Early 2025 round-up: More NAM adoption

I've gotten a bit behind the last few months covering all of the news with NAM, but 2025 is off to quite a start, with NAM showing up in a bunch of new products. I wanted to take a moment to recap the new additions to the list in an effort to keep up to date with everywhere that NAM has been adopted. This is really exciting to me!

Sonicake Pocket Master firmware v1.1 adds support for .nam files

The Sonicake Pocket Master [source]

The Sonicake Pocket Master is a palm-sized multi-effects unit that allows guitarists to build a signal chain with the standard guitar effects along with a variety of amp and cab models. Sonicake started with their own modeling approach, but added the ability to use neural amp models in place of their own amps (though the IR block isn't available at the same time, so NAMs will need to bring their own cabinet). At $65 as of this writing, it's very exciting to me to see NAM make it into such an affordable hardware unit.

It's amazing to me personally to see something like this get released and see NAM reach a true modern "budget" price point. In order to make this happen, Sonicake had to work with a very resource-constrained device chip. They ended up introducing a "conversion" step where they use a proprietary model that attempts to match the given .nam as closely as possible. In this sense, users can think of it as a "lossy" conversion that their models undergo in order to be played on-device. I'd like to sit down and make a video to demonstrate this in more detail but haven't had the time yet to do it in the level of thoroughness I'd like, but that's on my to-do list.

AnalogX Genesis uses NAM to model Studio Gear

AnalogX Genesis [source]

The AnalogX Genesis (free trial here) is a effect plugin focusing on outboard studio gear (compressors, limiters, saturation, etc). They've chosen to make their models using NAM, which is really exciting to me. Their website details all of the gear they've modeled, and their plugin allows loading a chain of models (including users' own NAMs).

Personally, I'm excited to see people pick up NAM for other applications beyond guitar--while guitar amplifiers were what I had in mind for myself as a guitarist, the technology is quite a bit more versatile than what I think most people expect (possibly because their experience has been influenced by companies' products that either can't be tuned appropriately for this use case), but this is a cool demonstration of what's possible in an application that's a bit outside where my own personal interests started, and I'm excited to get more into it.

GPU Audio releases their SDK, featuring a GPU-powered NAM demo

GPU Audio develop and sell a software development kit that enables builders to make use of GPU-like accelerators in audio plugins. Last year at ADC'24, GPU Audio led a workshop on the use of their SDK to implement a plugin that plays standard NAMs on the GPU, allowing them to rack up impressive numbers of instances of the plugin running at the same time.

Recently, GPU Audio have made their SDK available on GitHub for free use in demonstration/trial contexts (distributing a plugin is not covered by the free license). I'm honored that that GPU Audio have chosen to demonstrate their product using NAM as it attests to the opportunities they (and I) see for NAM to support new and exciting products in the digital audio space.

Sonulab StompStation is released as an affordable NAM player

Sonulab StompStation [source]

Sonulab have released the StompStation, a pedal that brings an affordable way to get NAM in a compact pedal format. Sonulab's approach involves a (lossy) conversion to a proprietary model architecture developed to run efficiently on the ESP32-S3 chip at the heart of the StompStation. Editing and model loading can be done via a desktop app, and the pedal can connect wirelessly.

Folks can find demos of the pedal on YouTube, and the pedal is now in stock and can be ordered from Sonulab, and I'm excited to try it out for myself.

Hotone's Ampero II Line adds support for NAM with new firmware

Hotone Ampero II [source]

Hotone recently released a firmware update for their Ampero II line of products. What's exciting to me is that they've implemented a NAM integration that allows users of their pedals to play (converted) NAMs on their products. From the perspective of an Ampero user, this should be super-exciting: your gear just got access to a whole ecosystem of models that you can confidently invest in since the underlying tech is open-source and therefore perfectly future-proof. I'm really excited to see how this unfolds as the benefit of going with NAM as an open-source standard benefits musicians long-term.

Valeton's GP-200 Line Firmware v1.7 adds .nam support

Valeton GP-200 [source]

The next day, Valeton introduced an update for the GP-200 their line of products with Firmware v1.7 that allows users to import NAMs and play them as a converted "Snaptone" model. Again, it's really exciting for me to see this become an option for users who are using these products already, and I hope that they see this as a big value-add for them.

Conclusion

...and that's all for now! Like I said earlier this year, 2025 is looking to be a very bright year for NAM. I'm really excited that there are all of these new ways for folks to be able to take NAM as a cross-platform standard and use it in their hardware and products of choice.

A word on "conversion"

I'd be remiss not to comment on what folks have coined the "conversion" approach to playing models, where the result doesn't process the sound in the same way as the original NAM model. NAM is known to be the most accurate data-driven modeling technology in this domain, and this is valuable to many people who have found that they value the difference that NAM brings.

As I've hinted at with my word choice above, I think that a good analogy is with audio files, where one has "lossless" options like WAVE files and lossy formats like MP3. It's understood that lossy conversions will lose something in the process; the goal is to minimize this and make it be worth it (not unlike how NAM aims to minimize how much it loses from the source it models!)

Clearly, there's a trade-off: if you want to play NAMs in a "lossless" way, there are fewer options. However, I don't think that more choices detract from products that have chosen to go the path of playing NAMs by using the open-source NeuralAmpModelerCore in their software, or by some other lossless conversion. I do believe that more choices is better, and I'm glad that NAM can be a catalyst for what I think is a net-positive outcome where more musicans have more options to make their music. These are steps in the right direction, and I'm excited for what's still to come.