NAL-NL3: What’s Changing in Hearing Aid Fitting—and Why It Matters

The NAL NL3 fitting algorithm is back in the news as GN Hearing becomes the first to roll it out, a process I expect will continue regardless of the more recent GN announcement. It’s only a matter of time before others do the same. With NL3 now in the wild, it’s the perfect time to break down what improvements it offers compared to the industry standard NL2. It was just over a year ago when I visited NAL in Sydney, a couple of weeks before NL3 was unveiled at AAA and Audiology Australia 2025. I met with Pádraig Kitterick, who was kind enough to go on camera in exchange for us holding it back until the public release. He did a great job, as you will see. Thanks also to Justin Zakus and Matt Cruteau, who fitted me at the time with a set of NL3 hearing aids to take home and try, making me one of the first people outside the studies to actually give the noise module a go. Stick around to the end for my impressions. Thank you for joining me.

Thank you for having me.

So this is really exciting, actually. Let’s talk about NL3, because in my mind it’s almost like a global standard fitting algorithm that’s known and used worldwide. What could you possibly do different?

So if we take a step back and think about NAL/NL2, it’s got lots of bells and whistles, lots of knobs you can turn, lots of parameters you can set, but it’s really just one formula that fits everybody.

Yeah, so meaning regardless of circumstances, NL2’s formula is running the same.

Exactly.

I mean, other than noise level, because gain is changing with the—

Exactly. So it changes the gain based on input level. So it’s a nonlinear formula, but actually it’s the same underpinning formula that’s working the whole time. What we realized and what we’ve come to the conclusion is that one size no longer fits all. So when we think about NL2, when we’re updating that, we’re saying that philosophy at the core of that—which is to maximize intelligibility of speech in quiet, but don’t make it too loud, so don’t exceed the loudness that a person with normal hearing might perceive—we’re retaining that for our Speech in Quiet formula for NL3.

What we realized is that there are needs out there where that philosophy breaks down. It no longer provides the best solution. And so we’re going to start introducing new formulae that will go alongside the Speech in Quiet formula, which we’ll call modules. We’ll keep adding to those modules over time.

So I guess to answer your question, there are two key things. One is we’re refining the core Speech in Quiet formula to address edge cases and make it easier to deliver that core philosophy. But the second is introducing something entirely new—where we change the philosophy to meet unmet needs that current hearing aid formulas don’t adequately address.

Okay, so when you talk about speech in quiet—like we’re here right now—NL2 and NL3 will look very similar. Some refinements, but basically similar in how nonlinear gain is applied.

Absolutely. We’re sticking with what works well and what has been widely validated. When you prescribe gain for Quiet with NL3, it will look quite similar, but there are important changes in certain cases.

When we asked clinicians around the world where NL2 runs into problems, we got very consistent feedback. One example is mixed losses—where the gain prescribed by NL2 can be too high to be tolerable. We’ve refined how we account for the conductive component and prescribe gain for the sensorineural component, resulting in a more achievable and acceptable gain profile.

So you’ll actually have a version for people with mixed losses?

Not exactly a separate version. When you input air and bone conduction thresholds, NL3 automatically applies a refined approach. It’s not radical, but it improves how additional gain is calculated. For example, NL2 often prescribed excessive gain at very low and very high frequencies, which was difficult to achieve and not well tolerated.

We also analyzed large datasets of real-world fittings and saw that clinicians were already adjusting away from NL2 targets. NL3 incorporates those real-world adjustments.

So you captured how clinicians actually adjust fittings and built that into NL3?

That’s exactly right. We’ve done similar work with high-frequency gain. Clinicians often reduced it, especially for new users. Instead of simply lowering gain, we developed a method to reduce it where it doesn’t meaningfully contribute to intelligibility—while maintaining performance and improving comfort.

Another key improvement is for reverse sloping losses, where NL2 often over-amplified low frequencies and under-amplified others. Again, we refined the formula based on real-world fitting data.

And this is all built into the software automatically?

Yes. Based on the audiogram, the system adjusts accordingly. And interestingly, clinicians around the world reported the same problem cases consistently.

That’s interesting—especially across languages.

Yes, and we’ve retained features like tonal language optimization because they’re highly valued globally.

Another key change is the use of more advanced computational methods. While NL2 was already based on strong science, we now have much more powerful tools—like reinforcement learning—to optimize intelligibility and loudness targets. These newer methods often produce results closer to what clinicians actually do in practice.

So this is almost like using clinician behavior as a training model?

Exactly. We’re using reinforcement learning to solve complex optimization problems, similar to how it’s used in other industries. We’re combining that with real-world fitting data to create more practical solutions.

Well, we’ve been talking about the core formula—what about modules?

Modules are where we apply a different fitting philosophy for specific use cases. For the first release, we focused on two key areas: people with minimal or no audiometric hearing loss, and listening in noisy environments.

We know many people struggle in noise despite having normal hearing. Research shows they can benefit from hearing aids, but traditional prescriptions like NL2 don’t provide meaningful guidance—often defaulting to zero gain.

So clinicians are left relying mostly on directionality?

Exactly—and that’s not enough. We’re addressing this by creating solutions that balance comfort, sound quality, and intelligibility, even for people without measurable hearing loss. It’s a trade-off, but one we can now manage more effectively.