

Reply #61 –
Quote from: greynol on
... is it proper to cite a tolerance of something expressed as a percentage to six decimal places when you are unsure where a number falls within a window of one percent?


It is not mandatory to read '.XXXXXX' as a percentage to six decimal places, but it could also express something like 'very high probability'. You might not like the style of replacing words by numbers, which may raise a impression of mathematical certainty where it actually isn't justified. I don't like that style, either. But continuing about exactly that aspect does only come across as nitpicking.

Your point still stands, though. Spoon continues, in what whatever form of expression, to call these problems negligible. But his reluctance to state how he came to that conclusion makes it rather look like not wanting to admit that the he is guessing.

I think the best thing would be if he actually crunched the data. If you get too personal he will just vanish to his own forum, where he can enforce by authority, that there is no problem.

Bit-perfect rips

Reply #62 –
This recalls me of Dark Shikari vs neuron2 on doom9 ... at first those childish fights are fun to read but in the end it's a waste of time for Spoon & Dark Shikari ... they got much interesting thing to code than wasting their time with frustrated moderators ...

Guess what ... we don't even care who's right ... for the curious, it's always fun to see a public execution.

Bit-perfect rips

Reply #63 –
Quote from: googlebot on
I think the best thing would be if he actually crunched the data.

Totally agree, except I don't think it is possible to trace where they come from except upon submission (Spoon, can you clarify?) I.e., the data on troublesome drives etc. must be collected first.



Quote from: googlebot on
If you get too personal he will just vanish to his own forum, where he can enforce by authority, that there is no problem.


Uhm ... that's here. On hydrogenaudio.org you might waste a lot of time pointing out logical flaws in the arguments of someone whose weaponry counts the delete button. This place has an attitude problem which fairly often keeps me from spending time writing something serious (although sometimes I forget where I am). I have never seen that abusive censorship on the dBpoweramp forums.

Bit-perfect rips

Reply #64 –
Quote from: Porcus on
Totally agree, except I don't think it is possible to trace where they come from except upon submission (Spoon, can you clarify?) I.e., the data on troublesome drives etc. must be collected first.


The already submitted data should be sufficient due to statistical anomaly, as lined out in post #47.

Lets also not forget, that the mods probably relieve us from much more trash, behind the scenes, every day, than we even notice. HA has a serious inflow of bullshit every week and the mods, especially greynol, are always at our service. They're still human beings and a lot of commitment maybe sometimes causes them to overshoot a little.

Bit-perfect rips

Reply #65 –
I appreciate that someone understands what we go through and sees through the smokescreen.

I'm merely raising issues which have not been taken into account regarding the effectiveness of AR; holes in the road to avoid which may or may not be of any concern to any given individual.  I have not made any claims that can't be supported.  That my style of argumentation is not appreciated does not make me wrong.

Other than to defend this point of view, I really have nothing more to say on the matter and would prefer not to waste anymore time on this.

@Porcus:
Regarding the delete button, please tell me which of spoon's posts I've deleted (google generally keeps a cache of our threads).  The fact is that you can't because I haven't.  There are rules that guide moderation here and while I am only human, I do my best to follow them.  The only posts that get deleted are those which do not add to the conversation (typically trolling and one troll post was removed).  Those that are off-topic generally get split to a new thread, provided there is some evidence that the discussion will continue.  Regarding my "logical flaws", can you point them out, or are you just blowing more smoke?
 

Bit-perfect rips

Reply #66 –
I am done with this conversation, it was a fishing episode...it is not AccurateRips role to correct drives which are technically faulty (drives which loose AccurateStream part way, or drives which swap channels).

I am not going to spend even 1 second determining how many discs out there finish on non-silence, for those drives which do not overread, to me 6 samples (perhaps the most common drive offset) are not that important at the end of the last track, for those that really care they know they need a Plextor drive which can overread. I also know that to detect drives which loose AccurateStream, or occasionally swap channels is not overly feasible when there are so many different pressings out there, if there was one pressing of a given disc it would be easy.

It could be argued that AccurateRip did more for secure ripping than any other piece of technology (re-reading, c2 pointers), I do not feel the need to spend an in-ordinate amount of time just to satisfy a few naysayers, right now I am working on something major (audio related), which might benefit 100,000's if not millions of people, you will hear of this in about 4-5 months.

Bit-perfect rips

Reply #67 –
Quote from: spoon on
I am done with this conversation, it was a fishing episode...it is not AccurateRips role to correct drives which are technically faulty (drives which loose AccurateStream part way, or drives which swap channels).
At least two of us already acknowledged this.

Quote from: spoon on
I am not going to spend even 1 second determining how many discs out there finish on non-silence, for those drives which do not overread, to me 6 samples (perhaps the most common drive offset) are not that important at the end of the last track, for those that really care they know they need a Plextor drive which can overread.
I don't care about this either, my point is that your answer to peoples' concerns is to tell them they don't need to worry by fabricating a statistical figure.  I don't know about other forums, but this is a scientific-minded community.  Numbers are generally expected to actually mean something here.

Quote from: spoon on
It could be argued that AccurateRip did more for secure ripping than any other piece of technology (re-reading, c2 pointers)
...as if this doesn't get argued?

Bit-perfect rips

Reply #68 –
Quote from: greynol on
Regarding the delete button, please tell me which of spoon's posts I've deleted (google generally keeps a cache of our threads).  The fact is that you can't because I haven't.


So ... your straw man did imply that this has already occured to postings by this particular member?

Oh, and Google caches do change as sites are updated. Of course I cannot prove that a non-existant posting did once exist.

Bit-perfect rips

Reply #69 –
This sounds like more trolling.  Would you care to further clarify your question?

Let's be clear:
None of spoon's posts were deleted, got it?

Not that it is any of your business, but this was the only post that was deleted from this discussion:
Quote
This discussion is very necessary because it illustrates the migthy power of Greynol, he who holds the secrets of the spheres, kepper of the truth & master of the world ...I humbly prostrate myself at his feet ...

Do you have any evidence to support the specious suggestion that I would delete a post in order to gain the upper-hand in an argument?

Bit-perfect rips

Reply #70 –
Quote from: greynol on
This sounds like more trolling.  Would you care to further clarify your question?


*sigh*

(a) It was not intended as a question.
(b) It was intended as an accusation.
© Of what? Straw man argument.
(d) Against whom? You.
(e) Based on what?
Fact1: I did not identify any user qua hydrogenaudio.org user. I did though mention the moderation rôle at the dBpoweramp forum.
Fact2: You respond by asking me to identify postings deleted, belonging to a particular poster
Fact3: Your second response adresses a particular thread only.

Obviously, the fact that Spoon has admin privileges at the dBpoweramp forum is not to say that he has been (or hasn't been) moderated in other forums.


You are actively trying to misinterpret a "there is an element in set X" statement as "there is an element in subset {spoon's postings} of X", and when this is pointed out to you as a straw man argument, you further try to restrict the statement down to intersection with the "this thread" subset. And frankly, I do believe that you are more than smart enough to know the difference.

Bit-perfect rips

Reply #72 –
You are deliberately misinterpreting statements into something you'd rather answer to and when get pointed out specifically in terms of sets and subsets how, in detail you are trying to restrict a statement into something easier refutable, then you answer "just trolling"? And this comes from one who has spent half this thread questioning honesty? Doesn't look good.

If you against all odds are really not able to distinguish an accusation like this from trolling, you should lock the gun away for a little while and consider the following suggestion:
Divide threads between those which you allow yourself to participate and those you allow yourself to moderate, and leave the intersection set empty. (And for what it's worth, I would rather see you participating with your knowledge.)

Bit-perfect rips

Reply #73 –
This thread has turned into an off-topic discussion whether moderating a forum requires besides specialized knowledge something that is called Social intelligence and whether one of the moderators has the latter.

Once again an initially interesting thread ended up in personal attacks and disparagement. Not really worth reading.
This is HA. Not the Jerry Springer Show.

Bit-perfect rips

Reply #74 –
I thought it was worth reading (since this is now a poll and no contributions are necessary.)

Please continue... 
The Loudness War is over. Now it's a hopeless occupation.