2022’s Finland EY Entrepreneur of the Year Douglas Castro on his passion for music and how he turned that into two successful businesses.
Douglas Castro had always wanted to mix music with mechanics. An avid metal musician for more than two decades, Castro attended college to study electronic engineering with the sole idea of combining his two passions into a successful business.
“I love music and my plan was always to make my own musical equipment,” Castro tells SiliconRepublic.com. “Basically, just the idea of combining these two passions was super appealing to be since I was 16.”
That’s how he ended up creating Darkglass Electronics in 2008, a hardware company specialising in tools such as pedals and amplifiers for bass guitars. The company is headquartered in Helsinki, Finland.
“[Darkglass] was kind of like all the work I did in Uni (university) when I was studying electronics.” He converted his thesis project and college work into products for his new company, he says.
“I realised that there was no great company for bass players.” Most players in the industry focused on guitars, leaving the bass as an afterthought, he explains.
Though he didn’t just want to make good equipment for bass guitars – he was striving to be the best. And according to him, the nicheness of the industry made that goal quite achievable.
“It was a very solid niche…like being the tallest building in a very small town,” he tells me.
Castro leveraged his relationships in the industry to create collaborations for Darkglass – counting musicians from leading groups such as Cannibal Corpse, Periphery, Foo Fighters and even artists who work for Taylor Swift among the company’s featured artists.
However, in 2022, he sold Darkglass to Korg USA, the US subsidiary of the Japanese musical instrument manufacturing giant Korg. This was because, by then, his second venture Neural DSP was already making headlines, with Castro being named Finland’s EY Entrepreneur of the Year that year.
Castro founded Neural DSP (short for Digital Signal Processing) in 2017 alongside his friend and fellow musician Francisco Cresp. The second company represented a “very natural transition from Darkglass”, Castro tells me, this time targeting a much bigger market, encompassing both guitar and bass players.
Neural specialises in creating a software and hardware ecosystem that aims to give musicians the flexibility of creating the sounds they want using portable equipment at relatively affordable prices.
“There were a few big gaps in the software side of products for musicians, Castro says. “We realised there weren’t really any good sounding guitar emulators.”
“The thing you plug your instrument to is as important to your sound as the instrument itself. Actually, probably even more, arguably”. Tools such as plugins allow musicians to create different tones without making considerable investment in instruments, which are often much more expensive to purchase.
Castro explains that Neural’s popularity exploded in 2018, with revenue at the new start-up jumping from a few thousand euros to tens of thousands of euros a month. And for a company that’s less than a year old, that’s a good outcome.
This relatively fast-found success made him realise that Neural could be “something huge, much bigger than Darkglass”.
A business for and by creatives
Similar to Darkglass, Neural DSP collaborates with artists and groups, including Tim Henson, Petrucci and Plini.
For Castro, the idea is to build tools that the artists will want to use instead of what they currently use. Their new ‘signature’ sound then acts as both advertising for the company, as well as earns them royalties.
This works differently for different artists. For some, the company needs to wade through a dozen managers and intermediary parties, while for some, they can hash things out over a beer, he explains.
“And I think that’s why each product turns out being quite unique and different from each other…We really try to capture the essence, not just in tone, but also in personality and aesthetics of every artist.
“And I think that’s something that resonates really well with the fans of each artist and why the ‘Archetype’ [plugin] series is quite successful.”
Neural DSP is a heavy employer of machine learning, using the tech to create digital amplifiers that would otherwise take much longer with just manual efforts.
“Traditionally, if you wanted to make a digital model of a guitar amplifier, you needed like one or more PhDs”. And if done manually, creating a single, good model of an amplifier would take around 12 to 16 weeks, Castro says.
That would not work for Neural, which is attempting to create dozens more tools in a shorter period of time. Besides time, there aren’t enough experts who specialise in this particular field, he explains.
“It’s really hard to find enough talent and we’re lucky enough to have one or two of the best in the world right off the get go,” he says.
“I did the math and I was like if we hired like every DSP engineer in the world, and we had the money, which we didn’t, it would still have taken years to do all of [what we did].”
So their idea was to automate the whole process by designing a tool that analyses an amplifier and records the signals as training data. After about 12 hours of training, the company then has a digital amplifier model that “made itself basically”.
For this niche task, the company hired key talent from top universities from Finland, who were exploring this field in their research. “We found a couple of engineers that were actually doing that funnily enough.”
Machine learning engineers are some of the most in-demand workers in the tech sector and that’s why Castro thinks that his staff is passionate about working with Neural DSP.
“If you’re a world class machine learning engineer you can make 10 times the money working at Nvidia, or pretty much any company now. So you need to have a really, really strong passion for music for you to…take [a] 90pc pay cut just to do this”.
Castro tells me that everyone working in Neural’s machine learning team is a musician. “So I think on the more competitive fields like AI and machine learning, the passion for music definitely makes us competitive – even if you know we don’t have Nvidia money to throw around.”
Over the years, Neural DSP products have garnered a cult-like following. And interestingly, as many parts of the music industry took a hit during the pandemic, stuck-at-home creatives were able to boost Neural sales by purchasing tech that allowed them to experiment with music right in the confines of their four walls.
While demand is not as high as during the pandemic, Neural has managed to almost triple is returns in recent years and expects to make between €40m to €45m in revenue this year.
Don’t miss out on the knowledge you need to succeed. Sign up for the Daily Brief, Silicon Republic’s digest of need-to-know sci-tech news.
