NAM for Gear Marketing: a Win-Win for Players and Brands?

More and more brands that sell traditional guitar gear are taking notice of Neural Amp Modeler (NAM), and it goes beyond implementing NAM support in their digital gear. NAM, along with TONE3000's community for sharing tones, is emerging as part of the marketing layer of the guitar gear world.

I wasn't terribly surprised to see the news in late May that Blackstar had launched verified captures since it seemed complimentary to their embrace of NAM and TONE3000's API for their Beam Mini. The June 29th announcement of retailer Andertons joining TONE3000 with a verified account was a little more unexpected by me, but I see it as evidence that TONE3000 is becoming more influential outside the world of NAM purists. With 1.05 million Youtube subscribers currently, Andertons are a rather large fish in the world of guitar gear content, making their joining noteworthy.

NAM works by using neural networks to learn to reproduce your gear's effect on a guitar signal. This creates a digital model that you can use later in a DAW plugin or a compatible modeler. Crucially, it can even run in web browsers. I think this makes NAM captures the newest frontier for marketing gear.

Terms like marketing, promotion, and advertising tend to have strongly negative connotations, but I think it's actually a win for guitarists and bassists if brands promote their gear with free NAM models because the promotion contributes to the democratization of tone and gear.

NAM as a buying aid

Even though we can order pedals and amps online these days with relatively little friction, we understandably want information before we commit to spending a lot of money. This can mean spending a decent amount of time looking for Youtube demo videos, online discussions, and reviews for gear. And while video can be a decent medium for showing off (someone else's) tone, all of those media I listed ultimately leave you dependent on other people to describe the feel of the gear. And that can leave a lot to be desired.

The ideal is to try all gear in-store before buying, but that isn't always practical.

NAM's proven, leading accuracy means that it's the closest thing to actually trying an amp or pedal yourself.

And at least at an anecdotal level, there is some evidence that guitarists are starting to think of NAM captures when making purchasing decisions. For example, here's an interaction I had in the comments for one of the tones I shared of my Marshall DSL5CR:

The gap filled by verified T3K profiles

TONE3000's community is full of dedicated musicians and gear enthusiasts who hold their work to a high standard. But it's hardly a world of uniform standards and practices. There are a surprising amount of choices that go into a capturing a piece of gear.

For full-rig captures, for example, are you trying to capture/reproduce what your ears hear in the room by adjusting mic choice and position, or are you using those mic variables creatively like you might when miking a cab for recording?

What settings do you make and how many different captures do you make?

There are countless choices that you need to make when capturing gear with NAM, and it's as much of an art as it is a science. Error-to-signal ratio (ESR) doesn't tell the full story even. Accuracy to what? It depends on your settings and choices.

Even with the more modularized approach of capturing heads and cabs (which are "modeled" with IRs and not .nam files) separately, there is still a lot of artistic discretion that goes into choosing which combinations of settings to capture.

I think all of this means that any NAM capture is inherently an interpretation rather than an objective representation of the gear being modeled. As a clear example, when I make models of my Marshall DSL5CR, I see myself as capturing the amp as I use it rather than providing a single, authoritative example of the Marshall sound. You have to trust that I have maintained my amp properly, and you need to accept my choices that influence the tone. There is clearly still room for Marshall to put their best foot forward as determined by their company.

If I were looking at buying a Blackstar amp, I would want to try both verified and community captures to get a full sense of the amp's character and personality before taking the plunge. Both sides are valuable.

What's the scale of reach here?

It's still early days, and I admittedly don't personally have data to support the idea that NAM is helping gear brands with ecommerce conversion (in fact, my experiments show that expectations should be pretty modest with smaller-scale affiliate content). However, the verified accounts of Darkglass, Andertons, and Blackstar have rapidly accumulated tens of thousands of downloads and hundreds of favorites as I am writing this article.

Also, I think these numbers are big enough that it's evidence that the relationship between modeling and physical gear isn't purely cannibalistic. Not every one of those 34,000+ downloads represents a user that would have bought something from Darkglass. But those people are potential buyers, and the ones that favorited the tones are potentially qualified buyers.

A web of tone?

Although T3K is the center of the online NAM universe right now, the ingredients exist for things to become more decentralized over time. NAM itself is open-source, and the T3K team also open-sourced their browser-based NAM player. This makes it feasible for amp and pedal makers, as well as retailers, to integrate nam into their own website's shopping and marketing experiences. My hope is that this powers a "web of tone" where a variety of loosely-connected sites and platforms can include interactive demos with NAM.

Can we avoid extractive surveillance capitalism?

Nobody likes ads. And although nobody likes their attention or time being auctioned off, an even darker side is the collection of data that goes into powering those auctions. This is one of the issues I started this site to look into (specifically as it applies to the lives and wallets of guitarists).

Eventually, TONE3000 will need to develop a business model that can actually make money. It isn't free for them to offer their training services for neural amp models, and there are numerous costs in running a platform like that at scale. We don't know what the future will look like here. It is possible that they will go in the direction of offering tools to allow creators to monetize captures on their site. That would be a good direction, I think. They will also have a lot of personalized data about API usage, favorites, and the creators you follow on there. At some point, this starts to become quite valuable in terms of customized gear recommendations and ads.

No matter what, TONE3000's core product relies on people being able to find tones! So it's not unreasonable that they might start building better recommendation systems to help people find tones outside of the latest feed and "trending" tones. For example, they could use their data to build a model that identifies clusters of similar users, and then recommend tones that are favorited by users in the same clusters to each other. This would likely be an excellent quality of life improvement to the platform (finding tones, rather than availability, is becoming the bottleneck), but it would also be a step towards personalized promotions that will make some people uncomfortable. Of course, this is all speculation, but businesses are incentivized to develop revenue streams when there are none.

But if the NAM models themselves are the ads and the site doesn't have a whole bunch of 3rd party ad/tracking stuff in it, then maybe we don't care? It would be pretty egregious to have to watch some 30 second McDonalds ad before every tone you try online, but I don't really see anyone (T3K, myself, others) going that way, and there are lots of alternatives.

Anyway, I'm not trying to cast any shade on TONE3000, as the team has done an incredible job. But when something is free, you do have to wonder about how you will end up being the product.

For what it's worth, I decided a while ago that I didn't want to sell the privacy of my audience for pennies on the dollar by running ads here at Tubes & Code, so affiliate links will be the main method of monetizing my interactive pedal demos (such as this article I published about two budget distortion pedals).

Conclusion

TONE3000, along with their community, has built up a great resource for guitarists to find digital models for free. In addition to enabling guitarists to make music with a wider variety of gear, it opens up a world of more informed shopping where we can try amps, pedals, and cabs digitally before buying them.

I think this is an exciting trend to keep an eye on, because guitarists get more choice and more competition for high-quality models available freely if more brands keep joining and releasing captures. If the ad is a free digital model of an amp, please send more ads my way!

For makers and sellers of gear, there is an obvious concern that guitarists and bassists might choose to totally forgo physical, analog gear if modeling continues its march to dominance. But there's a lot of exposure to be had on T3K, and my position is that brands have an opportunity to make you fall in love with the tone so you buy all the physical knobs and controls for it (aka the actual amp or pedal). I think the way to lose the biggest is to not play and miss out on chances to impress guitarists with your brand's tones. Even if modeling continues "winning out" over tube amps, we still need something to model, so traditional amp makers still very much have a place in the ecosystem as "tone curators." It's an evolution of the role they play today in the ecosystem though, not a total reinvention.

Car companies use racing games to help build loyalty and mindshare even though much of the player base might be decades away from ever buying said cars. NAM unlocks the same "interactive showroom" experience for guitar gear. I do think the various factors I've discussed here will push the market for new tube amps in a luxury direction over time.

NAM's open-source licensing provides the community with options in the event that parts of the ecosystem become too captured by corporate interests in a way that doesn't align with guitarists. Someone can always fork it or build their own platforms.

These days, it can be hard to impress people, and attention spans are saturated. But guitarists seem to collectively have a lot of enthusiasm for NAM and T3K. Sharing high-quality NAM captures seems like a cheap way to build goodwill!