Saturday, 16 January 2016

MEASUREMENTS: Stealth Releases of Good Remastering in Hi-Res Audio... (Alanis Morissette!)



The other day, a reader suggested that I check out the recent 2015 HDTracks remastered Alanis Morissette albums. I must admit that I'm not a big fan of Alanis (my sister was more into her albums), but back in the day, did enjoy the 2004 album So-Called Chaos. As I revisit the album again this past week, I was reminded of how poorly the original CD was mastered as witnessed below:


foobar2000 1.3 / Dynamic Range Meter 1.1.1
log date: 2016-01-14 08:26:51

--------------------------------------------------------------------------------
Analyzed: Alanis Morissette / So-Called Chaos (FLAC)
--------------------------------------------------------------------------------

DR         Peak         RMS     Duration Track
--------------------------------------------------------------------------------
DR5       -0.52 dB    -7.48 dB      2:53 01-Eight Easy Steps
DR6       -0.16 dB    -7.29 dB      3:53 02-Out Is Through
DR5       -0.04 dB    -6.81 dB      3:32 03-Excuses
DR6       -0.02 dB    -7.20 dB      4:03 04-Doth I Protest Too Much
DR5       -0.01 dB    -7.20 dB      3:41 05-Knees Of My Bees
DR5       -0.31 dB    -8.20 dB      5:04 06-So-Called Chaos
DR5       -0.38 dB    -7.08 dB      3:58 07-Not All Me
DR7       -0.63 dB    -9.12 dB      5:08 08-This Grudge
DR5       -0.13 dB    -7.32 dB      4:16 09-Spineless
DR5       -0.31 dB    -7.16 dB      4:36 10-Everything
--------------------------------------------------------------------------------

Number of tracks:  10
Official DR value: DR5

Samplerate:        44100 Hz
Channels:          2
Bits per sample:   16
Bitrate:           983 kbps
Codec:             FLAC
The waveform display above is of the first song "Eight Easy Steps". As you can see, there's significant dynamic range compression. Just by visual inspection, one can bet that the DR measurement would be significantly less than a DR10. Indeed, when put through the DR plugin in foobar, we see that this album as well as that first track measures a miserly DR5!

Sadly, this is not atypical for these CD's released in the early 2000's. Whether poor dynamic range CD releases contributed to overall decline in CD sales (not just correlate with this fact) would be hard to prove. However, I don't think it's unfair to suggest that music lovers who valued good sound quality would have found new music on CDs like these disappointing...

So, based on the readership suggestion, I took out my credit card and downloaded a copy of So-Called Chaos off HDTracks. Had a listen... Guess what... It looks like the record label has been holding out on us for the last decade!


foobar2000 1.1.6 / Dynamic Range Meter 1.1.1
log date: 2015-12-27 11:10:08

--------------------------------------------------------------------------------
Analyzed: Alanis Morissette / So-Called Chaos
--------------------------------------------------------------------------------

DR         Peak         RMS     Duration Track
--------------------------------------------------------------------------------
DR11      -0.10 dB   -13.39 dB      2:52 01-Eight Easy Steps
DR12      -0.10 dB   -13.48 dB      3:53 02-Out Is Through
DR12      -0.10 dB   -13.20 dB      3:32 03-Excuses
DR12      -2.27 dB   -15.34 dB      4:03 04-Doth I Protest Too Much
DR11      -0.10 dB   -13.26 dB      3:41 05-Knees Of My Bees
DR12      -0.10 dB   -14.26 dB      5:03 06-So-Called Chaos
DR12      -0.89 dB   -15.41 dB      3:58 07-Not All Me
DR12      -3.43 dB   -17.05 dB      5:08 08-This Grudge
DR12      -0.10 dB   -13.95 dB      4:16 09-Spineless
DR12      -0.11 dB   -15.09 dB      4:35 10-Everything
--------------------------------------------------------------------------------

Number of tracks:  10
Official DR value: DR12

Samplerate:        96000 Hz
Channels:          2
Bits per sample:   24
Bitrate:           2883 kbps
Codec:             FLAC
My goodness... What a difference compared to the fatiguing original CD release. Wider sound stage, better delineation between instrumental parts, better intelligibility of the lyrics, even the artificial "surround" effect is more "effective". Slower tracks like track 2 "Out Is Through" clearly demonstrates the nuances of her prosody and enunciation with the more dynamic master.

IMO, this is an example of the most significant progress the music industry can make at this point - not through expensive audiophile cables, weird power products, more USB doohickeys, not DSD, not hype around custom FPGA digital signal processing algorithms, and not likely MQA (clearly trying very hard with testimonial fluff here rather than answering a few simple questions in the comments here). All of those things, even if beneficial will more than likely at best be a thin "icing on the cake" on top of competent gear, a good sound room, and acoustic treatment/correction.

The fact that the audiophile press does not focus more time and attention on this simple fact with tremendous ramifications in terms of playback fidelity is an example of their irrelevance in being able to educate and advocate for improved musical fidelity; you know, something a little more relevant than tabloid journalism meant for entertainment or advertising. If the music labels can improve a recording such as this released in the heat of the Loudness Wars, then surely there's more than a decade's worth of music they can remaster properly!

Remember that even though an album like this is released as 24/96 (and it looks like there's a 24/192 option), this does not imply a "need" for the high-resolution container. This album is still not a genuine high-resolution recording. The HDTracks page clearly states "Notes: 44kHz sources mastered to 96kHz/24bit." I suspect that this was a 48kHz recording based on portions of the FFT looking like this:


Also, it looks like in creating the 96kHz mastering, it must have been processed through some analogue equipment thus picking up all the noise >24kHz, especially that 28.5kHz high-frequency signal that of course would not have been on any 44/48kHz digital source.

Though I would not consider this remastering a true "Advanced Resolution" recording as discussed previously, it certainly does qualify as an improvement over the standard CD release. This is why I'm surprised that HDTracks doesn't advertise this better (call it a "New, previously unreleased higher dynamic range remaster!"). While I'm not convinced the HDTracks "High-Res Mafia" is a particularly astute organization, I can appreciate the role and advocate that they publish the dynamic range values with the recordings they sell to indicate different and potentially better remasters as a valuable service to its customers - this goes for Pono and all the various music resellers also. I bet if more prominently advertised especially to Alanis fans that a superior master is available, it should pick up interest and sales.

Personally I'd love to see the RHCP's Californication or Stadium Arcadium get a digital remaster like this! Pretty please? (Anyone know if the 2015 HDTracks RHCP Studio Album Collection was an improvement?)

By the way, if I wanted to save disk space, I would not feel at all concerned/neurotic about using a good resampler/dithering program like iZotope RX to convert this 24/96 download to 16/48. No need for such a large 24/96 "container".

Okay. Off to the Vancouver Symphony tonight to have a listen to their performance of Zemlinsky's Die Seejungfrau (The Mermaid).

Have a great week ahead! Enjoy the music :-).

Addendum: Wonderful evening at the Symphony last night! I quite enjoyed The Mermaid - will have to look further into Zemlinsky's other works. When I'm in the right frame of mind, I really love how classical music can transport me into the flow of emotions - joy, love, adventure, tragedy - orchestrated by the composer/conductor/musicians...

The highlight of the night turned out to be the solo violinist Augustin Hadelich and Tchaikovsky Violin Concerto in D Major, Op. 35. Simply beautiful playing... He ended off his performance with an amazing encore rendition of Bach Sonata No. 2 Andante. Fantastically emotive combined with technical wizardry from the 1723 Kiesewetter Stradivarius - bravo! Couldn't help but grab a copy of his recent CD Sibelius, Adès : Violin Concertos (2014)... Looking forward to digging into the album this week.

I'm sure you've heard this from other audiophiles; it is really a treat to attend the symphony about every 2 months to get a reminder of the sound from a live acoustic performance. Valuable reminders of how essential dynamics are to the soul of music and how important it is to evoke the full palette of emotional range. A true "high fidelity" system should be able to produce these dynamics in an effortless fashion. How sad it is that these days, many of our favourite recordings are purposely made severely impotent.

24 comments:

  1. Is there any way to know if Jagged Little Pill was remastered this well? That album is only available in a 44.1/24 configuration, which is odd.

    ReplyDelete
    Replies
    1. Hi Williaty,
      Based on DR Database, there is an entry for 24/192 on HDTracks that lists it as DR12! I haven't checked it out personally...

      http://dr.loudness-war.info/album/list?artist=&album=jagged+little+pill

      Delete
  2. There are numerous releases I bought several times (LP, LP Remaster, CD, CD Remaster), and I am really fed up with with the music industry that keeps on fooling people. Just realized that the boxed versions of talking heads releases have a DR worse than the original CD releases.

    ReplyDelete
    Replies
    1. Yup. Me too :-(.

      These days I'm much more picky... I need a *really* good reason to buy yet another copy of "Kind Of Blue" :-).

      I'm resolve to not buy anything unless I have evidence that it's a better remaster either based on evidence like better DR or word of mouth from someone I trust.

      Delete
  3. Great analysis. I'd like to think there'd be a technical sidebar to Rolling Stone's or Mojo's reviews (where incidentally the average review score is 4/5, limiting the reviewers DR as well) or Stereophile's (maybe they do?) highlighting the techical quality of releases. Where else, apart from your eminent blog, can dynamic range tests/listings of material be found prior to purchase?

    ReplyDelete
    Replies
    1. Must bookmark this resource:
      http://dr.loudness-war.info/

      Invaluable!

      I believe that Hi-Fi News out of the UK will also publish analysis of hi-res downloads in their reviews... Since many of the bookstores have closed around here these days, I haven't been able to check!

      Delete
    2. Thank you! There had to be a site like that... A follow up question, I checked out a reissue SACD album on the dr.loudness-war.info site, and compared it to other measurements of another measurement of the SACD issue, and to a CD comparison. I was surprised to see that the SACD received different DR measurement levels, one on averaged 11, the other averaged 12, same top DR 13, and bottom DR 10. AFAIK there's only one SACD out there for this album, so it is highly likely that it is the same SACD issue in both measurements. One was measured with 176.4 sample rate, the other with 88.2, so it was rips to PCM that were compared. So could the difference in DR between the PCM rips of the same SACD issue be explained by the sample rate difference? Finally, I checked the CD issue from back in 1993, which is the one I have, and it shows higher DR across the board, of another 2 DR on the scale compared to the lesser SACD PCM rip.

      Delete
  4. Maybe i'm an young man with an old crancky mind, but i've only bought a new version of one of my discs one time in my life (i'm almost in my 32 years old...). Or maybe i value my money. I don't know.
    I just can't spend more and more money buying the "ultimate" version of an album and run it in Audacity with total paranoia, hopping that i wasn't scammed by the store/label/"hi-res mafia"/or artist.

    I may sound like and old broken LP, but until we have an standard medium or format (phisical disc, flac, alac...) and a standard in quality (44/16, 44/24, 48/16, 96/24...) and a standard in dynamic range, i'll will continue to buy CD's and downloads in 44/16 or 44/24. Thankfully 90~95% of the albuns i buy are classical music, and we know that this music doesn't suffer too much in the hands of modern dumb enginners.

    All that music industry is a huge circus, and it appears that they want us to be the clowns of the show...

    Best regards!!

    ReplyDelete
    Replies
    1. Well put. Clearly you're "over the hill" at 32! :-)

      Kudos on being a classical music lover at this tender age!

      Sadly, I've been too much of a sucker for "remaster" repurchases from Beatles to Dylan to Miles to The Police to Floyd... so I'm learning at age 44.

      Hmmm, sounds like I need to enroll in "Audio Suckers Anonymous".

      Delete
    2. Haha! Don't punish yourself so hard ;)

      I've made some mistakes and believed in "colorful audio unicorns" some years ago. Blogs like yours and other serious audio sites helped me to see the reality and how misleading (or shall we say... dishonest) some products are. Add that to my "music first" point-of-view of things (regardless of when it was recorded, what matter is the music) and i'm out of that "ultimate version" shopping.

      Have a good weekend!

      Delete
  5. Sorry for the somewhat off topic post, but since this pertains to hi-res downloads I thought I might post it here. Universal Music Group are water marking hi-res files that are not sold on their website. What's worse is it is audible. Some discussion and technical information in this thread: http://forums.stevehoffman.tv/threads/is-universal-music-group-using-audible-watermarks-on-digital-files.334991/page-3

    ReplyDelete
    Replies
    1. Thanks Brent. This is certainly very relevant and more so terribly disturbing!

      Wombat on Squeezebox Forum recently sent me this link as well:
      https://hydrogenaud.io/index.php/topic,89818.msg913427.html#msg913427

      Here's another link for folks:
      http://www.mattmontag.com/music/universals-audible-watermark

      With a convenient sample to listen to the effect.

      Shame on UMG.

      Delete
    2. I am taking an educated guess that it is actually quite a widespread issue, but audiophiles have not caught onto it yet. And even if they did my guess is we are too small a sector for it to matter; UMG and their subsidiaries probably control well over 60-70% of the hi-res downloads for mainstream music.

      Would you consider writing up a separate blog entry on the issue?

      Delete
    3. Hi Brent. Not sure I can add anything to this matter at the moment but will look around... Thanks for the suggestion...

      Delete
    4. This comment has been removed by the author.

      Delete
  6. "Personally I'd love to see the RHCP's Californication or Stadium Arcadium get a digital remaster like this! Pretty please? (Anyone know if the 2015 HDTracks RHCP Studio Album Collection was an improvement?)" no, it wasn't. dr5 (checked on jriver) or something...

    ReplyDelete
    Replies
    1. Oh well... Thanks for looking into this!

      Delete
  7. Thank you again, for two interesting articles (+ MQA)!

    Of the CD "Sibelius, Adès: Violin Concertos" I enjoy Sibelius a lot. Adès is too agitating for my general mood.^^

    Good information on iZotope RX. That - shrinking my "hi res" library in terms of disk space - is exactly what I want to do as soon as time permits. Can you give some advice on how to start? Currently, I am doing small conversions via the FLAC capabilities of foobar2000.

    In case you have not noticed my comment on your article about the UH-7000, part II: Tascam is currently looking at your findings...

    Best regards,
    Daniel

    ReplyDelete
    Replies
    1. I see that iZotope RX is rather pricey. For downsampling hi-res to save space, what does this app do that I can't do for free using XLD, Max, or even iTunes (I use a Mac; there are other free options for Windows and the same question would apply)?

      Delete
  8. There are caveats to interpreting dr.loudness and other DR metrics. EQ can make a big difference apart from compression/limiting.

    ReplyDelete
  9. Yeah, looks like the labels found out now that the "boombox" (2000's) and "iphone" (2010's) heads don't spend that much money as they hoped. And they turn back to the "critical listeners" pockets one more time with LP's and re-mastered CD's.
    Too little, too late?

    ReplyDelete
  10. I have my doubts about DRM numbers. For example, Ian Shepherd found that adjusting the Width control on one of his own mixes could cause DRM to vary by 5 points easily with no change in dynamics whatsoever. You have noted similar things yourself on 14 March 2014, referring to mixing bass to mono and getting DRM to jump 5 points. (And mixing bass to mono is exactly what mastering for LP does, every time).

    Final note: the two green waveform plots would be nice to see normalised for RMS (or LU) amplitude. This might show a bit more insight into dynamics than the plots you published, which say more about loudness to me at present.

    cheers

    ReplyDelete
  11. What are the tools you use to show the data of the tracks. The spectrum tool (purple) and the diagram of the peak values (green)?

    ReplyDelete
  12. If you allow me to respond to MEASUREMENTS: Tascam UH-7000 USB Interface (Part II: As an ADC) , dear Archimeo, did you ever consider this ADC; Behringer-SRC2496-BEHRINGER-ULTRAMATCH-PRO, since DSD and frequency rates above 96kHz are not really worthwhile it looks like a bargain interface.

    You asked 2 years ago for alternatives

    Some kind of review is here:http://lampizator.eu/LAMPIZATOR/TRANSPORT/behringer/Behringer.html



    https://www.amazon.com/Behringer-SRC2496-BEHRINGER-ULTRAMATCH-PRO/dp/B0002E50J0

    ReplyDelete