Photo Credit: Immo Wegmann
Following many ticked-off customer comments – and steadfast denials on the company’s end – D’Addario has acknowledged that generative AI pumped out the music featured in a demo clip showcasing its NYXL high-density electric strings.
D’Addario, having previously defended the no-AI position tooth and nail, just recently took to social media to confirm the machine assistance. As underscored by screen-capped Instagram comments, a number of musicians promptly identified and criticized that assistance soon after the company posted the video.
(The original’s been deleted, but one can still find said video – simply featuring a D’Addario employee showing the above-noted strings in action – in YouTube reactions.)
With a quick apology and a pledge to be more mindful in the future – a small gift-card giveaway in support of proper music and artistry wouldn’t have hurt, either – the matter might have fizzled out there.
Instead, though, D’Addario vehemently refuted the AI allegations on social media and by maintaining that the aforesaid employee had both written and performed the work in question.
“Again—this is not Suno or AI. This is multiple tracks layered, and heavily edited. We are a company for and made of artists and musicians, and do not use AI for our posts,” the business responded to one commenter who’d sarcastically inquired about the prompt at hand.
As D’Addario’s doubling down added fuel to the fire, it probably won’t come as too much of a surprise that tripling down caused the flames of controversy to grow taller yet. After wisely stepping away from Instagram for a moment – keep in mind that this is the condensed version of the lengthy episode – the company made the ill-advised choice to offer a 600-word defense of its existing stance.
Published as part of an image, the convoluted explanation, complete with an emphasis on the relevant employee’s having purportedly spent “hours fine-tuning his guitar performance with an excess of plug-ins to compress, auto-tune, EQ, and distort a DI guitar performance,” can be found in its entirety below.
With that, the evolution from PR hiccup to unmitigated disaster was complete. And while this description might seem dramatic, a cursory glance at the appropriate social profiles indicates that customers are pretty far from putting the series of events in the rearview.
From there, the 600-word defense made way for an updated comment admitting to the use of Suno on the “original track.”
“We’ve completed a full review and confirmed that Suno Studio was used to regenerate the original track. We shared inaccurate information, and we’re truly sorry,” D’Addario wrote in part.
Things will, of course, cool off – and, if the internet’s short memory is any indication, probably at some point in the not-so-distant future. (That said, BTS diehards still haven’t let the Grammys off the hook.) But the incident is a good reminder of the music world’s continued disdain for AI notwithstanding certain individuals’ and companies’ embrace of the technology.
Here’s the statement in which D’Addario acknowledged the use of Suno:
“AN UPDATE ON THE STRING DEMO VIDEO
We got this wrong. We’ve completed a full review and confirmed that Suno Studio was used to regenerate the original track. We shared inaccurate information, and we’re truly sorry.
Music is made by people-their ears, hands, taste, and emotions. D’Addario exists to give artists the tools to shape their own sound, not to blur the line between human creativity and AI-generated content.
We do not support AI-generated music, and our process will reflect that going forward.
We’ll require our employees and creative partners to disclose any use of generative AI, and we’ll more closely review everything we put into the world.
We also got comment moderation wrong. Honest criticism should not be erased. We’re changing how we moderate our channels so feedback can be heard while people are protected from threats, harassment, and exposure of their private information.
We know we can’t fix this with another statement. D’Addario exists because musicians trust our products with their sound, and we’re committed to earning that same trust in how we show up for this community.”
And the company’s earlier remarks:
“We are a music company. Built by musicians for musicians.
So naturally, we share the growing alarm that generative AI can remove the musician from the creative process. In a recent post featuring one of our products, many people questioned whether we employed generative AI in our demo track. The track was written and performed by one of our employees who was proud to create an original piece to show off a new string he helped to develop.
In creating this demo, we now understand what our employee did and how complex it was. He spent hours fine-tuning his guitar performance with an excess of plug-ins to compress, auto-tune, EQ, and distort a DI guitar performance. He also programmed drums and 12 synthesizers for accompaniment. He ran a Logic ‘So-Cal’ stock drumset through 4 compressor plug-ins blended with EZ drummer with all velocity set to the maximum of 127 to get a big sound. This subtracted all the dynamics and added a lot of digital noise. The result is a track that exhibits the hallmarks commonly found in generative AI music, and it is understandable that this sound would elicit such a response from the community.
Some of you will immediately understand what we’re talking about. But for those who may be scratching their head, there were excessive non-AI tools used to make the original composition sound ‘more produced.’ This added a lot of extra noise and inorganic details to the mix which people misunderstood as signs of AI-generation.
So, where did we go wrong?
The video we posted last week in response to the accusations that this was created by Suno featured a DAW clip which was a pre-mix and pre-master pile of stems that did not sound like the finished product, and we did not question it because much of our team is not familiar with audio production. You see, our original demo was recorded in a logic session set to 87 bpm with a 44.1K sample rate and exported as such.
This is where the confusion lies: while Generative AI was not used in the creation of this music, by virtue of using Logic’s ‘Mastering Assistant’ feature and LANDR’s Assisted Mastering, AI did assist with the mixing and mastering process post DAW export which caused noise and drift. Most of the digital artifacts that people online have taken issue with come from an unusual take on modern production techniques (4 compressors in series on the drums adding high end fuzz, auto-tune on the guitar for pitch accuracy adding aliasing, just to mention a few).
We’re experts in crafting the world’s best music accessories, not audio production. In the future, we will know what to ask and what degree of production is appropriate for our videos.
We’d also like to make a bit of a personal statement. We knew we had nothing to hide; we just needed the time to understand why our take on the situation was so different than others in the community before we responded further. We’re genuinely sorry for the delay in response and (the) confusion it caused.
Remember, we are a company whose products can only be used by people. If AI replaces musicians making music, then we’re out of business. Of course, that’s nothing compared to the chilling thought of musical expression being co-opted by machines. As a company made of hundreds of people, we desperately want to preserve human expression through music.
With respect,
The D’Addario Team”
Leave a Reply