Pink Floyd - Dark Side of the Moon - 50th Anniversary Atmos mix in 2023! (Standalone BD coming in October!)

[inspclouseau]

I originally posted the thoughts below in the poll thread, but I am reposting it here to avoid a new conversation in a thread that it's meant for voting only. Apologies.
Here's my view:

I've ripped both discs and these are my impressions:
- the 5.1 mix is exactly the same.
- DR values are identical, but with different peak and RMS values. Seems a level shift, but there's more...
- Frequency response is 'almost' identical
- spectrograms confirm it's the same DSD master converted to PCM. But they are not the same. See the screenshot below.
- I tried to listen to the 'differences' with a null test, but the two transfers cannot be aligned; they run at slightly different speeds.
Conclusion: it's the same 5.1 mix, which has been re-transferred to PCM at a lower volume (1.5dB on average). Minor differences in speed etc. are to be expected. I couldn't notice any differences that can be heard, just differences that can be seen via analysis software.

Spectrogram comparison (On the Run 2011 Blu-ray disc, left; On the Run 2023 Blu-ray disc). The two tracks were level-matched before capturing this screenshot.

Thank you for doing this.

 

[soesbeout]

Thanks for the interesting spectrogram I was also wondering how close to 2003 5.1 SACD mix is to the bootleg DVDA.

 

[inspclouseau]

There is no "new 5.1 mix" in this set, it's the same version that originally came out on SACD in 2003 and then again on Blu-Ray in the 2011 Immersion set.

And thank you too.

 

[ArmyOfQuad]

Why would the same digital mix run at different speeds? I can see that for an analog tape, but a common digital source kept in the digital realm should be able to line up perfectly.

 

[ar surround]

I don't like the look of that "ripple" banding on the new edition. To me, it suggests some kind of error in transcoding.

What is "ripple banding" as shown in the second spectrogram? Is it those alternating horizontal narrow dark and wide light colored bands?

 

[edisonbaggins]

Question for the group, regarding the 7.1.4 (9.1.4?) rip that was uploaded recently:

Doesn't Atmos play back on a particular system in an attempt to reproduce "objects" where the engineer "placed" them?
Instead of being deadlocked to certain music in certain channels, no matter what?
If this is true, won't our systems all playback different sounds from particular channels, based on the system's understanding of the array layout?

That is my impression of how Atmos is supposed to work.

Yay? Nay? Maybe?

Thanks for any shared knowledge.

 

[zeerround]

Money, the way they did it (7.1.4):

The I would do it (as an remix from their 7.1.4, using AI stem separation):

I used UVR5 separately on the fronts and rears to get 6 steams each.

I left the rest of the original 7.1.4 channels as they were.

I used centercutGUI to split the front vocals into C and sides.

I mixed:

C vocals into C
S vocals into Fronts
Rear vocals into height rears
Front Guitar into Front
Rear Guitar into Rears
Front and Rear Bass mixed together and "stereoized" with a crossover and mixed to sides
Piano (just a few notes here and there) front and rears to heights front and rears.
Drums front and rears to height front and rears
Other front and rear (pretty much only the Sax) flipped front for rear and mixed into heights

It still needs some level adjustments and remastering, but this was a quick and dirty experiment to see what might have been possible.

 

[zeerround]

Question for the group, regarding the 7.1.4 (9.1.4?) rip that was uploaded recently:

Doesn't Atmos play back on a particular system in an attempt to reproduce "objects" where the engineer "placed" them?
Instead of being deadlocked to certain music in certain channels, no matter what?
If this is true, won't our systems all playback different sounds from particular channels, based on the system's understanding of the array layout?

That is my impression of how Atmos is supposed to work.

Yay? Nay? Maybe?

Thanks for any shared knowledge.

Yes however, in this case, the dolby reference player won't play anything in the 9.1.6 channels, beyond 7.1.4. I think they mixed it with the objects firmly in 7.1.4 channels.

Other mixes, for sure, I can get more out of 9.1.6. 9.1.6 adds 2 "wide" fronts and 2 "mid" heights, but if nothing is ever panned there or passed through there when moving, they would be empty.

 

[JediJoker]

What is "ripple banding" as shown in the second spectrogram? Is it those alternating horizontal narrow dark and wide light colored bands?

Yes. Music doesn't look like that.

 

[HomerJAU]

Doesn't Atmos play back on a particular system in an attempt to reproduce "objects" where the engineer "placed" them?
Instead of being deadlocked to certain music in certain channels, no matter what?
If this is true, won't our systems all playback different sounds from particular channels, based on the system's understanding of the array layout?

Correct. The 7.1.4 wav is what is the decoder outputs to each speaker when the user has a 7.1.4 setup. The objects have been decoded into their assigned positions which may mean they will play back in multiple channels simultaneously so they appear to be between speakers.

If a user selects 5.1.2 layout the wav file contains all the objects decoded to that layout etc.

Objects must be decoded to channels/speakers for playback. It’s the only way we can hear them.

 

[fredblue]

Money, the way they did it (7.1.4):

View attachment 90386

The I would do it (as an remix from their 7.1.4, using AI stem separation):

View attachment 90388

I used UVR5 separately on the fronts and rears to get 6 steams each.

I left the rest of the original 7.1.4 channels as they were.

I used centercutGUI to split the front vocals into C and sides.

I mixed:

C vocals into C
S vocals into Fronts
Rear vocals into height rears
Front Guitar into Front
Rear Guitar into Rears
Front and Rear Bass mixed together and "stereoized" with a crossover and mixed to sides
Piano (just a few notes here and there) front and rears to heights front and rears.
Drums front and rears to height front and rears
Other front and rear (pretty much only the Sax) flipped front for rear and mixed into heights

It still needs some level adjustments and remastering, but this was a quick and dirty experiment to see what might have been possible.

wow!! i'd love to hear your version!!

 

[styxrogan]

this thing sounds phenominal on Blade One Metas

 

[The56Kid]

I happened to listen to the ATMOS mix, and, although is not bad, AP's mix is still the foundation for me and its omission from the box set shows how much PF cares about their fans (una mierda pinchá en un palo), so here is my mierda back at them, they can stuff their overpriced, biased and revisionist box up their limey culos (and I mean no disrespect to English folks, I have many awesome English folk in my life)...
RW and DG "Brain Damage" indeed...

Que monton de pendejos!

 

[PissedOffPeoN]

I have to agree after listening to the immersion alan parsons mix again. Least i got both =-).

 

[edisonbaggins]

Correct. The 7.1.4 wav is what is the decoder outputs to each speaker when the user has a 7.1.4 setup. The objects have been decoded into their assigned positions which may mean they will play back in multiple channels simultaneously so they appear to be between speakers.

If a user selects 5.1.2 layout the wav file contains all the objects decoded to that layout etc.

Objects must be decoded to channels/speakers for playback. It’s the only way we can hear them.

That makes sense.

So, that leaves differences in room treatment, speaker placement, room correction, system quality, one's own hearing and maybe some other factors as to why those of us with the same speaker arrays, such as 7.1.4, might be hearing this mix differently, same as with other formats, like Quad and 5.1.

 

[Albert Dagger]

Following up on my previous post here and in the poll thread (apologies for posting in the poll thread), I've done a few more tests.

I compared all 10 tracks of the 2011 5.1 mix and the 2023 5.1 mix. Spectrograms and track speeds are different. Therefore, performing a null test to hear the differences proved difficult. As you can see from the screenshots below, the alleged transcoding problem applies to all tracks. The speed difference between 2011 and 2023 mixes also applies to all tracks. The variation in speed is so minimal that it can't be heard, but it is a symptom that the 2023 5.1 mix is not simply a copy and paste of the 2011 5.1 mix on a new disc. There have been some errors in the process of converting from DSD to PCM, or along the way.

Just for context, the DR values are below just to highlight the 2023 5.1 mix is slightly quieter than the 2011 5.1 mix.

--------------------------------------------------------------------------------
Analyzed: Pink Floyd / The Dark Side of the Moon (5.1 Mix)
--------------------------------------------------------------------------------

DR         Peak         RMS     Duration Track
--------------------------------------------------------------------------------
DR13     -13.42 dB   -34.27 dB      1:08 01-Speak to Me
DR11      -2.10 dB   -19.21 dB      2:50 02-Breathe (in the Air)
DR12      -2.26 dB   -21.76 dB      3:34 03-On the Run
DR12      -1.92 dB   -18.52 dB      7:04 04-Time / Breathe (reprise)
DR11      -2.00 dB   -21.32 dB      4:44 05-The Great Gig in the Sky
DR12      -1.67 dB   -19.06 dB      6:17 06-Money
DR11      -2.15 dB   -18.15 dB      7:57 07-Us and Them
DR10      -4.11 dB   -18.59 dB      3:25 08-Any Colour You Like
DR11      -1.66 dB   -20.01 dB      3:51 09-Brain Damage
DR10      -1.96 dB   -18.12 dB      2:14 10-Eclipse
--------------------------------------------------------------------------------

Number of tracks:  10
Official DR value: DR11

Samplerate:        96000 Hz
Channels:          6
Bits per sample:   24
Bitrate:           8074 kbps
Codec:             FLAC
================================================================================

I am pretty confident this is not the way this mix was supposed to be presented on the Blu-ray disc of the 2023 box set, and the only reason why I am not complaining too much is because I also own the 2011 box set.

I have also compared the two rips of the 2023 stereo remaster (one from the 5.1 Blu-ray disc and the second from the Atmos Blu-ray disc), and I can confirm the two rips are identical, and therefore, the only issue with this box set lies in the 5.1 mix, in my view.

 

[NYMo]

Yeah....different speed is always down to wrong sample rate....but has the pitch fallen as well ?
If one lays the tracks underneath each other and then use global mute to select between the two different versions the pitch change should be obvious.

 

[Albert Dagger]

Yeah....different speed is always down to wrong sample rate....but has the pitch fallen as well ?
If one lays the tracks underneath each other and then use global mute to select between the two different versions the pitch change should be obvious.

The speed difference is negligible, no pitch difference, but it didn't allow me to do a null test easily.

 

[CPR]

Downloaded, 50th Anniversary edition of DSOTM/Dolby Atmos/Amazon Music
It sounds fine, on my Sonos Arc; but I'm disappointed, with the lack of WOW factor
I guess streamed Atmos won't do it for me, like DVD/A, or BluRay audio discs
I'm NOT paying $300, to listen to the BluRay/Atmos disc
Also, the only way I can play it back, is through my IPhone 12
The Amazon Music app, is useless for Atmos, on my Apple TV 4K, or my Sony 77" TV,
as well as my Mac
C'mon, Amazon; get with the program
Thoughts?

 

[ted_b]

Thoughts?
1) Well, don't throw the baby out with the bathwater. The uncompressed DSOTM Atmos mix shows off well, but as I posted in a couple of the other DSOTM Atmos threads, it is ever so slightly warm (read: rolled off?) and that lack of immediate pop shows itself poorly when compressed via streaming. But there are plenty of other Atmos streamed mixes that sound wonderful via the ATV4K (but...see no. 2)
2) If you have an ATV4K then get and use Apple Music, at least for the trial period, and collect your favorites that you may want to get on download or physical disc later (the DSOTM box price is an outlier, albeit still disappointing).