Hmmm...Does Atmos Mess with EQ, Not Just Space?

[Anderton]

These days, I release Atmos Binaural and conventional stereo mixes on my YouTube channel. I do the Atmos mix first, create a copy of the song, and use the Dolby Renderer to downmix it to stereo. It's good, but not perfect. Sometimes there will be some slight level variations. I want the stereo mix to sound as good as possible, but I also want it to sound as much like the Atmos binaural mix as possible. So I compare the two, and tweak the stereo mix to match track levels as closely as possible.

 

After doing so, I noticed that even though the levels were right, and the imaging was as close as possible to what Atmos does, something was missing aside from the stereo mix not being able to use space the same way. It sounded like there were EQ issues. So, I used iZotope's Tonal Balance Control to run a couple curves.

 

Check it out. The top one is Atmos Binaural. The bottom one is the rendered stereo, with some tweaks to match levels of instruments and such. Note that Atmos boosts the bass a bit, dips around 500-600 Hz, and gives a little lift around 2 to 3 kHz. As the bottom graph shows, those changes were not in the stereo render.

 

 

So, I created a compensation EQ curve. Note the scale is +/- 12 dB, so the changes are not as drastic as the might appear at first glance.

 

 

After applying it to the stereo mix...well, looky here. It looks much more like the curve on the Atmos Binaural render, and it sounds more like it, too. 

 

 

So, is Atmos hyping the EQ to make a "better" listening experience? I don't know, I'm just asking questions  All I can say is that with those EQ changes, the stereo mix came much closer to achieving parity with the Atmos binaural one.

 

 

 

 

[Dave Bryce]

Re: Atmos Binaural, I’ve made this observation myself.

 

It’s especially noticeable for me listening to Atmos remixes and the respective stereo fold downs of older material on Apple Music, then listening to original stereo mixes and switching back and forth to and from Spatial processing.

 

When listening to actual Atmos mixes on a multiple speaker setup, I’m not sure the hype thing is an issue. Not so easy to A/B, though.

 

dB

[Anderton]

20 hours ago, Dave Bryce said:

When listening to actual Atmos mixes on a multiple speaker setup, I’m not sure the hype thing is an issue. Not so easy to A/B, though.

 

True that. But I wouldn't be surprised if Atmos involves a little hyping of the frequency response in general. 

[Philbo King]

Binaural, as in HRTF? If so, I wouldn't be at all surprised there were EQ FX apparent. 

 

Pinnae and earholes are not necessarily FR flat.

[Anderton]

HRTF defines how sound waves are modified as they travel in space to the listener. The brain then is able to differentiate among the subtle spatial cues...so far, so good. 

 

But, if those sound waves are travelling from speakers to the listener's ears, aren't they being influenced by the same factors? Systems like Waves Nx and Slate's VSX have the goal of using binaural audio to emulate the sound of listening to speakers. So, if you're listening to speakers, wouldn't the sound be the same as if you were listening to emulated speakers on headphones?

 

My experience is that listening to the downmixed stereo over speakers does not have the same EQ changes as listening over headphones with binaural music. But, as Dave points out, it's difficult to do a direct comparison. That said, EQing the stereo with the same curve as the Atmos binaural produced a sound like Atmos when listening to stereo over speakers. That's not an exhaustive study by any means, but it was enough to pique my interest.