I fully understand the adjective "better" is subjective, and the question in the title requires a followup question. Better, compared to what? I use the word "better" to mean a sound that I, and likely most of us who like great sound quality, would consider better. Usually this means a recording that isn't dynamically squashed, among other somewhat illusive qualities. My comparison is between the traditional stereo versions and the immersive / Atmos versions of several albums.
Why look into this? A couple reason. First, I noticed a significant difference between a couple Atmos mixes and the stereo mixes. I thought, perhaps there's more to this immersive audio thing than just more channels. Second, it's not lost on me that an Atmos system is a fairly large commitment both financially and with respect to space. My goal here is to provide members of the Audiophile Style community with information.
Note: As part of this, I'll also show how much information is in the Atmos height channels. This obviously varies recording to recording, but based on my research, I've found that music uses more of the channels more often than movies.
Let's dig into this one and take a look at what I've found.
I first noticed something was up because of the large difference in volume between Apple Music Atmos content and "everything" else. This is talked about around the internet, with most people not understanding that it's actually a good thing. Apple places strict requirements on Atmos content created for its streaming service. The requirement that everyone audibly notices immediately is described by Apple below.
"The loudness level of a Dolby Atmos mix must be at or below –18 LUFS (Loudness Units relative to Full Scale) in order to play back correctly on a system or device compatible with Dolby Atmos."
In addition to Apple, music distributors such as CD Baby, also list similar requirements.
- For your release to be branded “Dolby Masters” on Apple the immersive source audio must meet the following requirements:
- The Dolby Atmos master file must be provided as a BWF ADM file using 24fps timecode for all tracks.
- All audio must be 24-bit LPCM audio at 48kHz.
- All deliverables must be conformed and synced to the original stereo reference masters.
- Target Loudness value should not exceed -18 LKFS measured as per ITU-R BS. 1770-4.
- True-peak level should not exceed -1 dB TP measured as per ITU-R BS. 1770-4.
- For albums where gapless playback is intended between tracks:
- Each album track must be delivered as an individual BWF ADM file.
- Each track boundary must be no more than half a frame (1,000 samples @ 48 kHz) earlier or later than the same track boundary in the corresponding stereo deliverable.
- There must be no additional silence at the end of each track when compared to the same track from the corresponding stereo deliverable.
- Please note that while ATMOS supports up to 128 channels per track, we can only facilitate submissions that do not exceed 25 GB in total size.
Here are some examples, with objective data, showing that immersive Atmos audio mixes can be quite different from stereo mixes, but not always. Pearl Jam's latest album Gigaton was mixed to stereo and Atmos. The differences between the two are huge. It's like a completely different album.
This is the track named Alright, stereo 24 bit / 96 kHz, official lossless streaming and download stereo release.
Dynamic Range (DR) = 7, Dynamic Range (R128) = 8.1 LU
Below is the same track, 24 bit / 44.1 kHz, official lossless Apple Music stereo release.
Dynamic Range (DR) = 7, Dynamic Range (R128) = 7.9 LU
The above two versions of the track look nearly identical. Now for the interesting part. Below is the 24 bit / 48 kHz, official Apple Music Atmos release, streamed in the Dolby Digital Plus codec. This is 7.1.4, but could also be played on any number of Atmos systems because the technology is adaptive (2.0, 5.1.2, 9.1.6, etc...)
The top two (of 12) waveforms are the left and right channels. It doesn't take an audio engineer to notice that these look like completely different tracks. The Atmos version has absolutely no clipping, and leaves plenty of breathing room for the audio. Given the huge difference in dynamic range, I checked to see if all the Atmos channels had an impact that skewed the results. After removing all channels of the Atmos mix except the front left and right, the dynamic range of this actually increased (13 (DR), 14.1 LU (R128)).
Dynamic Range (DR) = 13, Dynamic Range (R128) = 12.0 LU
Taking this a step further, meaning better, I looked at Moby's album Reprise. The better part of this is that I have the lossless TrueHD Atmos version to compare to the Apple Music version, as well as stereo. Is Apple Music making Atmos better or are Atmos versions outside Apple Music as good or better? Of course it all depends on the album, but in this case, it appears that Moby released the same version to Apple Music as he released on Blu-ray. The Blu-ray being lossless TrueHD as opposed to Dolby Digital Plus.
Let's look at five versions of the track named Everloving.
Below is the stereo 24 bit / 48 kHz, official lossless streaming and download stereo release. You can see clipping toward the end of the track, and the dynamic range compression pushes the music beyond the point of no return.
Dynamic Range (DR) = 7, Dynamic Range (R128) = 14.2 LU.
Below is the 24 bit / 48 kHz, official lossless TrueHD Dolby Atmos Blu-ray release in a 7.1.4 channel configuration. You can see breathing room in the same location of the track, looking at the top two channels (left channel on top and right channel below that).
Also note the information in the Atmos height channels. Atmos channels are the bottom four channels in the screenshot. Traditional multichannel is incapable of reproducing this information. You can see that each of the four height channels contains slightly different audio. This is because of how Atmos is mixed as objects in a 3D space rather than channels. It isn't possible to take a stereo track and up-mix it for Atoms, similar to how some high resolution albums are created by simply converting CD to a higher sampling rate.
Dynamic Range (DR) = 12, Dynamic Range (R128) = 13.2 LU.
Below is the 24 bit / 48 kHz, official lossless TrueHD multichannel 7.1 Blu-ray release (non-Atmos). The only difference is the lack of height channels that are present in the Atmos version.
Dynamic Range (DR) = 12, Dynamic Range (R128) = 13.2 LU.
What about the 24 bit / 48 kHz, official lossless TrueHD Atmos stereo mix from Blu-ray? Here you go. Even the Atmos stereo mix is much better than the non-Atmos stereo mix. The DR(128) of this Atmos stereo mix is slightly better than the 7.1.4 mix.
Dynamic Range (DR) = 12, Dynamic Range (R128) = 15.2 LU.
Below is the 24 bit / 48 kHz, official Apple Music Atmos release, streamed in the Dolby Digital Plus codec. This is 7.1.4, but could also be played on any number of Atmos systems because the technology is adaptive (2.0, 5.1.2, 9.1.6, etc...).
This mix appears to be the same as the official lossless Atmos 7.1.4 mix released on Blu-ray. Cheers to Moby for creating a great album that also sounds great.
Dynamic Range (DR) = 12, Dynamic Range (R128) = 13.6 LU
Looking at Bruce Springsteen's Western Stars, songs from the film, we can see something a bit different. The lossless TrueHD Atmos 7.1.4 mix of Hitch Hikin' has greater DR, but significant less DR128. Looking at the waveforms shows us more breathing room for the Atmos mix and major use of the center channel. The front left and right channels look like a different song altogether. The surround channels are also used significantly, while the LFE channel (4th from top) is nearly empty.
This is the track named Hitch Hikin', stereo 24 bit / 96 kHz, official lossless streaming and download stereo release.
Dynamic Range (DR) = 9, Dynamic Range (R128) = 10.8 LU.
Below is the 24 bit / 48 kHz, official lossless TrueHD Dolby Atmos Blu-ray release in a 7.1.4 channel configuration.
Dynamic Range (DR) = 12, Dynamic Range (R128) = 6.5 LU.
Over the years, we've all played several formats of our favorite albums. The format in and of itself, isn't indicative of sound quality. With Dolby Atmos, it's the same as it ever was. Atmos can surely be used for terrible sounding albums. However, given Apple's -18 LUFS requirement to upload Atmos content, the fact that Atmos mixes must be created with original unmixed tracks (not up-mixed from stereo), and my limited research into sound differences between stereo and Atmos mixes, I'd say we are off to a great start with immersive audio and Atmos.
If more albums are mixed like Pearl Jam's Gigaton, Atmos could be THE audiophile format of the future. Breathing room, waveforms that don't clip, and large increases in dynamic range usually equate to better sound. I can't wait to continue digging into Atmos / immersive audio and listening to more of my favorite music.