The Dolby Atmos Music Target Curve. And on being an audiophile "tonmeister".

[For the post this week, let's take a pause from the Pacific Audio Fest 2024 discussions (last 2 weeks). Next week we'll have a Part III discussion about audio shows.]

Standardization is generally good.

I've discussed in previous posts that Dolby Atmos defines loudness characteristics for recordings and this has helped preserve dynamic range in multichannel/Atmos albums following their guidelines. Likewise, we can benefit from standardization of the frequency response as this will allow us to better "translate" the sound from the studio into our sound rooms. This idea is not new for audiophiles. The ill-fated (with good reason) MQA promised that we could experience the "sound of the studio", but they never could deliver in a large part because there was never any standardization for essential parameters like the frequency response, among other failings. In the same way, back in the day (~2014), this was the promise of hi-res audio by Neil Young and Pono that you would somehow get the "finest digital copy" and this would represent the sound the artists intended from the studio; clearly they overpromised and under-delivered by 2015.

It's one thing to standardize between the studio and home ("inter-system") playback, but it's also important to standardize "intra-system" - that is, between all the speakers in your set-up when you're doing multichannel such that each speaker is calibrated to the same timbre. This is important in a multichannel layout since we do not want "objects" moving through space changing tonality because of mismatch between speakers around the listener. The Dolby Atmos Music Target Curve is Dolby's recommendation for the music studio frequency response. This curve evolved from the X-Curve, an 'X'perimental frequency response based on some early empirical work exploring the performance of actual theater sound systems starting in the 1970's, and evolved over the years. See this paper for more details, note some similarity of the Dolby Atmos Music Target Curve to the SMPTE 202M - 1998 X-Curve.

As with all things in audio, nothing wrong with asking whether this curve is the "best". 🤔

From the latest September 5, 2024 Best Practices for Dolby Atmos Music Studios document - here's Dolby's frequency response curve:

While there is an expectation for the target curve to resemble what you're seeing above, note that Dolby specifies the frequency response for speakers to "extend from 40Hz to 18kHz with no greater variation than +/- 3dB" in the music production studio. We all know that nothing's perfect, so there's wiggle-room.

We can easily extract the parameters from the curve above (also see this thread on the Dolby support forum). Here's my edited version of the values with simplifications (based on numbers from the support forum from MattD) and expansions that I use at home for my full-range 20Hz-20kHz system:

These values plotted out results in the graph at the very top of this post with +/- 1dB glow applied around the official Dolby values. Dotted-line below 32Hz to represent the natural roll-off of bass-capable speakers/sound systems without a subwoofer to sustain sub-bass down to 20Hz. While Atmos productions can use 96kHz samplerate in the Master files, the encoding is always at 48kHz so maximum frequency response is 24kHz, and lossy EAC3-JOC content as streamed from services like Apple Music "Spatial Audio" typically cuts off around 21kHz.

Using my multi-seat measurement previously discussed, let's then apply this room curve to the Audiolense multichannel room correction. Here's a look at the Target Designer window and the specific data points in the Chart Editor window:

Note that my Data values use -6dB at 1kHz as the "0" point. Instead of single 20Hz point, I used 18 & 23Hz control points to achieve desired smooth roll-off.

My Paradigm Signature SUB1 subwoofer has no issues going below 20Hz in the room but I prefer not to have excess rumble way down there and would rather the quick roll-off below 20Hz.

[Nah, no need for adding 7.83Hz Schumann Frequency here on planet Earth. Could be important beyond near-earth orbit though - here's an inexpensive unit to bring to the Moon.🙄]

In the same way, it's best to roll-off the ultrasonic frequencies relatively quickly with 22kHz and 24kHz points to make sure the tweeter doesn't receive much energy up there (still no evidence for need of super-tweeters after all these years).

Here are the measured (uncorrected) and simulated (corrected) frequency response curves for all 5 main channels + LFE (at +10dB level):

As we can see, Audiolense has achieved much improved tonal consistency between the speakers. This degree of consistency in frequency response between speakers would be very difficult (if not impossible) to achieve without powerful DSP correction like this. You would literally need exactly matched speakers, perfect symmetry and very tight control over room acoustics!

In the time domain, here are the measured and simulated step responses across 75ms, 5.1 multichannel:

Note the large LFE curve, +10dB louder compared to other speakers.

Again, much improved time-domain performance with Audiolense. Not possible with just simple EQ.

As usual, I'll ZIP the .cfg files together with the .wav filter files to use in Roon with convolution DSP as presets:

While the Dolby Atmos Music Curve is obviously aimed at multichannel reproduction, there's certainly no reason why it cannot be used in 2-channel stereo applications. For example, with that general shape, we can create a similar target curve in Acourate using its design facilities:

Notice the application of pre-filtering in the low frequencies already reflected in the flatness of the measured response below 300Hz in my example, as discussed recently.

The rest of the Acourate filter-creation process is discussed here. And like with Audiolense, though not shown, there will be much improved frequency and time-domain performance.

So how does it sound? I guess it sounds as Dolby intended. 😉

The curve with its +1dB "hump" in the bass, sloping into 10kHz (-4dB) and steeper decline into 20kHz (-7dB) sounds reasonable to me. Obviously nothing too dramatic about this kind of target although some who are used to a flatter treble response might subjectively want more sparkle in the "presence" and "brilliance" frequencies above 5kHz. For "bright" sounding recordings, this can tone down some harshness. For example, over the last week, I was listening to older Sheena Easton albums like Best Kept Secret (1983, DR10) with songs like "Telefone (Long Distance Love Affair)" and "Almost Over You" which sounded significantly better with that high-frequency EQ attenuation. As discussed in the past, exploring high-frequency "tilt" might be worth considering.

I've also seen some studio pros express a wish for greater accentuation of the bass to +3dB (instead of just +1dB); for example, we can see that idea in this discussion where it looks like the guy added something like +6dB or more into the bass! Hmmm, if one mixes with that kind of a bump in the studio playback, wouldn't bass sound thin with standard calibration in other studios or consumer playback for those of us who more closely follow the Atmos standard?

BTW, I believe Sonarworks is quite popular in the pro world. I noticed the slight difference with Sonarworks' Dolby Atmos Music target EQ in the treble where it looks like it hits -6dB by 10kHz and then just a little more to -7dB into 20kHz. Notable, but not a big deal. On the sub-bass end, they roll-off to -6dB at 20Hz.

This target curve is a guideline, and Dolby states that the playback can be somewhat flexible depending on room size and equipment, so by all means, I believe one should experiment and adjust to taste (trying various room curves is nothing new here, as discussed in previous posts like this). Of course, there is still something to be said about excess deviation from the "standard" and not directly translating the sound from a studio that is calibrated "by the book" to Dolby's recommendations.

Being your own home tonmeister... 🤔

To end, in many ways, I think the audiophile who has studied the performance of his/her room, carefully assembled/optimized the sound system, who has learned to be aware of his/her own listening abilities and preferences, plus has curated the music he/she enjoys (the various domains of sound quality) has very much learned to be his/her own "tonmeister"; the "sound guy/gal", the "sound master" who considers the variables and achieves a kind of mastery over sonic qualities.

As with any hobby, there are things I hope we've learned and skills acquired over time. If a friend asks us:

"Hey friend, you've been an audiophile for years, tell me what you've learned? What are you better at now than before?"

What would be your answer?

Personally, I think it would be sad if all an audiophile could say after years of being in the hobby is that we've learned to name some popular or even esoteric companies to buy from. Learned that cables make a "big" difference supposedly. That $100,000 speakers sound better than a mere $10,000 pair because it costs more. That there are "quantum" technologies in the form of PWB foil stickers that make sound better, or that Synergistic Research fuses somehow are beneficial. That vacuum tubes and vinyl LPs are the pinnacle of sound quality cuz they have that "analog sound" (how dare they corrupt with DSD256!?). That in an audiophile system, we don't dare use "noisy" computers, or USB ports, or damage the pristine sound with DSP! That blind listening tests "don't work". That R2R and FPGA tech is always better than mere sigma-delta chip DACs just because (even though many of these same audiophiles think that DSD/SACD is great). That there is a special class of audio reviewer with "golden ears" out there who can point us straight and worth paying attention to because they know better.

No friends, it must not be like that! Those ideas are obviously a few of the insanity-tinged myths that can drag us down into the "rabbit hole" of madness. If there is a curse to being an audiophile, it is this potential of losing one's sanity because of so much misinformation and advertised promises out there. Instead of being empowered to be potent enough to achieve mastery, I see many audiophiles giving into the claims of others who simply want you to buy something - directly or indirectly satisfying their financial interests.

"Soundmen will be trained in music, acoustics, physics, mechanics and related fields to a degree enabling them to control and improve the sonority of recordings, radio broadcasts and sound films." -- Arnold Schoenberg (1946)

Rather, I hope as audiophiles, let's remain rational and respectable. We can learn the technical skills of running our own measurements (at least appreciate the power of doing so and how important independent verification is), advance our listening skills to tell whether claimed differences are actually meaningful (doing a few blind tests tend to instill wisdom and humility), and we can then apply that understanding towards improving the sound in our rooms, choosing the devices we want based on measurable qualities, and of course speaking openly and with some authority against the gross misinformation so pervasive in this hobby. The replacement of simplistic myths and fairy tales with truthful understanding based on concrete evidence has always been "the good fight"; a fight on many fronts these days, whether it might be to look beyond myths and prejudices in geopolitics, or our microcosm of mythical audio fantasies. Claims of "that's what I believe" or "that's what I hear" without evidence or even plausible explanation, often serving to fuel neurotic discontentment, are a dime a dozen. Be a knowledgeable, intelligent, capable, proactive, and effective, tonmeister.

A side quote from that link on myths and politics worth noting here:

"The sheer volume and intensity of popular culture, and its cannibalisation of traditional history and myth, has created a kind of Dreamland, where very limited personal knowledge and limited hard information is overwhelmed by a mass of popular culture stereotypes, distortions and contradictions. Now this isn’t another complaint about “disinformation:” the issue is much more fundamental than that. Our culture, including our political culture, no longer knows how to distinguish between (at least approximate) fact on one hand, and sheer invention on the other, because the two have become inextricably linked and confused, and each feeds off the other."

Also:

"Likewise, myths draw their strength from the need for them in the first place. No-one is convinced of the validity of a myth by patient investigation. Rather, the validity of the myth is taken for granted, and events are fitted into it, with more or less difficulty, as they occur."

Sounds like a lot of the typical audiophile culture doesn't it? The magazines, YouTube media, reviews, advertising claims, produce little actual "hard knowledge" most of the time. Instead, they feed off each other with allegations and opinions, to create a "Dreamland" of what constitutes good sound quality or the benefits of "High-End" audio. The myth of the highly audible digital "jitter" is a great example of one that just won't die despite "patient investigations" and evidence in the last decade from multiple sources. There seems to be almost a need within audiophile media to find fault with digital audio; and the dreaded jitter fits within that narrative.

Thankfully, we can find knowledge in the form of objective measurements and blind tests to get ourselves out of that Dreamland orbit. I highly recommend trying.

This week, I've also been digging into Townes Van Zandt's 4-CD box set Texas Troubadour (2002, DR12-13 average). Nice compilation if you're interested in Americana/folk/country! Tragic life story of mental health issues and substance abuse, but beautiful lyrics, touching music...

Hope you're progressing well down the audiophile path, and most importantly, enjoying the music, dear audiophiles.