Propellerhead, makers of the unique Reason software studio, have come from nowhere to award-winning developer status in just eight years. We catch up with one of the key men behind this small but influential Swedish company.
The past decade or so has been a dynamic and productive time in the history of music software, and the pace of change has become ever faster as virtual instruments have gained in number and popularity. One of the first software instruments to achieve widespread acceptance, and even cult status, was Rebirth RB338, a virtual recreation of the classic Roland TB303 Bassline synth and TR808 drum machine — both almost equally sought-after and expensive at the time of the 1997 debut of Rebirth, which offered a way to get those trendy sounds without the hassle and expense. The company behind this prescient launch was a small Swedish outfit called Propellerhead Music Software, previously known only for their innovative Recycle sample-manipulation software. Rebirth paved the way for their rapid rise to fame in the hi-tech music world.
'The Props', as they're casually known to users of the active forum/message boards at the company's web site, originally numbered just two, as founder member Ernst Nathorst-Böös explained when we met with him at the 2002 Frankfurt Musikmesse. Ernst himself has a long history in the MI business, starting back in the 1980s.
"I started out in music equipment distribution, working for a company that's now called the Fitzpatrick Import Group. We imported a lot of stuff into Sweden and some into Scandinavia — Linn, Simmons, Emu, Synergy, Sequential, and so on. In 1986 I left that company and worked as a technical editor for two Swedish music magazines for a while. Then I started a documentation company [specialising in producing technical manuals, including the Steinberg Cubase documentation]. Then, in 1991, I met Marcus Zetterquist, who is one-third of the company. We worked on some projects together before we formed Propellerhead in 1994."
That same year, Recycle was launched, initially in conjunction with Steinberg, who marketed the software as the ideal companion to Cubase, as it offered a simple and elegant way of gaining flexible control over the tempo of audio loops. Around two years later, the third piece of the Propellerhead jigsaw fell into place when synth and software designer Peter Jubel, who had created the software for Clavia's Nord Lead synth, joined Ernst and Marcus. Rebirth RB338 was the result, and its success made possible the product that the Propellerhead team seems to have been born to create: 2000's Reason, a complete electronic music studio inside a computer that was received with strong critical and user acclaim and went on to win a 2002 MIPA (Musikmesse International Press Award).
How Do They Do That?
Propellerhead is still a small software house, with a staff just 16 strong, and its products are still very much guided and directed by Ernst, Marcus and Peter, as Ernst explains. "It's not a formalised process, but it usually works like this. The three of us talk about what we want to do with an object or a new product. After that, when we have a really basic infrastructure, we bring in everybody for brainstorming meetings. Then the three of us take all that information, what everybody has said, and try to make it fit together and make sense. Then it goes back and forth from there. It's still very much the three of us that decide the products, although individual sections of the product can be done by many different people."
Ernst himself is not a programmer, but uses his expertise in other areas. "I sometimes say it's like when you were a teenager and you want to form a band: there's someone who can play a guitar and there's someone who can sing, and then there's someone who can't do anything and he gets to play the bass! And that's me. I'm in the background, just making sure it all stays together, keeping the drummer under control! I'm taking care of the business side."
Design & Development
As long-term Reason users, we were keen to find out more about the background to the program, and Ernst is happy to oblige. "Actually, Reason is the product we wanted to do when we made Rebirth, and there were some prototypes of things that looked more like Reason than Rebirth. But we were only three people at the time, and there was no way we could pull it off. It's crazy that we later pulled Reason off with five or six developers, as we did. Also, the computer technology at the time of Rebirth wasn't really there... it was too early. Reason is a bit like Rebirth on steroids. It's the same concept, but everything is much bigger."
One of the most interesting and fruitful aspects of Reason (which resembles a rack of hardware instruments, for anyone who hasn't seen it), is the use of the 'analogue' CV and gate interconnection and control system, and of virtual cables. We wondered who came up with these ideas, which have proved so popular with users. "CV and gate I think we had all along, as a concept. But the cables was Marcus' idea — that's a typically Marcus thing. He came to me and said 'You know what we've got to do? We've got to do cables!' And Peter and I said, 'No way, that's so stupid!' The rationale for doing it was actually down to three things. The first is that it's a familiar way of working and people understand it; the cable can only work in one way. Another thing is that it actually solves a lot of design problems. What do you do if you need to route something to multiple destinations? You need a cable splitter. Or if you want to merge signals, you need a mixer. For every problem that appears, there's a natural solution. And the third, of course, is that it's just totally cool."
Ernst's mention of cable splitters begged a question for us: where are they? We've been wanting a CV/gate splitter ever since Reason came out. Ernst's answer gives a bit of hope: "It's lacking. We're aware of it. I can't talk about the future updates, but it's one of the things that we know is missing!"
From the start, Propellerhead's products have been released simultaneously on Mac and PC, and the development inside the company occurs at the same time, as Ernst explains. "It's done totally in parallel. We have this framework — the technical term is a 'class library' — which we've built ourselves. At the lowest level there are the operating system-specific things, like 'how to save a file under Windows', 'how to save a file under Mac OS X', and so on. Then on top of that we have 'how to save a file the Propellerhead way'. So someone who wants to save a file only needs to know about that. On top of that there's the actual application, whether it's Recycle or Reason. So we have people working on features for the class library and when they submit their work it's instantly available on all platforms. Most of the programmers have both Windows and Mac machines on their desktops, and the program compiles continuously on both platforms."
As the company has grown, issues not directly related to their products arise: "One of our biggest problems is how to keep releasing interesting products, of high quality, at a decent rate, and grow as a company. Because every time you hire a new developer — it doesn't matter how talented or experienced — it's going to take at least six months before he's really productive, because he's got to get into our code base; it's a monstrous project. So during the first time period, he's not productive and he's dragging down someone else, who has to teach him. It's a paradox, because every time we hire someone our productivity goes down in the short term. It's a process that's really hard to manage. We look for general experience with programming, experience working with large software projects, using C++, coding for at least three years..."
Anyone who uses Reason as well as other music software should have noticed how moderate the program is in its demands on the host computer — not always the case for virtual instruments. Ernst puts it down to having people in the company with the right background. "Most of the DSP stuff is done by Peter Jubel, and he's got a very extensive background, starting with analogue electronics. He built vocoders and things, and then moved over to digital electronics hardware. Basically, he's the guy who designed [the software for] the Clavia Nord Lead. Then he moved on to applications software. So that means we have an understanding of the whole process. It's all about knowing where to spend your resources and where you can cut corners. And if you understand analogue electronics, you have that understanding when it comes to actually designing DSP electronics. I think some of the companies coming in now started up in computers. They went directly for that, which means that some of them might lack that depth which we are fortunate enough to have."
Though projects are controlled and co-ordinated by Ernst, Marcus and Peter, it seems that everyone on the Propellerhead team is welcome to have an input. At the time of our interviews, Niels Larsen was the company's marketing manager/press liaison, but as a musician and sound programmer himself (he's provided factory presets for Reason), he was obviously in a position to make comments during the software's development.
"The Subtractor synth filter I had some influence on, because when I heard it before the last betas [of the original version of Reason], I thought it didn't really bite. After a little bit of conversation, Pelle [designer/programmer] went back to it. I think what had happened was that he'd been tuning it too much. It's like mixing. After a while you don't hear that you've added too much bass, or whatever. So then we started looking at the Roland Juno 106, how that behaved. It was just a manual adjustment to get the 'behaviour curve' modified. It took him probably a couple of hours.
"The other thing I had an influence on was the stereo in Malström [the new synth in Reason v2]. Malström is very flexible in its routing. You can route the oscillators to the two filters in several different ways, and the filters can be used in parallel or serially. If you use them serially, the point is 'what happens to the stereo image?', and of course it's no longer there: the stereo image has to be a combination of the two filters in parallel, and maybe of the oscillators. So what we did was put a little Spread knob on it, so that the patch programmer can decide whether to use stereo or not, and actually it works very well."
Knights Of The Graintable
Reason v2 and its new devices has been very well received by users. A common reaction amongst people getting their first look at the new Malström 'graintable' synth, where sounds are based on one of 82 preset graintables (or indeed, any single grain from one of the preset tables!), has been to wonder whether they can make their own graintables. We asked Propellerhead if there were any plans to allow users to do this, or to add more factory graintables to the current collection. Niels Larsen told us that "The process of creating graintables is too involved and academic for us to be able to create a sensible tool for users. So, sorry, no. We talked about the possibility of adding graintables in the future, but we currently feel that we tried all the material we could think of and selected what worked. At the moment we think it's probably best to leave it where it is — but who knows?"
We also noticed that (uniquely amongst Reason synths and samplers) the powerful Malström has a polyphony limit of just 16 notes, in contrast to the 99-note potential of other devices. Niels comments that "Malström is a more CPU-intensive device than, for instance, Subtractor, so every voice counts. By reducing the polyphony we make sure that even a sloppy synth programmer will not bring Reason 2.0 to its knees. It goes hand in hand with our philosophy of optimising the product so it works on even modest systems. Imagine a batch of Malström patches with a polyphony of 50 and a long release time. You wouldn't need to play too many chords before the computer couldn't handle it any more. Sixteen notes is the sensible setting for this device."
The Future For Reason
The v2 update for Reason was pretty eagerly anticipated by what is a lively and vocal user base, some of which spent time in the months preceding the release posting their thoughts on what it should include at the Propellerhead web-site user forum. Amongst the smaller items on various wish-lists, there are three major recurring requests common to many users: audio recording; MIDI output, so that Reason's sequencers could control external instruments; and VST plug-in compatibility. At the risk of being immensely boring to Propellerhead personnel, who are well aware of what users ask for, we felt we had to ask what the official position on these issues was!
First, on audio recording, Niels Larsen: "I think it's out of the question if we just said 'a little bit of audio'. That's the sort of thing that we just don't do. We have a lot of people saying: 'You could just put in a couple of tracks' or 'Why don't you do an ADAT-like device and call it Recorder?' A nice idea, but it's the wrong way to approach an issue like that. If you put an ADAT into the rack, it's like having a little 16-channel Mackie mixer and an ADAT connected to it. At what point do you become more frustrated and begin to say 'I want to do more than that, I want to do editing. Why don't I have editing?' That instantly becomes an issue. Then it becomes 'Oh, I need more mixing facilities...' And then Reason as a concept, as a product to make music, might change too much. That's why audio tracks at this time is really a no-no, because we feel we'd be ruining what we created rather than enhancing it."
Ernst adds his thoughts. "People ask us 'What is your hang-up with audio, what is your hang-up with VST plug-ins? Why don't you just do it?' And I don't think there is a hang-up. We'd like to think that Reason is not a 'me too' product. There were certainly soft synthesizers before it, but something that we did — we're not exactly sure what it was — took it to another level. If we ever do hard disk recording, I'd like to try to pull off the same trick again! I don't want to just do a 'me too' thing for adding hard disk tracks to Reason; even though it would probably be practical, it's not what we are about. It's not right for us. We'd just be an inferior Cubase, and why would we want to be that?"
Perhaps, rather than incorporating the feature into Reason, a Propellerhead application could be purpose-developed to run in tandem with Reason? Niels doesn't rule out the possibility: "I think that's an interesting option, particularly with ReWire. I could see us having products that were the best for each recording task, rather than having one product that tries to do everything. Another thing about audio recording is that we have to be realistic about what we are capable of. If your expertise is synthesis and you have to start writing an audio editor, it would probably take several years, and then you have to question whether it would really be, after all that, so much better than anybody else's. But maybe we will."
MIDI output is next on our list of annoying questions — as Niels observes: "Ah... another one of those questions that are asked a lot. The thing with MIDI Out is that people say 'Oh, it's really easy', but it's not, actually. If you include MIDI Out, you have a host of other things you need to take into account. For example, do you need SysEx handling? We don't have that because we only need to control our own devices. The other thing is all the other tools you'll need if you are going outside to an environment that is uncontrollable and unpredictable. It would be a key strategic change for us, but also it involves a host of complications other than just making an output. For example, you'd probably need to have transpose on every track, certain other playback controls, probably a MIDI mixer as well..."
Finally, there's the issue of VST plug-ins, which certain users would like, perhaps principally to provide access to different effects. "Of course, a lot of people would love us to do it," says Niels. "But one reason is that we don't want to become a product that is a VST host, and particularly for VST plug-ins that might not be as efficient as Reason. We take pride in making reliable products, and in making them efficient, so that the user can do a lot on just one computer. One single plug-in could bring the whole thing down! And we don't want to do that. What we'd rather do is keep making instruments that can satisfy our users' appetite. We favour ReWire [for accessing VST plug-ins], not because it's our technology, but because it has become what we wanted an interface to be. Now we have Sonar on board [with ReWire support], and Cakewalk have done a great implementation, I have to say. Also Emagic, Ableton Live, Cubase, Nuendo... So it's getting there, bit by bit."
Pushed on the company's plans for Reason in the future, Ernst is circumspect: "The only thing I can say is that we will try to keep the program as focused as it is. A lot of people would like us to go into all territories and do everything that the big companies do. We're not sure that's a good idea, at least not for us. We can't do everything, so better for us to concentrate on what we do well and try to establish links — like, for example, with ReWire — to other products that provide the pieces that we can't. The future will be a natural evolution of what we have today."
They're not called Propellerhead(s) for nothing, as a story Niels Larsen tells us illustrates: "Three of the programmers, Dan, Marcus and Magnus, had a band years ago where the rule of the band was that you could only be part of it if you were playing your own applications! It didn't really have much to do with music — they spent more time in the debugger than anything else! But a lot of Magnus's work was born at that stage, leading to where Malström is today."
A big issue for any software house these days is software piracy. Companies come up with various ways of combating it, Propellerhead's being one of the least intrusive to the legitimate user. How big a problem is piracy for them?
Niels: "It's hard to quantify. It's big, of course. But we're very pleased that so many people choose to do the right thing. We try also, rather than introducing copy protection that frustrates users and yet is instantly crackable, to make service and support easy for the people who have done the right thing and bought the program. We don't want to frustrate them with installation processes that are more convoluted than downloading a pirate version, if you know what I mean. It's almost a punishment with some software products to actually buy them. We really want our users to have an easy life. We try instead to add value to the process of registration. We have a very good registration database, and for our registered users we hope we will make our service much better.
"It's also about having the material that's included with the program. People will want the sounds. We have the Soundbank, that's 500MB, and the new version has the Orkester [24-bit orchestral samples] collection for the NNXT sampler. The sheer size of that will prevent some downloads. We get the occasional post on the board that says: "Oh, I've lost my Soundbank CD, can anyone do me a copy?" I usually write 'Official reply from Propellerhead: Just register your software, and we'll sort you out.' Of course, you never hear back. We can also check, based on the songs that are being posted, whether they've been made using a pirate copy. On our archive, there are no songs made with pirated programs."
"No one really knows what the software issue is all about", adds Ernst. "I can illustrate with a story. I have three children; the oldest is 16. When Napster arrived, he started downloading songs like crazy. And he got to find out about so many new bands that he never heard about, which also meant that he started buying CDs like a lunatic. Then Napster disappeared, and he stopped buying as many records as he used to, because where is his source now? There's the radio, but he's not interested in the music that's played on the radio. And MTV? No! I'm saying that I can understand what the record industry is saying, but I can also show examples of the opposite. I think it's the same thing with software piracy. It is a major problem for us, but no one actually knows the mechanisms behind it. No one knows how many of these people bought the program who wouldn't actually have been your customers otherwise."
It would have been difficult to predict, 10 years ago, just how fast and relentless the rise of software instruments would become once they reached the kind of critical mass that occurred a couple of years ago. Developers like Propellerhead must have reflected on the impact their products are having on the hardware instrument. Is hardware becoming obsolete? Niels sticks up for the traditional synth: "Just for generating sounds, maybe. But they have other advantages. When you have a hardware synth, it's much more intuitive. Even with the best controller surfaces today, there's still a lot of setting up that puts people off. But the main reason hardware synths should not become obsolete is for live work."
Ernst comments: "It's no secret that software sales are increasing and hardware penetration is going down. The balance is shifting really fast now. It seems like some of the big companies are almost in denial. They're facing some major changes. You still need hardware, but take one of these Oxygen8 controllers and hook it up to a Titanium Powerbook, and I don't think you can argue that it's a more powerful and fun solution than any normal 'workstation'.
"But If I take it on a world tour, on stage with Madonna, with a Powerbook, then I'm not so sure. There are still issues. We're still struggling with things like drivers and audio cards, and efficiency and latency, and all those things. We're still struggling with them, but we're getting there."