BREAKING: Dirac Co-Founder Mathias Johansson Resigns, Accepts New Position with HARMAN

(September 10, 2024) In what may ultimately represent a seismic shift in the industry, Dirac's Chief Technology Officer and Co-Founder, Dr. Mathias Johansson, has announced his departure from the company he founded some 23 years ago.

Johansson shared the news in a LinkedIN post published less than an hour ago. With an official end date of September 16, 2024, his next venture is to lead HARMAN International’s Signal Processing & Applied Intelligence Lab.

"A new exciting chapter begins for me today," states Johansson. "Having devoted more than 23 years to the company I co-founded in 2001, I have decided to make my next move and spread my wings outside of Dirac. I am excited to share that on September 16, I will begin leading Harman International’s Signal Processing & Applied Intelligence Lab."

According to HARMAN International's CTO, Armin Prommersberger, "Mathias' extensive experience and visionary leadership in digital signal processing and human-machine interaction are ideally suited to our future technology direction, representing the essence of the tech culture we are building: an entrepreneurial mindset combined with extraordinary technical skills and a strong focus on collaboration."

Prommersberger says he's admired Johansson since their first meeting almost 20 years ago.

Johansson's departure from Dirac isn't acrimonious. In fact, he indicates that Dirac remains an innovative force within the market and that he'll assume an advisory role for the foreseeable future. "Founding a company is like having a baby. It's a 24/7 commitment that’s always on your mind, day and night," says Johannson. "I am proud of everything Dirac has achieved over the last 23 years, and I remain as optimistic and confident about Dirac’s future today as ever before. All the arrows are now pointing in the right direction for the company. To me, this means it is time to let this baby fly without me. Children grow up, and at some point, they need to tackle the world on their own, even if it’s painful to let go. To this end, I will remain an advisor to Dirac’s CEO for some time."

Johansson holds a Ph.D. in Philosophy and Signal Processing from Uppsala University and an advanced degree from the Stockholm School of Economics and Executive Education. During his time at Dirac, he's served as CEO (2009-2021), Chief Product Officer (2021-2023), and Chief Technology Officer (2023-2024).


Related Reading:

• Mathias Johannson of Dirac Joins AV NIRVANA Live, 11AM EST on Wednesday!
• Onkyo and Dirac Strengthen Ties, Tease What's to Come
• Dirac Live Bass Control Becomes Available for Denon and Marantz Receivers

 

[Sonnie Parker]

That's a pretty serious move... and likely a VERY good catch for Harman.

 

[Gerry Iaria]

That's a shock, but he must have had his reasons

 

[Todd Anderson]

Having spent some time chatting with Mathias over the years, he doesn’t strike me as someone who would spin an untruth in a situation like this. I think it's simply a career move to freshen things up. His LinkedIn post is pretty clear about his intentions. I guess it was time for a change!

Dirac is in good hands. Nothing seems to indicate otherwise.

 

[Todd Anderson]

^ Just to note, I’ve received inside confirmation: these assumptions are spot on.

 

[tjcinnamon]

great vision but atrocious quality control. Hope he’s not in charge of any product implementations

 

[JStewart]

great vision but atrocious quality control.

That may or may not be fair, but it strikes me that the vast majority of the issues come from use with different brand and model devices. I’d like to think Dirac provides specifications to the manufacturers that will work with their software and on the manufacture’s end things don’t always work as intended.
Beyond that there’s still been hiccups. I don’t find that surprising. How can one possibly test every windows or Mac quirk that can crop up? I see the same with other software, like intuit quickbooks for example, perhaps the most widely distributed business accounting software.
I’ve personally written simple code using tools from Microsoft that gets broken with one of their countless updates or worse, a new OS version. Very time consuming to figure out a cause and workaround for a nothing small business person as I was. I’m glad I wasn’t in software development. Some days I’m surprised this stuff works at all. Like, what was up with this forum yesterday? I created a post with photos and when I hit post reply they were gone. Pages loaded excruciatingly slowly, with missing data, or not at all. Glad it‘s working again today.

 

[Gerry Iaria]

That may or may not be fair, but it strikes me that the vast majority of the issues come from use with different brand and model devices. I’d like to think Dirac provides specifications to the manufacturers that will work with their software and on the manufacture’s end things don’t always work as intended.
Beyond that there’s still been hiccups. I don’t find that surprising. How can one possibly test every windows or Mac quirk that can crop up? I see the same with other software, like intuit quickbooks for example, perhaps the most widely distributed business accounting software.
I’ve personally written simple code using tools from Microsoft that gets broken with one of their countless updates or worse, a new OS version. Very time consuming to figure out a cause and workaround for a nothing small business person as I was. I’m glad I wasn’t in software development. Some days I’m surprised this stuff works at all. Like, what was up with this forum yesterday? I created a post with photos and when I hit post reply they were gone. Pages loaded excruciatingly slowly, with missing data, or not at all. Glad it‘s working again today.

Windows works in mysterious ways

 

[Sonnie Parker]

Dirac is not equal in all implementations, as the manufacturers control each implementation. We briefly discussed this with Gary Blouse (StormAudio) during our YouTube interview a few weeks ago. This is why DLBC sounded better on my MK3 processor than my HTP-1 processor. I also remember having the Audio Control Maestro X9 in here for review. I could not finish the review because of Audio Control issues with the Dirac implementation. AC even sent me a new unit, but it had the same issues. Several of the guys from Dirac worked with us relentlessly to resolve the issue, but ultimately, AC had to fix their implementation issues, not Dirac. That doesn't mean Dirac is completely free of issues from time to time, but outside of the AC issues I had with their X9, my countless interactions with Dirac's software have been very good from day one.

 

[Todd Anderson]

Sonnie is right. Dirac basically provides the playbook. It's up to each individual manufacturer to code that playbook into their own hardware.

In the case of StormAudio, they do all of their coding in-house and have a much more robust platform capable of running Dirac with more precision. My understanding is some – not all, but some – outsource the coding to a third party. In those cases, if the implementation is riddled with problems, the process to correct them is a lot more complicated.

So, I don't think it's necessarily a Dirac problem. I think it speaks more to the complexity of the software and the fact that manufacturers have difficulties properly implementing it on their various platforms.

 

[Gerry Iaria]

Agree

 

[tjcinnamon]

Sonnie is right. Dirac basically provides the playbook. It's up to each individual manufacturer to code that playbook into their own hardware.

In the case of StormAudio, they do all of their coding in-house and have a much more robust platform capable of running Dirac with more precision. My understanding is some – not all, but some – outsource the coding to a third party. In those cases, if the implementation is riddled with problems, the process to correct them is a lot more complicated.

So, I don't think it's necessarily a Dirac problem. I think it speaks more to the complexity of the software and the fact that manufacturers have difficulties properly implementing it on their various platforms.

This is the software before it even sends to the AVR. It’s had a pervasive issue with windows where it says that the measurement doesn’t have enough data.

They have a work around that works for some and not others with using the mic in exclusive mode but Umik’s use a boost and exclusive mode doesn’t like that.

With only a few speakers you can luck into a good measurement set. However, with a 9.1.4 system, it’s difficult.

I’ve been using Dirac since 2.0, like close to a decade. It’s been buggy since day 1 irrespective of the AVR. Totally the software side

 

[JStewart]

I’ve been using Dirac since 2.0, like close to a decade. It’s been buggy since day 1 irrespective of the AVR. Totally the software side

I’m going to respectfully disagree. I’ve also used Dirac since Ver 1 on a pc then on MiniDSP Nano AVR, Nad T758v3, and Monolith HTP 1. Never experienced anything remotely close to a problem with a measurement set. I‘ve followed the Dirac thread since the day Markus started it on AVS so I’ve seen the problems there as I know you have. The difference between me and those that have had problems using a Umik and HTP-1 is our computer. To me that rules out the software per se.

I just don’t think it’s reasonable to expect a developer to account for every bit of internal hardware, divers, conflicting software and widows insanity that happens. I do think they could be quicker about code development that traps and handles exceptions they are aware of. The “imprecise measurement “ problem comes to mind. The biggest frustration there was the user had no clue what caused it and how to rectify it. For experienced users we didn’t why it was added as a feature. Had our measurements always been imprecise and we just didn’t know it?

I would also criticize their customer support process. At one point it was necessary to set up a separate account with Dirac just to submit a ticket and it was not clear on their website as to where and how to get help. It should be as easy as possible for customers to communicate. Hopefully the process is streamlined now.

Here’s a thought. Dirac should up the price and include a mic and pre-configured laptop. Problem solved. New problem. Price.

 

[Sonnie Parker]

Yeah, if you are experiencing a lack of measurement data, something else is going on. I've had it in every stage for home and auto, and the only time I have ever experienced this (since day one) was with the UMIK-2 and my MK3, which I quickly ditched and returned to the UMIK-1. I've had equipment issues and Windows issues, but very few Dirac software issues that were not addressed fairly quickly via an update. I did have some other issues one time with some measurements, and I discovered my UMIK-1 bit the dust for some reason. Fortunately, I had another UMIK-1, and all was good. I have three UMIK-1 mics now... two I use for two different rooms and another for backup. I suspect I've run well over a thousand Dirac measurements... it's gotta be a crazy number.

 

[Todd Anderson]

... had a pervasive issue with windows where it says that the measurement doesn’t have enough data.

. I've had equipment issues and Windows issues...

Hmm... it sounds like you guys need to use a Mac.

 

[Sonnie Parker]

Ohhhh noooooo! lol

Go to the REW support forum and see the astronomical number of issues with Macs.

Google search "dirac live issues with mac"... they are probably more notorious for issues than Windows.

 

[Mike-48]

only time I have ever experienced this (since day one) was with the UMIK-2 and my MK3, which I quickly ditched and returned to the UMIK-1.

Sonnie: Interesting! You may remember, I won a UMIK-2 in an AV NIRVANA contest. I have experienced weird error messages in REW when using it, and now I wonder if the mic may be the issue. Have you ever got to the bottom of issues with the new mic model? Even to the point of understanding when it works best and when it causes trouble?

 

[Sonnie Parker]

Sonnie: Interesting! You may remember, I won a UMIK-2 in an AV NIRVANA contest. I have experienced weird error messages in REW when using it, and now I wonder if the mic may be the issue. Have you ever got to the bottom of issues with the new mic model? Even to the point of understanding when it works best and when it causes trouble?

I have not dug into it any further. When I kept having issues, I mentioned it to Matt at StormAudio, and he immediately asked me if I was using the UMIK-2. I had just gotten the mic, and he said don't use it. I don't remember why, but I'll ask him about it and see what he says. I don't know if it was the mic's fault or if it was a compatibility issue with the MK3.