Do lossless files differ if downloaded from store vs. ripped from CD? Options

[kirbydoo]

Note: I had posted this thread on another audio site and copied the title. For whatever reason, it got cut off. The correct title of this thread is "Is there a difference between an .aiff/.wav downloaded from the Internet and one ripped from a CD?"

I have read threads in the past about how ripped copies of tracks from CDs in lossless formats are slightly lesser in quality than the original copies on the CD, but now I am wondering if there is a difference between copies downloaded from the Internet and ones on a CD. I am assuming that a lossless track downloaded from the web is a perfect copy, considering there is no grey area to consider (the quality of the CD, how efficient your DVD ROM drive is, what program you are using to rip and how well that works). I buy music from sites such as Beatport, Discogs, and Bandcamp, all of which seem to give pleasing results. I guess I am trying to ask, what is the difference between how you obtain lossless music from the web and how you obtain it from a CD? I know, there is a bunch of technical stuff that goes into ripping music from a CD, but how do online music stores sell lossless music?

This post has been edited by kirbydoo: Jan 13 2013, 21:32

[saratoga]

I guess I am trying to ask, what is the difference between how you obtain lossless music from the web and how you obtain it from a CD? I know, there is a bunch of technical stuff that goes into ripping music from a CD, but how do online music stores sell lossless music?

You'd have to ask the people who are producing the lossless content and selling it over the internet. My guess is that some is simply ripped from CDs and thus identical, while other content is sourced from elsewhere and thus not necessarily identical. But again, it depends on the content source in question...

[db1989]

I had posted this thread on another audio site and copied the title. For whatever reason, it got cut off.

This software limits titles to 70 characters. Your original title is far too long in this context. I’ll see whether I can edit-in a shorter one that makes the same point. Edit: great success

I am assuming that a lossless track downloaded from the web is a perfect copy, considering there is no grey area to consider (the quality of the CD, how efficient your DVD ROM drive is, what program you are using to rip and how well that works).

Not at all. As saratoga said, the lossless seller may well be ripping from a CD themselves.

My guess is that some is simply ripped from CDs and thus identical

saratoga knows what I’m about to say, but just to state it in print for other readers: Even if the store rips their files from exactly the same pressing of the release in question, which is a reasonably big if in itself, identical results can only be guaranteed expected with confidence if they are ripping by a secure method. One would hope that would be the norm – but, as saratoga said, to be certain, you would have to ask them, or consult their documentation about the origin of their files if they are conscientious enough to provide such a thing.

This post has been edited by db1989: Jan 13 2013, 22:14

Reason for edit: A guarantee? In SCIENCE?

[spoon]

I am quite certain that the companies which did the mass ripping of the older stuff, they would not have used secure rippers, it was a job done for $, they would have wanted to complete as quickly as possible (so would you if you had 100K discs to rip, as much of the effort went on getting ISRC + UPC codes correct).

This is a hunch I have, I hope to be able to test it later this year.

[greynol]

Expected? I really wouldn't approach it that way. Many people rip in burst mode with error free results quite regularly. That there is some form of verification or method to demontrate consistency does not change the outcome of the initial extraction.

I would base my expectations on the condition of the discs being ripped.

This post has been edited by greynol: Jan 13 2013, 22:56

[Wombat]

I bet not even the vendor himself can exactly tell you about the real origin sometimes.

I can tell you about one experience i had with one artist.
I bought an album that had very audible clicks in a silent part that after looking over it showed some samples broken like it was read wrong and badly interpolated like in the good old days before EAC and alike rippers arised.
After i complained at the store they told me they only sell the lossless files as they were given to them by the artist.
After 1 week there still wasn´t a solution so i contacted the artist directly after being told by the store.
Some days later i got the corrected version for download. What i recognized is that in the signatures of the artist himself always had an -send by an "i" device-
A second album i purchased by the same artist i can verify against the AR database with CUEtools and one of the songs is NOT accurately ripped but has no audible defect either.

What i say now is only a guess. It may be as simple as these lossless files sold are some iTunes rips the artist did himself because it was the easiest way to do so. You see in this case self-rips may be more secure.

[spoon]

You would hope that these label provided discs were in top condition (they should not have had a 3 year old scraping the disc on the floor as a UFO chasing batman), so that would leave manufacturing defects as the likely cause of errors.

A drive which burst rips, can often produce very small errors (be it a frame, or multiple frames), on these poor manufactured discs, where as if the drive revisits that area the correct result can often be achieved.

What is very much dependent on how many errors there are, the drive used to rip, such projects would almost certainly have used autoloaders, so would not have chosen a drive to base ripping on, some drives can be very poor at ripping audio (when compared to good drives)

I would expect that in the online stores which sell audio (mp3, aac, lossless), there are around 1% of tracks with errors (on the cat which was supplied by ripping), these errors might be quite small (a single frame with interpolation), so go unnoticed.

This post has been edited by spoon: Jan 14 2013, 00:05

[mixminus1]

...and then there's Universal Music Group, which applies very audible watermarking to many of the files it distributes to third-party online vendors, including supposedly lossless files.

In such cases, files ripped directly from the CD would most definitely be audibly superior.

[Nessuno]

My experience is that I bought many albums from eClassical web store, verified them against AccurateRip db, when found an entry, using XLD "open folder as CD" function and some of them matched completely. The ones that didn't match, often had a mismatch only in the last track (so I guess there could be a reason for that, which I couldn't figure out). Occasionally there are also mismatch in another track than the last, so maybe the hypothesis of Spoon and Greynol (not securely ripped but still nearly perfect results) applies.

I wrote a mail to their customer service asking for explanations, receiving no answer.

BTW: they give also the pdf booklets of the original CD edition, which especially for older releases are actual scanning from printed paper, sometimes rather misaligned: burst scanning mode!

This post has been edited by Nessuno: Jan 14 2013, 00:38

[spoon]

>so I guess there could be a reason for that, which I couldn't figure out

The last track is often more difficult to rip than the other tracks, especially if the disc is near its audio capacity. You can look at the confidences when ripping CDs, you might have around 200 for most tracks, with a drop off on the last track (perhaps only 5, but still noticeable).

This post has been edited by spoon: Jan 14 2013, 00:34

[Rescator]

My advise to those buying lossless tracks for digital download, and are concerned abut quality or simply a audiophile to ask the seller if these are encoded from the original master or from the CD master.

Assuming that they will not lie, then "original master" is what you want as that has most likely not the CD mastering processing that CD masters usually have gone through (compressing limiting).
Also, if they say it's the "vinyl master" that one also has processing (vinyl is usually bandpass limited to between 40HZ-16KHz to ensure the vinyl behaves as expected, you don't want the needle behaving incorrectly after all).
Likewise there may be masters that are processed and intended for lossy encoding (mp3, aac, etc.) later.

What is most desired is the master as it is in the "final mix". Resampling from 384KHz to 192KHz, 96Khz or 48KHz and encoded to FLAC is probably acceptable to most as no other manipulation of the audio is done.
However I can't recall ever seeing anyone stating what/which/how the master for FLAC (or other lossless) releases is originating from.
As capitalism goes it might benefit a seller the most to keep quiet if it's simply a copy of the CD they are selling.

[db1989]

Assuming that they will not lie, then "original master" is what you want as that has most likely not the CD mastering processing that CD masters usually have gone through (compressing limiting).

I beg to differ. Stating that the loudness war only affects CDs is giving false hope. Do you have supporting evidence or even reasoning for this claim? Despite there being no trend that I am aware of for non–CD-sourced material being miraculously free of the problem?

The kind of squashing that afflicts so many modern releases will, in all but exceptional cases, be done before any down-conversion to the 16 bits and 44100 Hz of CDs. Do not expect to find wonderfully dynamic releases simply by bypassing one format. This is an industry-wide problem.

[Rescator]

Assuming that they will not lie, then "original master" is what you want as that has most likely not the CD mastering processing that CD masters usually have gone through (compressing limiting).

I beg to differ. Stating that the loudness war only affects CDs is giving false hope. Do you have supporting evidence or even reasoning for this claim? Despite there being no trend that I am aware of for non–CD-sourced material being miraculously free of the problem?

I was implying that if a retailer is selling lossless tracks and state they are from the original master (rather than the cd master) then it is more likely that there is no "cd processing". Pretty much all cd's I've encountered in later years has had peaks at 0dB or very close to it. An flac (for example) from the original master is most likely to also have the headroom of the master.

But as I said, sellers will most likely not want to state too much of the actual specs.
Which is shame as in other industries rather extensive specs are shown for a product. (like soundcards for example).

It would be nice if the origin/type of master was specified, and what headroom and what additional processing (if any) like for radio play has been applied. Maybe RMS or EBU R128 info and so on.
High Quality lossless audio retailers are the ones that need to take that step though. A few special interest ones outline even the gear and process used (like certain classical live recordings) which is very nice.

But to bring things back to the original poster's question:

No/Yes/Maybe/probably/no idea/ask the seller (take your pick of answer )


EDIT Some links of interest:
http://productionadvice.co.uk/vinyl-mastering/
http://www.linnrecords.com/linn-what-is-a-studio-master.aspx
http://9to5mac.com/2012/02/28/mastering-en...oser-to-the-cd/


This post has been edited by Rescator: Jan 14 2013, 01:31

[kirbydoo]

Wow. Silly me. I never considered all of these factors until recently experimenting with EAC and re-ripping all my CDs I had originally ripped with iTunes. I will say that I definitely noticed some differences in certain CDs - there were little bits and pieces that EAC picked up that iTunes basically ignored altogether. It is so easy to fall prey to the belief that something mainstream or seemingly "corporate" is trustworthy and is selling you something you can rest assured is of quality. I would hope that Beatport or any of these other digital music sites aren't just buying random CDs and ripping them without care for any potential flaws in the discs etc... I think Beatport in particular gets copies from record labels, because they sell a lot of stuff that is extremely underground. I think it's unlikely that certain artists' material would come from a physical CD considering the nature of underground music... The album art is extremely makeshift regardless of the musical content etc... You can kind of get an idea of the $ put into specific stuff just by looking at the album art and the name of the label it comes from. As for Discogs, I am thinking they do the same thing, but they're weird. They'll show a certain album under one label and then they'll show the same album under another label, and one of them will be "unavailable" due to territory restrictions or something like that. Bandcamp is simply a platform for independent artists to share their music over the Internet in a more convenient way for listeners (they sell stuff in all different file types). I think some artists on that site upload their music from their own computer while others are under certain labels so I'm not sure how they distribute their music on there.

My next question is this - if I have a CD that has little "nicks" or what looks like pieces of glitter embedded in the CD, and I rip it using EAC and I get a 100% track quality, no errors reported and AccurateRip says it is accurate, then can I trust it?

[Wombat]

I was implying that if a retailer is selling lossless tracks and state they are from the original master (rather than the cd master) then it is more likely that there is no "cd processing". Pretty much all cd's I've encountered in later years has had peaks at 0dB or very close to it. An flac (for example) from the original master is most likely to also have the headroom of the master.
...
EDIT Some links of interest:
http://productionadvice.co.uk/vinyl-mastering/
http://www.linnrecords.com/linn-what-is-a-studio-master.aspx
http://9to5mac.com/2012/02/28/mastering-en...oser-to-the-cd/

You seem to got pretty excited about music reaching near 0dB while i din´t find music that sounds better when lowered in volume after finding out it should produce oversamples.
Also as you quoted Linns marketing writings i doubt they leave much headroom on any of their releases be it "Master" or not. We had a Claire Martin sample here that was freely available if i remember right and this was pretty loud stuff.

This post has been edited by Wombat: Jan 14 2013, 01:56

[Wombat]

Wow. Silly me. I never considered all of these factors until recently experimenting with EAC and re-ripping all my CDs I had originally ripped with iTunes. I will say that I definitely noticed some differences in certain CDs - there were little bits and pieces that EAC picked up that iTunes basically ignored altogether. It is so easy to fall prey to the belief that something mainstream or seemingly "corporate" is trustworthy and is selling you something you can rest assured is of quality. I would hope that Beatport or any of these other digital music sites aren't just buying random CDs and ripping them without care for any potential flaws in the discs etc... I think Beatport in particular gets copies from record labels, because they sell a lot of stuff that is extremely underground. I think it's unlikely that certain artists' material would come from a physical CD considering the nature of underground music... The album art is extremely makeshift regardless of the musical content etc... You can kind of get an idea of the $ put into specific stuff just by looking at the album art and the name of the label it comes from. As for Discogs, I am thinking they do the same thing, but they're weird. They'll show a certain album under one label and then they'll show the same album under another label, and one of them will be "unavailable" due to territory restrictions or something like that. Bandcamp is simply a platform for independent artists to share their music over the Internet in a more convenient way for listeners (they sell stuff in all different file types). I think some artists on that site upload their music from their own computer while others are under certain labels so I'm not sure how they distribute their music on there.

My next question is this - if I have a CD that has little "nicks" or what looks like pieces of glitter embedded in the CD, and I rip it using EAC and I get a 100% track quality, no errors reported and AccurateRip says it is accurate, then can I trust it?

It is pretty unlikely you hear tiny differences between an EAC and iTunes rip. The few milliseconds per track that may be different between these rips may result in a clearly audible click or may already been interpolated very well. But we talk about some milliseconds distributed over some minutes of a song.

[db1989]

I was implying that if a retailer is selling lossless tracks and state they are from the original master (rather than the cd master) then it is more likely that there is no "cd processing".

And what exactly is this “cd processing”? What special processes are mastering engineers feeling obliged to use before pressing onto a CD that they otherwise would omit, other than the obvious down-conversion if necessary?

Pretty much all cd's I've encountered in later years has had peaks at 0dB or very close to it. An flac (for example) from the original master is most likely to also have the headroom of the master.

Why is this facet of the loudness war any more relevant to CDs than it is to “the original master”?

It would be nice if the origin/type of master was specified, and what headroom and what additional processing (if any) like for radio play has been applied. Maybe RMS or EBU R128 info and so on.
High Quality lossless audio retailers are the ones that need to take that step though. A few special interest ones outline even the gear and process used (like certain classical live recordings) which is very nice.

Too bad that many “High Quality” stores are focussed instead upon gimmicks and snake-oil such as 96 kHz and 24 bit, then.

—

My next question is this - if I have a CD that has little "nicks" or what looks like pieces of glitter embedded in the CD, and I rip it using EAC and I get a 100% track quality, no errors reported and AccurateRip says it is accurate, then can I trust it?

I don’t see any reason why not. If you have any further questions about different subjects, please create one or more new thread(s) for them rather than diverging quite widely from the topic here.

[kirbydoo]

It is pretty unlikely you hear tiny differences between an EAC and iTunes rip. The few milliseconds per track that may be different between these rips may result in a clearly audible click or may already been interpolated very well. But we talk about some milliseconds distributed over some minutes of a song.

Hmm... I'm not sure. You're probably right... I'm not too tech savvy with this stuff. I just, could swear I heard more detail in the harmonizing of this particular hook of a song. The vocals were more "dynamic", I guess is the right word to use. I thought that because EAC has the whole secure mode feature and the fact that it rereads the disc to prevent errors that the results would be different.

[kirbydoo]

My next question is this - if I have a CD that has little "nicks" or what looks like pieces of glitter embedded in the CD, and I rip it using EAC and I get a 100% track quality, no errors reported and AccurateRip says it is accurate, then can I trust it?

db1989

I don’t see any reason why not. If you have any further questions about different subjects, please create one or more new thread(s) for them rather than diverging quite widely from the topic here.

Okay, sorry about that. I thought someone might say that but I figured that while I had a lot of people coming and going in this thread that I would ask that question.

This post has been edited by kirbydoo: Jan 14 2013, 02:12

[Wombat]

Hmm... I'm not sure. You're probably right... I'm not too tech savvy with this stuff. I just, could swear I heard more detail in the harmonizing of this particular hook of a song. The vocals were more "dynamic", I guess is the right word to use. I thought that because EAC has the whole secure mode feature and the fact that it rereads the disc to prevent errors that the results would be different.

Very unlikely, but since you feel much more confident with the technical aspects of your EAC rip your brain will tell you it sounds better. No doubt about this. It would be interesting to have a sample of this particular song of course.

This post has been edited by Wombat: Jan 14 2013, 02:13

[kirbydoo]

Hmm... I'm not sure. You're probably right... I'm not too tech savvy with this stuff. I just, could swear I heard more detail in the harmonizing of this particular hook of a song. The vocals were more "dynamic", I guess is the right word to use. I thought that because EAC has the whole secure mode feature and the fact that it rereads the disc to prevent errors that the results would be different.

Very unlikely, but since you feel much more confident with the technical aspects of your EAC rip your brain will tell you it sounds better. No doubt about this. It would be interesting to have a sample of this particular song of course.

I guess it is an unfortunate case of mind over matter.

This post has been edited by kirbydoo: Jan 14 2013, 02:36

[db1989]

It is pretty unlikely you hear tiny differences between an EAC and iTunes rip. The few milliseconds per track that may be different between these rips may result in a clearly audible click or may already been interpolated very well. But we talk about some milliseconds distributed over some minutes of a song.

Hmm... I'm not sure. You're probably right... I'm not too tech savvy with this stuff. I just, could swear I heard more detail in the harmonizing of this particular hook of a song. The vocals were more "dynamic", I guess is the right word to use. I thought that because EAC has the whole secure mode feature and the fact that it rereads the disc to prevent errors that the results would be different.

That’s not how read errors work. Please have a look over #8 of the terms of service and bear it in mind during future posts. It seems the placebo effect is making you hear distinctly musical differences where they cannot possibly exist. Even if one of the rips were highly erroneous, the chance that these differences would be characterisable in terms of music-theoretical concepts is so small as to be effectively impossible.

[greynol]

"Very unlikely" is overly-generous. Furthermore claims of audible differences must be backed by double-blind test results per the rules of the forum. You agreed to these rules when you registered, kirbydoo. We expect that you follow them.

[Wombat]

I guess it is an unfortunate case of mind over matter.

Maybe You may try to do you a favour and perform some proper abx test of these samples. It is pretty fascinating how many of these felt differernces go away then. In this case a simple bit-compare of both samples already may lighten up the case. Thats why i asked for a sample. Something like a small part of the songs snipped out with a wav editor for example.

Edit:

"Very unlikely" is overly-generous. Furthermore claims of audible differences must be backed by double-blind test results per the rules of the forum. You agreed to these rules when you registered, kirbydoo. We expect that you follow them.

Hehe! now that i wanted to be a nice guy the first time in ages to better my unkind image...

This post has been edited by Wombat: Jan 14 2013, 02:48

[greynol]

We all kinda descended at once; and you can pretty well count on me playing bad cop.