MUSINGS: Beyond "High Resolution" recordings. Into "Immersive" and "High Reality" content.

The other day, I was reading this interesting article about someone upgrading his computer gaming gear to the latest and greatest, only to find that it ultimately made no meaningful difference for today's games that he's playing. No surprise? (As discussed a couple months back, I upgraded to the nVidia RTX 4090 but not for gaming.)

As tech hobbyists, it's important not just to be aware of what we could buy, but more importantly, what is it that buying the hardware would achieve.

Human sensory systems are not infinite - whether it's the angular resolution of vision, or how many frames per second we need for gaming before it's no longer noticeable (upper limit I've seen maybe 75fps?). So too, there's only so much audio frequency bandwidth or dBs of dynamic range we need to more than saturate our auditory limits.

To me, it's not good enough to just be fed information from the usual magazines or advertisers or reviewers, it's more important to understand what it is that the technology is able to do and by doing so, we understand what we really need, and have a chance to anticipate what we might be looking for ahead instead of being unduly externally influenced.

In this spirit, let's take some time to consider the road ahead. After all these years, where is progress to come from? Let's think about this...

Enough resolution?

I. Have we already extracted most of what we can from stereo high-resolution audio?

Over the last decade, we have seen a fantastic merger of computer technology and the audiophile world with digital streamers and music servers now commonplace. Streaming has become >80% of audio consumption. We've seen digital fidelity objectively rise with each generation of DACs to the point where I believe that technically, performance has reached such high levels that in proper volume-matched blind listening (regardless of short or long term), it's highly doubtful there would be audible differences that actually matter.

Even better, we're now seeing Class D amplifiers at prices that provide very competitive performance, even to the point where devices like the Topping PA5 Mk II+ amplifier verifiably reproduce dynamic range and distortion comparable to hi-res DACs at 1W into 4Ω at very reasonable prices! Amplifiers like these should satisfy even the most demanding audiophile's need for fidelity and adequate loudness when paired with reasonably efficient speakers in small and medium rooms.

Yes, I know these ideas of "adequate fidelity" is likely foreign to those who read high-end magazines and imagine that DACs and amplifiers can always "sound better". That every few months, there's yet another "best sounding" component. After all these decades with improvements in amplification and DACs especially, this belief in incessant audible improvement obviously cannot always be true regardless of how much money a company asks for their "latest and greatest" component.

On the music content side, we've discussed for years now about the unlikely "value of hi-res audio". In general, as consumers we need to be more mindful of the quality of the music we buy. It's very rare to see pop, rock, electronica, modern multi-track recordings demonstrate a noise floor below the abilities of dithered 16-bits. You'll have better luck if your musical tastes are of acoustic jazz or classical recordings but even then there's definitely no guarantee.

In search of something "new", anachronistic technology like vacuum tubes have been implemented in devices to add a certain sound which can be enticing even if strictly this does not enhance "fidelity" as in transparency. Audiophiles do not necessarily prefer a flat frequency response and I do believe that a highly resolving audio system can sound "thin" or "clinical" to some listeners. For example, Rick Beato demonstrated the change of sonic characteristics in a recent video with compressors and I think we can appreciate that effect processors (whether digital or analog) can subjectively improve sound even though this means adding distortion to the original capture.

Yes, physical media like LP's can certainly sound very good if starting with a good mix/master despite obvious fidelity limitations of extracting information from grooves pressed on a piece of vinyl. Again, adding noise and other distortions from such a technology can be both euphonic and the playback ritual psychologically satisfying. LP playback is clearly not objectively high-resolution by today's standards and I see these euphonic preferences as lateral moves at best if not at times regression, not representative of actual progress in high fidelity.

Have we achieved most of the resolution potential from current "high-resolution" 2-channel digital recordings? Yeah, I think it's pretty clear that we have. Regardless of whether you spent US$10,000 on a fancy DAC or US$200 with an objectively high resolution device, honestly folks, it's not going to sound much different if you're seeking high-fidelity sound. Likewise, assuming you match a good amp with speakers of appropriate impedance and sensitivity, there's really not much difference. Don't believe me? Beyond measurements and other tests (such as listening to the AMPT recordings from decades-old CD players and compared with modern gear), go have a listen to some systems at a local dealer showroom or audio show (like this) at some point and see if you're really that blown away by the sound quality compared to a reasonably-priced modern system with good speakers in a decent room at home.

Focus on your speakers. Focus on your room set-up. Make sure the digital transfer is bit-perfect, listen to enjoyable music with adequate dynamic range (good production quality) and I believe you're all set.

II. Expansion to 3D simulated reality

So how do we proceed if we're at the phase where "high-fidelity" can be achieved quite easily these days with many objectively high-resolution hardware products?

Well, in the beginning, there was mono. No matter how good the bit-depth, how broad the frequency bandwidth, or how little distortion there was, the sound came from a single channel (or two stereo channels with the same content, hence a phantom center). Sure, with some imagination, a listener might be able to "feel" that he's inside the recording venue; this is a contortion performed by the "mind's eye" of faith. Over the years, I have seen audiophile articles espousing the preference for monophonic sound (like this). Seriously though, in 2023, other than intentionally wanting to experience the nostalgia of original mono sound of recordings from the '50s, I'll happily pick the current stereo mixes of Beatles albums every time, for example.

By the 1960's we have stereophonic 2-channel albums that can convey right and left signals, combining to create interaural timing and level changes that create the illusion of space while also achieving a good phantom center on adequate sound systems when we're seated within the central "sweet-spot". Realize that some of the earliest "stereophonic" systems were actually 3 channels of audio - Left-Center-Right. An example would be Fantasound used with the release of Disney's Fantasia way back in 1940 to show off the soundstage that Leopold Stokowski would hear when conducting the pieces throughout the movie. Many RCA Living Stereo (like Fritz Reiner & Chicago Symphony Orchestra material from the mid-'50s) and Mercury Living Presence (like Antal Doráti & London Symphony from early '60s) recordings also had 3-channel origins. The pop vocals of Nat King Cole from the '50s (Just One Of Those Things, Love Is The Thing, The Very Thought Of You) sound phenomenal as 3-channel stereophonic sound if you can hunt down the multichannel SACDs.

The issue with standard 2-channel stereo is that it's simply not natural. Even the center image isn't anchored to any true central location and we can easily tell if speakers, the DAC, amps are not very closely level balanced or if we sit slightly off center - this is why early on, 3-channel stereo was felt to be beneficial especially if the coveted sweet-spot has been taken. Live stereo recordings can certainly sound great, but how often are we fooled by 2-channel reproduction as actually sounding "real"?

Every day, we are immersed in sounds coming from all around us. Sonic reality is a 3-dimensional one which gives us cues about the rooms we're in, where the band is playing, who and what inhabits the space around us, how far a friend's voice is when she's speaking to us, and kinetically from where and how fast sonic sources are moving around us. Even though we only have two ears, every day our brains are trained to interpret a 3D sonic world thanks to the localizing effect of our individually-shaped pinnae (a major component defining our own HRTF). 

If we have achieved reproduction in the frequency and temporal domains with low distortion (as per hi-res transparency), what else can we do? The obvious answer is to better reproduce what we actually hear every day in our lives in 3D, beyond the limits of the ubiquitous 2-channel stereo which has been commonplace among consumers since the mid-1960's - coming up to 60 years now.

Over the years in this blog, I have not been shy about voicing the idea that I quite enjoy multichannel audio. I expect all my movies to be multichannel. Many multichannel SACD and DVD-A sounded great through the early 2000's. I expressed hope that multichannel would expand to interest more listeners back in 2019 with the idea that this could start with DSP-reconstructed binaural playback for the masses of headphone listeners, including younger audiophiles. And since 2021, with Apple's "Spatial Audio" streaming, we're now in the midst of that transformation which provides not just a more-real experience to be had with multichannel sound systems, but also a creative opportunity for artists to express fully "immersive" soundscapes routinely as part of their default production process.

With 3D content available for consumers to choose to enjoy, this would open up an opportunity for hardware manufacturers to find ways of reproducing that material convincingly, economically, and conveniently through innovation and good engineering. It will no longer be about simply which hi-res DAC to use (since they're all quite equivalent IMO) but more sophisticated considerations are needed such as how a manufacturer implements playback, how we optimize speaker options, placement, and use of advanced DSP for binauralization of headphone playback or maybe "beam steering" of sound radiation patterns.

While distortion measurements and frequency responses will always be important, how we measure live acoustic "high reality" playback capabilities will clearly need to incorporate other types of acoustic metrics.

How does the older guy know the younger guy is listening to dance music?
Maybe he just loves his classical virtual concert hall experience? 😉

III. Why is the traditional audiophile media against 3D audio?

With what I've said so far, is any of this potentially threatening to audiophiles who desire the best-sounding, most natural sound reproduction at home? I hope not. It should actually be the opposite!

Obviously the image above is meant to be humorous rather than mean-spirited or "ageist". But I think there's nothing wrong with calling out a certain audiophile attitude I see out there among "seasoned" hobbyists of a certain generation who have rather opinionated, subjective, views about genres of music, certain types of hardware, or newer technologies. 

This is why I do not understand why some in the audiophile media have already come out to dampen interest in multichannel or Atmos audio. It's one thing to criticize a technology if it is demonstrably suboptimal, inferior, with objective evidence (as per MQA), but as a class of technologies, I've seen multichannel "immersive" audio being negatively portrayed in the traditional audiophile media with usually nothing more than subjective idiosyncratic reasons.

For example, this past week, Stereophile published their article "The Beatles' Last Stand" by Jim Austin and Tom Fine where they concluded that:

"For those of us who care about perfectionist audio, Atmos, as conceived by record-company executives, is not the answer. We should hope for its demise."

Really? Are you guys actually "perfectionists" who use your ears to listen? Or are you now playing the technical numbers game, trying to complain without actual listening tests that 768kbps EAC3-JOC Atmos lossy streaming (the basis of streamed multichannel from Apple Music, Amazon, Tidal) is somehow not up to your "golden ears" standard? Sure, I'd prefer lossless TrueHD-Atmos over EAC3 as well, but the benefit is nowhere as big as the jump from 2-channel vs. 5.1+ reproduction!

How about this video from Steve Guttenberg a little while back?

Does it matter whether there were multichannel rooms at Axpona or Munich High End, Steve? Do you think temporary audio show rooms like those in Munich are even capable of showcasing good 2-channel quality, much less be able to do justice to a proper multichannel setup with all the noise and people shuffling in and out? Does it matter how many of his "Patreons" currently have multichannel systems? Have we learned anything factual or enduring from a video like that when we look back at it in 5 years?

Guttenberg asks "Who is this (Atmos) for?". Well, it's for you, me, audiophiles, and the future of what might be, Steve. You might not see it (or hear it) yet, but having artists access the tools today to create sounds that not all of us might have the equipment to play will pave the way for hardware innovations that the present-day old-skool audiophile media representatives like you lack the imagination to foresee. Yeah, back in the late '50s and early '60s, there was also resistance to transition from mono to stereo.

And some of this seems to have expanded into the younger (but still of the traditional attitude) audiophiles; folks like John Darko - as per the recent article "The madness of Dolby Atmos for music". (There's also this podcast with Lavorgna.)

Darko seems to suggest there are few multichannel or Atmos mixes out there which is of course not true. There are way more than "the sound of a hundred or so albums" obviously. My own collection currently consists of >600 multichannel albums on the music Server computer streamed using Roon (hey Roon, adding EAC3/TrueHD decoding or bitstreaming would be appreciated). While I would not suggest that audiophiles simply run out and make their systems multichannel-capable immediately, it does seem a bit extreme to call Atmos "madness", right?!  If you don't think you'll ever need it or want it, that's OK. I don't think I'll ever want a tube amp in my main system either, but I'm happy to have a listen, test things out, and can appreciate what it might do for some folks (like this).

IMO, with articles/videos like these, I see gross stagnation and confusion within the "traditional" audiophile media; I'm not sure what exactly they're advocating for. Just because a system is more complicated - more amps, more speakers, more wires, more space - so what!? Many audiophiles already have huge speakers and all kinds of wiring, multiple monoblocks, a couple of tonearms on their turntables, well-treated sound rooms, etc. It's not like LP playback doesn't often add bits and pieces (like a phono preamp), or adds an inconvenience factor, yet some of these guys gush over this stuff.

If we are "perfectionists" (as per Stereophile), then what's the matter with talking about multichannel systems that require more money to build, and more components, to achieve a more realistic 3D soundscape? It's not like such a system would lack backward-compatibility with current content and still will produce amazing 2-channel playback!

My suspicion with Guttenberg and Darko is that these guys don't actually have an adequate sound room to run a decent multichannel system in. I wouldn't be surprised if it's more about them adopting a negative attitude because they can't have it themselves so they're just trying to project that attitude forward - limited foresight, sour grapes, boys?

Who knows what Stereophile's position will be in the years to come. I'm glad at least historically Kal Rubinson had written extensively on multichannel audio and at one point in the past, previous editor J. Gordon Holt was also enthusiastic about the potential of going beyond two channels (circa 1994). I agree with Holt that we have been stuck in "the two-channel cul-de-sac" for way too long. With relatively little progress other than finding creative ways to make money, the high-end industry seems to have ended up ruminating on subjective non-utilitarian hype.

Given the terrible track record of the audiophile media supporting nonsensical snake oil, or their support of useless "technologies" like MQA, and their promotion of "hi-res audio" without putting critical thought into value, I'm happy to take a different stance on multichannel/Atmos and the potential of a brighter future looking at this in late 2023.

Through history, it's not like any successful consumer audio technology ever needed the approval of audiophiles anyway because we're a small group even if we can be outspoken, and we haven't shown leadership as a hobby for decades. If anything, the traditional audiophile magazines mainly promote navel-gazing, expensive, anachronistic lifestyle products that "look" a certain way or project an attitude of luxury more than delivering hi-fi quality.

Heck, given what we've seen in the last decade, broad support and approval from the magazines and typical audiophile "influencers" might even be "the kiss of death" for any technology that's trying to break through! 😱

IV. The road ahead...

So how do we proceed from here? While we cannot know the future, I think it includes:

1. Maintaining objective high fidelity (transparency to the source).

2. Embracing technologies that can provide high reality immersion.

On the "high fidelity" side, we can also look at the use of modern digital room correction which we've talked about before (like Acourate or Audiolense). Look at custom filters such as Mitch's work at Accurate Sound. This could also include techniques like Dirac, Trinnov, or Audyssey, etc.

On the "high reality" side, this usually means embracing various other forms of digital signal processing for the spatial presentation. At the most basic level when all we have are two channels, this could look like ambiophonics which reduces distortion from crosstalk while feeding our minds with the perception of external sound spread across at least 180° by extracting the interaural level and timing information already in the recordings. Whether with a basic RACE-type DSP plug-in or a fancy BACCH system, this sounds very good to me and I encourage others to give it a try. It's the kind of externalized sound I wish I could hear from headphones.

Ideally with speakers, good quality multichannel systems represent the "gold standard" for realistic reproduction of concert halls and albums conceptualized to give the listener an immersive experience (material like Jean Michel Jarre's Oxymore, with streaming multichannel and binaural versions). This means using good DACs, good amps, great speakers, in an excellent room; the usual things that also apply to 2-channel audio. Of course, in time we will see better soundbars and products like the Amazon Echo Studio which are not ideal but can still produce interesting sounds.

Headphone users can employ binauralization of multichannel content. As I mentioned a few years back, the quality will depend on how well the DSP parameters match one's unique HRTF. Hopefully in the future, we can get quick, easy, accurate customized measurements of our own ears that work well (decent try from Creative's Super X-Fi though).

Obviously, I make no claims that I can prognosticate the future with my crystal ball. I'm sure there will be mind-blowing technologies that I cannot imagine currently. Nonetheless, the ultimate goal is that artists and audio engineers should be encouraged to provide content which manufacturers can then design products to reproduce accurately for consumers to hear the intended spatial sound-field in the comforts of home or through their headphones.

That's IMO how we get closer to "the real thing" - the true "absolute sound", maybe? 😀 

V. In summary...

Audiophiles, do not fear as the music industry has started to embark on alternate media encoding systems that provide more information than decades-old 2-channel audio.

Just as we expect to hear 2 channels in a file or extracted out of the LP grooves today, one day, I believe the norm will be that of 360° immersion where instruments and vocalists can be positioned anywhere in the space surrounding the listener. This freedom is a reflection of progress and what we would expect from our creatives, music producers, and record labels.

With this extra information, the hardware reproduction paradigm will grow to include new algorithms whether incorporated in new headphone technologies, sound rooms, theater systems, even the evolved, convenient soundbar.

Take home points:

1. There will always be 2-channel stereo - nothing to fear even if this is all you ever want. When stereo was released, this didn't suddenly stop anyone from enjoying mono. Just like the transition from mono to stereo, immersive sound-field reproduction is not going to immediately supplant the dominance of 2.0 as the base standard but to start a gradual transition - in fact, it'll be slower because for most situations, 2-channels are all we need.

IMO, there is no need for neurotic, dramatic, or otherwise emotional attitudes. As discussed a few weeks ago, downmixing multichannel content is a blessing for those of us exhausted by the decades of dynamic compression in typical CDs and hi-res downloads.

Thank you Dolby for standardizing integrated loudness at -18 LKFS or less and -1dB true peaks or lower. Hopefully streaming services like Apple Music will continue to enforce this. These parameters will be significant for those of us who value sound quality! This has even made the disastrous-sounding Iggy & The Stooges Raw Power (DR1! CD) sound somewhat alright as the multichannel/Atmos mix; I assume Iggy Pop approved this.

[Not every multichannel stream sounds good though, check out The Pussycat Dolls' PCD for a listen to what more dynamically compressed Atmos sounds like. Ironically, given his recent articles complaining about dynamic compression, Darko should rejoice at the generally improved dynamic range with Atmos tracks, especially given that he also seems to like synthpop and EDM music - they sound fantastic in immersive audio!]

2. Those calling multichannel playback a "gimmick" need to realize that 2-channel stereo is not any more natural or any less of a gimmick/trick/ploy.

To me, in principle it's all the same! Placing speakers throughout the room to better replicate the soundstage is no different whether you have 2 or 12 (eg. 7.1.4)  loudspeakers. Two channels can fill the front right and left horizontally, while with 12 loudspeakers you can represent nuances of the horizontal, depth and height making the sound more believably dimensional for you and others enjoying the music, plus the subwoofer to experience the subterranean low notes on the pipe organ.

As audiophiles we should consider potential improvements like these worthy of our exploration. Especially as perfectionist audiophiles!

3.  Atmos, while the most popular system currently is not the only game in town. What we are witnessing with Atmos is the next evolution of multichannel technology from the company. Atmos is a marketing term that refers to Dolby's form of object-based audio. We've had matrixed analogue quad in the past, then Dolby Pro-Logic and its various DSP progeny from the 90's. Dolby Digital and DTS began the transition to discrete multichannel in theaters and the home by the mid-90's initially on LaserDisc. Object-based encoding with Atmos and dts:X became available in the early 2010's with consumer products shortly afterwards (Atmos was released first in theaters for Pixar's Brave in 2012). This is all part of a technological evolution over the decades, transforming sound quality even as we audiophiles became entrenched in our 2-channel cul-de-sac. That is, until now as this next generation of surround sound encroaches into the popular streaming music platforms. I encourage Spotify to join the multichannel party and Qobuz to expand lossless multichannel streaming!

Production tools for Atmos have become quite easily accessible and the cost is not particularly high (look into Logic Pro with integrated Atmos for example). In fact, as a home hobbyist, I can mix in Atmos now and monitor over headphones even without a full multichannel speaker system. This ease of having accessible tools will drive artists and audio engineers to embrace the system and ultimately make the content ubiquitous.

In the music streaming world, also do not forget about the competing MPEG-H 3D Audio, typically known as Sony's 360 Reality Audio (Tidal, Amazon Music) which has the ability to encapsulate multichannel bed tracks, objects, and higher order ambisonics (see recent 2022 paper for more info). Good to have competition for Atmos to strive for better tools and lower costs.

4. Regardless of actual age, stay "young at heart", audiophiles. Otherwise if all we care about are 20th Century artists, genres, technologies, and techniques, then as a hobby, "audiophilia" deserves to go extinct. I don't care about your age, whether you have dark hair, white hair, or even any hair, nothing else signals to the world that one is just plain old and irrelevant than holding on stubbornly to unjustified closed attitudes. I certainly hope that we as audiophiles broadly are not like this.

Immersive, spatial, "sound-field", "high reality", multichannel playback likely will require that we change the hardware we'll need in our rooms and will of course sound different. This content will dare the listener with a different perspective on the music. Artists can intentionally create experiences we've never heard before from them. Take this as a challenge to keep an open mind and see if there's room in your exploration of music and sounds to find enjoyment even if you're coming at this skeptically today!

For another perspective on this, have a look at Chris Connaker / The Computer Audiophile's article "Embracing Immersive Audio".

Further reading:
EBU Multichannel Project Group report on the "universal" 5.1 audio signal format (2002).

Teruo Muraoka - Examination of Multichannel Sound-Field Recomposition Utilizing Frequency-Dependent Interaural Cross Correlation (FIACC) (2007)

For some technical discussions on Atmos, see "How Dolby Atmos actually works! Marketing vs. reality". Also see the Dolby Atmos Renderer guide (2018), section 24.3 on Spatial Coding for some info on how the encoding system dynamically groups objects as "clusters".

Bosun Xie - Spatial Sound - History, Principle, Progress and Challenge (2020)

For further discussions on this topic, see the post "MUSINGS: Fidelity, Immersion, and Realism (FIR) - Levels of Audiophile Attainment" in January 2024.

--------------------

Happy Thanksgiving weekend to the Americans!

I hope you're enjoying some wonderful music as we enter December.

One last thing. I noticed the new flagship Topping DX9 15th Anniversary DAC is out; IMO, a beauty. It basically supports everything with balanced line and headphone outs, Bluetooth with LDAC. Nice color screen with dancing effects like VU-meters - prepare for more of this from other manufacturers! Perhaps the most conspicuous absence is support for MQA. As expected, with the next generation of products, that codec's influence has passed.