Pro Audio and Personalized Sound with Reveal+

I introduced Audeze Reveal+as part of my LCD-2 Classic review, and Embody Immerse Virtual Studio with the LCD-XC review. Audeze's goal with the original Reveal platform is for audio mixed on an Audeze headphone to be sent to others with the same or another Audeze product and them having the same experience. As you can imagine, this is critical when collaborating with music artists on the international stage. Several established recording studios also provided Audeze feedback about how their final product was listened to significantly more on headphones than speakers, so the studios had to be aware of how the mix would sound on various headphones. Here too does Reveal comes in handy with various headphone profiles. It would thus be fair to say the goal was primarily the pro audio world. Where it also benefited audiophiles was Reveal coming with its own DSP presets for various popular PC source programs, such as Roon, and it soon became a whole other thing with the update to Reveal+ and Embody collaboration. Please read through the two pages linked above for background information on these programs, as well as how they operate as plug-ins with different prosumer and consumer-grade software tools. This page assumes you are somewhat familiar with the end-user operations of Reveal+ and focuses on how it got to that point in the first place. Note that the images used below are taken or adapted from the Reveal+ White Paper v1.2 provided by Embody unless specifically mentioned otherwise.


A typical studio mixing environment involves using various software plug-ins with the digital audio workstation (DAW) or playback software. There are other factors in how a uniform listening and workflow process is ensured, including the effect of the speakers, room treatment, and your individual ears—I stay away from speaker coverage whenever possible because you need to have a dedicated setup for them and room treatment is quite expensive. The last step about individual ears is why so many different standards are employed for even headphone and earphone measurements, with my particular setup complying to the IEC 60318-4 standard colloquially just referred to as IEC711. But suffice it to say that even if the first three steps are done, that last part is extremely hard to generalize beyond measurements.

The goal of Reveal+ is to simplify the chain by creating a headphone-based virtual chain, with steps 1–3 of the original chain handled by a custom plug-in that knows to look for specific headphones it is tuned for, which has the benefit of removing the physical studio environment altogether and replicating a studio environment with headphones. I spoke to several people in the industry who testified to how useful something like this either was or would have been during the quarantine and lockdown periods caused by COVID-19, as the term "bedroom producer" was actually a thing. Reveal+ and Embody's IVS models steps 1–3 through what is called studio measurements. Step 4 is separately catered to by modeling how the individual hears audio in the real world, and the new plug-in is then tuned accordingly.


There are a few different trains of thought on how to get room coloration out of the equation for studio monitoring, with some taking the software route, such as Sonarworks with SOUNDID Reference, although taking actual room measurements of your work environment is the only valid solution. Embody did so with a head and torso simulator (HATS) placed in a few different professional recording studios and the HATS dummy representing a generic human being. The room factor is incorporated by having audio sweeps from the mixing console and speakers in the studio captured by the HATS dummy placed in the mixing chair. This includes spectrogram measurements to gauge the room decay of these speakers, and binaural measurements from the HATS dummy. So while you won't have your own studio or room reflected here without your own measurements, provided everything else is accounted for, this has the added benefit of getting a feel of what these pro studio environments are like.


Incidentally, the HATS dummy in question here is the excellent Brüel & Kjær High-Frequency Head and Torso Simulator Type 5128, arguably the current gold standard in representing the human ear and more, and it costs a lot of money too. Too much for my wallet, but I appreciate Embody using it and adding more credibility to their workflow.


Here are two example frequency spectra of the binaural room impulse measurements taken on the B&K 5128 in two different studios, showing differences in the lower and higher frequencies. This relates to what is called the head related transfer function (HRTF), whereby different people, or in this case a simulated person, perceives sound differently from others, and in different ambient environments. As such, the generic HRTF from the HATS dummy isn't valid for your ears and needs to be replaced by your own individualized HRTF for the virtual studio signal chain to succeed.


Here too different solutions from different companies exist. We have covered some already, including SoundID from Sonarworks that generates a preferential EQ-based filter set after A/B testing at various frequencies. This is not going to qualify for professional use without both profiles for the headphones or earphones and SoundID Reference, and is it based on a discrete number of steps and EQ filters even so. Then there's Creative's Super X-Fi headphone holography that asks for a picture of your pinna and generates one of several discrete profiles based on their experience and collected user preferences data. Both cases are limiting, placing you into a user group with ideally the exact same HRTF. However, the HRTF is more likely just close enough given engineering constraints. Embody takes the latter route too, with the setup process for Reveal+ requiring a photo of your right ear taken from the front-facing camera of your phone. In an ideal world, everyone has in-ear measurements taken, as I had the luxury of doing for SXFI, but that is not practical. Where Embody aims to differentiate itself from others is by generating a unique HRTF for every single person based on said photo, with some of my alma mater colleagues responsible for developing the algorithms used to "run millions of nonlinear calculations in just a matter of seconds by harnessing the power of cloud computing and parallel processing" to generate your HRTF magnitude plot—an example is shown above.


I'll skip a couple of steps since I am not an audio engineer and discussing the replacement of the generic HATS HRTF with the personalized HRTF is beyond the scope of this article. The final version of the plug-in to headphones signal chain from earlier aims to have every single component here represented, which means compensating for the headphones being used, too. The EQ is probably where most audio engineers and others competing with Embody will have their own opinion, with some claiming the need of an EQ profile to flatten out the frequency response of headphones and others wanting things to be done on a headphone-by-headphone basis to generate profiles that simply correct tonality errors. Others in the middle, and I include Embody here, want to correct for the physical coupling between the headphones and individual person's ears and head. As such, Audeze first provides an EQ profile based on what it thinks is appropriate for the headphones to meet its own target, and Embody takes this profile and adds it to the chain seen above. If all works well, you now have a plug-in with your own HRTF profile, the specific headphones profile, and different recording studios for you to reproduce the chosen studio environment wherever you are using headphones. The various options can be decoupled too, so simply choose the HRTF profile and listen to music tuned for your ears far beyond any EQ filters.