Bogus Downloads!

[Andre Marc]

Any links to this review?

http://www.theabsolutesound.com/articles/download-roundup-led-zeppelin-led-zeppelin-ii/

"Now comes the promise of the best digitized Zep yet. Both HDtracks and SuperHiRez.com have made the band’s entire catalog available at 96/24 resolution. Further, the services are offering two different masterings of the albums. I was anxious to find out if these releases would finally substantially up the sonic quality of Led Zeppelin’s digital output."

He states there are two different masterings. There is NOT. There are two VERSIONS. Deluxe with bonus tracks and just the original album. This kind
sloppy reporting is unacceptable.

WRONG.

 

[FrantzM]

http://www.theabsolutesound.com/articles/download-roundup-led-zeppelin-led-zeppelin-ii/

"Now comes the promise of the best digitized Zep yet. Both HDtracks and SuperHiRez.com have made the band’s entire catalog available at 96/24 resolution. Further, the services are offering two different masterings of the albums. I was anxious to find out if these releases would finally substantially up the sonic quality of Led Zeppelin’s digital output."

He states there are two different masterings. There is NOT. There are two VERSIONS. Deluxe with bonus tracks and just the original album. This kind
sloppy reporting is unacceptable.

WRONG.

And the reviewer goes on to write

The deluxe editions cost about half again as much of the remastered mixes. For that you get not only better sound but also a live version of pretty much the entire album.

He got some flak from the commentators though ...
Will download these as soon as I get to Miami...

 

[TBone]

http://www.theabsolutesound.com/articles/download-roundup-led-zeppelin-led-zeppelin-ii/

"Now comes the promise of the best digitized Zep yet. Both HDtracks and SuperHiRez.com have made the band’s entire catalog available at 96/24 resolution. Further, the services are offering two different masterings of the albums. I was anxious to find out if these releases would finally substantially up the sonic quality of Led Zeppelin’s digital output."

He states there are two different masterings. There is NOT. There are two VERSIONS. Deluxe with bonus tracks and just the original album. This kind
sloppy reporting is unacceptable.

WRONG.

Thanks for the link ...

Well, I have LZII in so many guises. Different LP pressings, a Page mastered CD, some cuts in HDCD, and a host of digitized LPs to CDR, including the famous "RL" version. I also have an original digitized R2R copy on CDR & DSD. They can all sound very different.

"But forget all that, because now we have the high-res downloads—and they rock. To be sure, they rock to varying degrees, so let’s dig into them. I mentioned the availability of two versions. The first, dubbed “remastered,” is apparently not taken from the Jimmy Page masters. Nonetheless, this version sounds a lot more like what we all remember from the vinyl and easily outshines either of the CD versions. Boy, it sure is nice to hear some bass! Most importantly, the missing highs are largely restored. Channel separation is also better, which helps to create some nice soundstage width, if not depth. I would still wish for better dynamic contrast. All of these comments apply to both the SHR and HDtracks releases, which are indistinguishable."

I wonder if he's referencing the Barry Diament mixes as the "Remastered", because his description kinda (although still wrong) fits in comparison to the Page "deluxe" remasters. The "deluxe" CD (Page remaster/mix) of LZ2 was recorded hot,hot,hot, w/some obvious compression, it's far LOUDER and perhaps more "radio" friendly. On the other hand, the early Diament CD mix were recorded at much lower levels, tonally leaner they offered MUCH superior dynamic contrast/headroom.

 

[rbbert]

Page was involved in both "remasters", one with George Marino in the '90's and the most recent ones with John Davis. Barry has always said his are "flat" transfers. Unfortunately, The Absolute Sound frequently just can't be believed, and unless one already knows differently there's often no easy way to tell.

 

[Andre Marc]

Thanks for the link ...

Well, I have LZII in so many guises. Different LP pressings, a Page mastered CD, some cuts in HDCD, and a host of digitized LPs to CDR, including the famous "RL" version. I also have an original digitized R2R copy on CDR & DSD. They can all sound very different.

"But forget all that, because now we have the high-res downloads—and they rock. To be sure, they rock to varying degrees, so let’s dig into them. I mentioned the availability of two versions. The first, dubbed “remastered,” is apparently not taken from the Jimmy Page masters. Nonetheless, this version sounds a lot more like what we all remember from the vinyl and easily outshines either of the CD versions. Boy, it sure is nice to hear some bass! Most importantly, the missing highs are largely restored. Channel separation is also better, which helps to create some nice soundstage width, if not depth. I would still wish for better dynamic contrast. All of these comments apply to both the SHR and HDtracks releases, which are indistinguishable."

I wonder if he's referencing the Barry Diament mixes as the "Remastered", because his description kinda (although still wrong) fits in comparison to the Page "deluxe" remasters. The "deluxe" CD (Page remaster/mix) of LZ2 was recorded hot,hot,hot, w/some obvious compression, it's far LOUDER and perhaps more "radio" friendly. On the other hand, the early Diament CD mix were recorded at much lower levels, tonally leaner they offered MUCH superior dynamic contrast/headroom.

I had all of Zep's albums in pretty much every format..early pressings on LP, the FIRST CD (!) the George Marino CDs, etc, and the new ones.

Clearly, the new ones have less dynamic range, but not offensively so. There is a fine line in getting the SNR correct. These are hissy, old tapes, no way around that.

Making them really quiet would make that so much more noisy when you crank them up.

I think the new remasters are more punchy and vibrant than any other release.

Just to clarify..there were no remixes..these are the original mixes. Do you have any versions of the first 5 albums were the multiracks remixed?

I know of none myself.

BTW, this is just the tip of the iceberg, this download reviewer is off the deep end.

 

[Andre Marc]

Page was involved in both "remasters", one with George Marino in the '90's and the most recent ones with John Davis. Barry has always said his are "flat" transfers. Unfortunately, The Absolute Sound frequently just can't be believed, and unless one already knows differently there's often no easy way to tell.

Exactly. Disregard any TAS download reviews IMO.

The process for these remasters, and I own them all, was very transparent. tape>192/24>mastering>96/24.

 

[Brucemck2]

Exactly. Disregard any TAS download reviews IMO.

The process for these remasters, and I own them all, was very transparent. tape>192/24>mastering>96/24.

If you were going to download just one version tonight (which I'm hoping to do), which version would you choose and where would you get it from? (It's a bit hard for someone not versed in LZ recording history to parse this thread.)

 

[BlueFox]

If I have understood this correctly, there is only one version of the latest remasters. However, each album remastered so far has two versions of the albums. One version is the original album remastered at 96/24. The other is the same album with extra songs never released before. Some of these are only 44/16. Personally, for LZ I, I bought the extra song version, and didn't care for the extras. For LZ II, III, IV, Houses of the Holy, and Physical Graffiti, I just bought the original album remastered at 96/24. I bought mine from HD Tracks.

Hopefully, that makes sense.

 

[Bruce B]

Here are the 2 files that started the whole thing. As you can see, there is not really any content above 18k, even though it was a hi-rez recording in DSD.

 

[Bruce B]

Now, taking those same 2 files, I increased the gain +6dB.. Clearly you can see content above 30k.

 

[Andre Marc]

If you were going to download just one version tonight (which I'm hoping to do), which version would you choose and where would you get it from? (It's a bit hard for someone not versed in LZ recording history to parse this thread.)

Are you referring to Led Zeppelin II? Or all 6 currently available new masters?

Well the question kind of answers its self. If you are a casual fan, get the "single disc" versions instead of the deluxes. The bonus tracks, while
interesting, probably won't much value, and you will save 7 or 8 bucks.

I would download based on price to be honest.

Check the price at Superhirez.com, HDtracks (use a this is fraud) ProStudioMasters (use a this is fraud), and livedownloads.com.

The last title, Physical Graffiti, was 25% cheaper at livedownloads.com.

 

[Andre Marc]

If I have understood this correctly, there is only one version of the latest remasters. However, each album remastered so far has two versions of the albums. One version is the original album remastered at 96/24. The other is the same album with extra songs never released before. Some of these are only 44/16. Personally, for LZ I, I bought the extra song version, and didn't care for the extras. For LZ II, III, IV, I just bought the original album remastered at 96/24. I bought mine from HD Tracks.

Hopefully, that makes sense.

Actually, no.

The only Deluxe version with 44.1/16 and 48/24 tracks is Zep I, and the bonus tracks are a well known, well bootlegged Paris concert from 1969.

There is one other bonus track on Led Zep II that is 44.1/16. All other bonus tracks across the currently available 6 releases are 96/24.

BTW, most of the bonus tracks are rough mixes, alternate vocal takes, etc, but not really any unreleased songs.

 

[BlueFox]

Are you referring to Led Zeppelin II? Or all 6 currently available new masters?

Thanks for the reminder, there are 6. I, II, III, IV, Houses of the Holy, Physical Graffiti. Thanks for the alternative download sites. I have been a little irked at the high price, even with a this is fraud, at HD Tracks. A dollar here and there adds up.

 

[BlueFox]

Actually, no.

The only Deluxe version with 44.1/16 and 48/24 tracks is Zep I, and the bonus tracks are a well known, well bootlegged Paris concert from 1969.

Which is the only one where I bought the extra song version. I put my sentence in the wrong place. Thanks for clarifying that.

 

[Andre Marc]

Now, taking those same 2 files, I increased the gain +6dB.. Clearly you can see content above 30k.


View attachment 20427



View attachment 20428

Interesting!

Bruce, playing devil's advocate..what benefit does DSD bring to the table with a solo instrument recording like this with a narrow frequency range?

Why not just used 96 Khz, 24 Bit?

 

[Andre Marc]

Which is the only one where I bought the extra song version. I put my sentence in the wrong place. Thanks for clarifying that.

No worries!

The one bonus track that really blew my mind was the alternate, early mix of "Stairway to Heaven" included on IV. Totally different
presentation. Amazing to compare it to the final mix.

 

[Bruce B]

Interesting!

Bruce, playing devil's advocate..what benefit does DSD bring to the table with a solo instrument recording like this with a narrow frequency range?

Why not just used 96 Khz, 24 Bit?

The client liked the way those particular converters did DSD.
Some people like the way music sounds in DSD.

 

[Andre Marc]

The client liked the way those particular converters did DSD.
Some people like the way music sounds in DSD.

I do too!!

 

[tailspn]

DSP - Digital Signal Provider (ie:HDTracks or NativeDSD)

Hi Bruce,

While we really appreciate the mention of nativedsd.com being a Digital Content Provider (a download site business), I'd like to clarify we're just an example of a DCP (Digital Content Provider). and not one of the subject DCP's of this "bogus downloads" thread.

We also receive the Naxos distribution of several labels, including Wilson, but exclusively the sliced DSD .dff files. To these, I convert to .dsf and metadata tag them using a combination of Merging's Pyramix DSD Converter and JRiver's tag editor. The finished tagged DSD files that I upload to Amazon Web Services S3 are the identical music data content as received from Naxos.

I understand this thread is about the 176.4KHz FLAC files, which nativedsd.com does not offer. I just wanted to ensure nativedsd wasn't thought guilty by association

Thanks again for the mention!

 

[jeremya]

Thanks for the revealing analysis, Bruce!

Lest people draw the wrong conclusions about FLAC, I'd like to discuss this finding a little more:

2. Whether Naxos is doing it or the individual DSP's, someone is mucking around with your files. You can see the different file sizes, which means they are using different data compression ratios for the FLAC files.

Let me start by saying Bruce's conclusion is 100% accurate (someone is definitely mucking around with the files), however I'm concerned that people might misinterpret FLAC's "data compression" with sound compression algorithms, especially given the context Bruce provided (the pictures showing audibly compressed and damaged waveforms).

I've always understood FLAC to be a 100% lossless compression algorithm, but I've heard lots of speculation as to whether or not it impacts the resulting sound, etc. I'm going to posit that FLAC is not the problem and will seek to demonstrate that to you all here. I'm also not asserting that Bruce is suggesting that FLAC is in any way a problem. More on that later.

Bruce's observation is factual and sensible. Smaller FLAC files will cost you less to distribute than larger ones. If the compression is truly lossless, then it behooves you to crunch every file down to its smallest possible size. Lower cost for you + faster download for the end consumer + no loss of quality = everybody wins.

To prove the "Lossless" part of FLAC's "Free Lossless Audio Compression" to myself, I did a round-trip experiment with a hi-res PCM (DXD) file of Nidarosdomens jentekor & Trondheim Solistene's "Arnesen: MAGNIFICAT 4. Et misericordia". I performed the following steps:

1. Convert WAV (DXD PCM) -> FLAC with flac tools 1.3.1 at compression levels 0 (lowest), 3, 6 ('recommended'), and 8 (max)
2. Convert the resulting FLAC files -> WAV (again with flac tools 1.3.1; this is known as a "round-trip" conversion and the output WAVs from this step should be identical to the INPUT WAV (DXD) from step 1. Incidentally, this conversion is the same thing that happens when you "play" a FLAC file. The FLAC is first "de-compressed" into the original WAV (raw PCM data) and fed to your DAC.

Analysis:
Resulting files & sizes as follows:

 624,907,628 2L-Magnificat.wav       (ORIGINAL file given as INPUT to flac encoder)
 483,149,157 2L-Magnificat.0.flac    ("0" compression - lowest)
 423,566,900 2L-Magnificat.3.flac    ("3" compression)
 409,314,887 2L-Magnificat.6.flac    ("6" compression - recommended)
 400,708,221 2L-Magnificat.8.flac    ("8" compression)

 624,907,628 2L-Magnificat.wav       (ORIGINAL file given as INPUT to flac encoder (same as above))
 624,907,628 2L-Magnificat.0.wav     (from Magnificat.0.flac)
 624,907,628 2L-Magnificat.3.wav     (from Magnificat.3.flac)
 624,907,628 2L-Magnificat.6.wav     (from Magnificat.6.flac)
 624,907,628 2L-Magnificat.8.wav     (from Magnificat.8.flac)

Interpretation:
Output file size comparison suggests reveals the "round-tripped" files are probably identical regardless of compression level used. This data supports FLAC's claims that the compression level only impacts the amount of time it takes to compress the files, and the compression is truly "lossless."

Analysis:
Looking deeper, we can do a bitwise comparison of the file contents with a command-line tool that is built into Windows called "FC" (file compare):

>fc /?
FC [/A] [/C] [/L] [/LBn] [/N] [/OFF[LINE]] [/T] [/U] [/W] [/nnnn]
   [drive1:][path1]filename1 [drive2:][path2]filename2
FC /B [drive1:][path1]filename1 [drive2:][path2]filename2

  /A         Displays only first and last lines for each set of differences.
  /B         Performs a binary comparison.
.
.
.

File Compare will let us do a binary (bitwise) comparison and will tell us (bit-for-bit) if there are any differences whatsoever between two files. Here are the results of comparing the SOURCE wav file with the ROUND-TRIP resulting WAVs:

>for /R %i in (*.?.wav) do (fc /B 2l-magnificat.wav %i)

>(fc /B 2l-magnificat.wav C:\Users\Jeremy\Desktop\2L-Magnificat.0.wav )
Comparing files 2L-Magnificat.wav and 2L-MAGNIFICAT.0.WAV
FC: no differences encountered


>(fc /B 2l-magnificat.wav C:\Users\Jeremy\Desktop\2L-Magnificat.3.wav )
Comparing files 2L-Magnificat.wav and 2L-MAGNIFICAT.3.WAV
FC: no differences encountered


>(fc /B 2l-magnificat.wav C:\Users\Jeremy\Desktop\2L-Magnificat.6.wav )
Comparing files 2L-Magnificat.wav and 2L-MAGNIFICAT.6.WAV
FC: no differences encountered


>(fc /B 2l-magnificat.wav C:\Users\Jeremy\Desktop\2L-Magnificat.8.wav )
Comparing files 2L-Magnificat.wav and 2L-MAGNIFICAT.8.WAV
FC: no differences encountered

Interpretation:
WAV->FLAC->WAV ("round-trip") bits are IDENTICAL regardless of compression level.

Visual & Aural Analysis
For those who prefer it in pictures, I loaded the original WAV and the round-tripped level "8" (the most highly compressed (which should show the most deviation if any was to be found)) WAV into Audacity:

View attachment 20607

The original is on top. The WAV->FLAC(8)->WAV result is on the bottom.
I also listened to them both through a pair of highly-resolving Sony MDR-Z7s, a LH Labs Geek Out 1000 DAC, and a pre-production MIT Labs Vero Reference headphone cable (which I have on loan).

Interpretation
Visually the waveforms appear identical.
Sonically the waveforms sound identical.

Re-Compression
Lastly, some people might wonder what happens if you further FLAC compress a FLAC file. I was too, so I took my "0" level compression file and squeezed it down to an "8" level file and then did the round-trip and compare. The significant output follows:

>flac -V -8 2L-Magnificat.0.flac -o 2L-Magnificat.0-8.flac
2L-Magnificat.0.flac: Verify OK, wrote 400708221 bytes, ratio=0.829

>dir *flac
483,149,157 2L-Magnificat.0.flac         (INPUT file, created with "0" compression)
400,708,221 2L-Magnificat.0-8.flac       (OUTPUT file, created with "8" compression) 

>flac -V -d 2L-Magnificat.0-8.flac       (converting FLAC to WAV)

>dir *wav
624,907,628 2L-Magnificat.wav            (SOURCE file)
624,907,628 2L-Magnificat.0-8.wav        (from 2L-Magnificat.0-8.flac)

>fc /B 2L-Magnificat.wav 2L-Magnificat.0-8.wav
Comparing files 2L-Magnificat.wav and 2L-MAGNIFICAT.0-8.WAV
FC: no differences encountered

So, there we have it. No difference in a "re-compressed" FLAC file either.
(The eagle-eyed ones will have noticed that the file size of this "2L-Magnificat.0-8.flac" file and the "2L-Magnificat.8.flac" file are also identical... and as you might suspect a FC will reveal that there are no differences there. The FLAC algorithm is internally consistent.)

Conclusions:
FLAC compression, re-compression, and decompression at any level does not affect the resulting PCM data at all.
FLAC should not be held accountable for the any perceived or actual differences in the waveforms that Bruce discovered.*

* (Please note: I'm not asserting in any way that Bruce was ever trying to suggest otherwise; I'm trying to ensure that people don't draw the wrong conclusions from the data Bruce offered or use this data to give FLAC a "bad name." FLAC is not nor ever was the problem here!)

The following points Bruce raises are very much worthy of deeper investigation:

3. Also whatever they are using is altering the sonics because of the different loudness, which is audible. You can see this by Min/Max Sample values, RMS value and such.
4. There are different filter algorithms, taking out the noise compared with the waveforms of the original/DSP 3 and the others.

Someone needs to take these file vendors to task and let us know who is honoring Bruce's (and others') amazing efforts to give us the highest possible quality music... and who is not. People are mucking about with Bruce's work and that's deeply unsettling to me as a consumer. I've spent non-trivial money on just about every reputable download vendor. It would be nice to know I've got the "real deal" and not some filtered, squashed, mangled version of the music I paid for.

The way forward...
Let's focus our attention on the real culprit: a lack of integrity with regard to how our music is being processed by the various vendors. I'm not a supporter of DRM, but in this particular case what I would advocate strongly for is digital signing. It would be a boon to the consumer if a mastering engineer like Bruce could digitally "sign" his work in a manner that is easily verifiable by the end consumer (such that any tampering with the file would cause signature validation to fail). Then consumers would have a way of knowing whether or not the file they got was "mucked with." I've got some ideas on how to achieve this in a non-burdensome manner, but I'm keeping it close to my chest for now.

I'd like to thank Bruce for bringing this information to light. What an unsettling discovery!

I'd also like to thank tailspn and nativedsd.com for the transparency about how they're processing the files and their pledge to do no harm do the music data therein. These are the kinds of policies and processes we need (en masse) to ensure that customers are getting what they're paying for!

Shouldn't music be "worry free"?