Guy here, who programmed the C++ implementation of Operator: It was a pleasure to build the instrument together with Robert, and I learned a ton from him.
In the 2009 upgrade I replaced the aliasing wavetables with bandlimited ones, generated using IFFT, one per octave. With 2x oversampling, it became aliasing-free as long as you didn’t use FM. When adding the IFFT, the feature of drawing harmonics also became obvious.
Fun fact: The four oscillators were calculated in parallel using SSE intrinsics. It’s the only time I’ve ever been able to improve the performance of something using that particular technology.
For me personally, Operator is a pinnacle of my engineering career - It is one of the most-used synthesizers in the world, though of course, there are much better ones out there.
Ableton is an epitome of software design innovation. I think few people understand how groundbreaking the Session View vs Arrangement View design was to advance the workflow of both produced and live music. Subcomponents like Operator created design patterns that are widely adopted by most VSTs today. Kudos to the Ableton team for crafting a product that is so beloved.
Session View (aka "clip launching") represents more than just a workflow - it represents an idea of what music is.
Some people might want to spend their time arguing over whether or how much or when that idea is correct. I'd prefer to note that in the 25 or so years since its debut, the idea has reshaped so much contemporary music whether it is correct or not.
The idea that you build up music out of sections that potentially repeat, then move on to another section ... it all seems so obvious to us techno-centric geeks. It actually isn't how most of the world has traditionally conceived of music, but the language that has been traditionally used has often had a structure that has made it easy to strip out the complexities, and the result is ... Ableton Live's Session View.
Yes, 22 years ago. And while they’ve expanded some plugin features, astonishingly little has changed/improved about the base Live since then. Even minor quality-of-life features like setting a default audio interface is not even on the roadmap for 22 years.
I think part of that is probably the clientele, musicians aren’t exactly famous for asking for more from their tools. Maybe I’m a grump for not thinking that “session view vs arrangement view” (which is really a bare minimum digital mimicry of the popular music making interfaces of the 20th century: the magnetic tape and a vinyl loop) is enough as the dominant paradigm for music making in a world now driven by global, recursive, abstract navigation of/ negotiation with interconnected, automated logic processes…
But I don't wanna be sarcastic.
The interesting thing about that is that no one in my (limited) experience uses the session view as it was intended to be used for music composition. After trying it for a bit, everyone seems to revert back to using the linear Arrangement view. Session view is still useful in some performance cases, but it makes me wonder if it would make sense to have it as an optional view, and not as the default view for all sessions.
Very interesting. When my buddies and I started playing in Ableton after years in Adobe Audition (circa 2010?), we immersed ourselves in Session view for writing and piecing everything together, building songs top-down instead of left-to-right. It completely changed our entire workflows, and that seemed like the point.
I'm only ever in arrangement view when I'm finalizing the order of the sections before final mix down.
I fell in love with the Session view when I started using Live in the early 2000s. But the thing is, even the most basic production nowadays uses a ton of automation, variations, break downs, single shots, etc. Session view just doesn't work for that. It's nice to be able try out various combinations in the Session view but I switch to building bigger structures almost instantly.
I found it faster to just record clips directly into the Arrangement view and not waste time on moving between different views. Right from the beginning, basic things like chopping audio or e.g. removing the last kick in a bar are a pain to handle in the Session view (creating a copy without that last kick). And once I start touching automation - which is basically from the start - Session view becomes irrelevant.
It’s trivially easy to make arrangement view your default view. You simply hit tab to switch the view to arrangement and then overwrite your default template in the ‘File’ menu. Now you’re set.
I didn't mean "default" in the sense that it's the first view in a new session. It's one of the only two default views. It would be useful to me if I could open a different view on my second screen.
Same here, everyone I've introduced to Live has been completely stumped by session view and preferred linear arrangement. I don't understand why it's the default view other than hanging on to the idea that this software is first a live looping tool and DAW second.
I use the session view for vocal recording. Recording multiple takes to new clips in session view, then copied into the arranger, it lets me comp several takes without messing up the final arrangement, and is great for project organisation.
I produce music since the early 2000s and used to own a music studio. And to answer your question: not everyone, "no one in my (limited) experience" uses the Session view for composition.
I was so impressed by Bitwig since the beta. Just couldn't find anything major enough to make the switch from Live. Also, M4L is such an amazing ecosystem, I would miss it.
Any features in more recent Bitwig versions that stand out?
I’m not sure how recently they were added, but the Grid (modular environment) and the modulation sources are really powerful, particularly for adding organic/generative elements to a production. I’ve heard that Bitwig routing is another workflow enhancement, but haven’t used it too extensively.
It would be cool if Bitwig would introduce a scripting environment similar to Max for Live.
If Bitwig would add user-modules to the Grid, along with a scripting language, it would be perfect.
Where they both fall down is in less capability for the piano roll and midi for more complex compositions and styles (orchestration) and although Live has made some moves towards improving that recently, they're both still way behind in that area.
Totally agree regarding the piano roll. Same with the sample editor.
Saying that, DAWs just have different specialization. E.g. for a long time midi support in ProTools was extremely limited. Cakewalk already had an excellent piano roll in the late 90s.
I wish... Live is the only reason I have a windows partition anymore. I just can't quite get it to work in wine/emulation...
When I'm forced to upgrade to windows 11 I simply won't have Live anymore, because I won't be doing that, and so I haven't purchased the latest version and am starting to experiment with the alternatives. Makes me sad though, it'd be so great to have on Linux.
I use it in a VM with vfio pci pass through of a gpu. Works fine but I'll probably move to bitwig anyway, I don't respect the decision of not having Linux support (they even run Linux on their push devices).
Oh I love FM synths! I'm working on a customisable one in my spare time lately for the kids' school, as the music teacher was complaining that the students have been using all the same samples over and over. Feel free to have a peek! Desktop only. Source code is hopefully nice and clean too: https://chrishulbert.github.io/you-synth
Ah Operator. This synth is so deep. Not only is it a fantastic FM synth, but it does subtractive synthesis well too. Also, it really is impressive how the UI manages to fit all those parameters. I mostly use it for cool synth leads. Here's one of my favorite videos on Operator https://youtu.be/rfeY0_k1ctk?si=s68Lr033cHf34a4M by Robert Henke himself.
FM is one of the most "naturally digital" synthesis method to implement, it's trivial once you have an accumulator and sin table working. The simplest form (and arguably the easiest to sound musical), can be expressed with a one-liner formula:
If you're interested in the qualitative distinction between subtractive and additive synthesis this is a fun practical example of why you might use the latter that I like a lot (implemented in operator no less :) )
I always tell new Ableton users to make one practice track using only Operator for every sound. It’s very educational and surprisingly for the new users, effective.
somebody please make a plugin version of operator for both windows and mac, there are zero good options for fm synth plugins in 2024 except for sytrus on windows if they even still sell it. exacoustics GHOST is looking very promising though, just still in its teething stages
Dexed is good but it's primarily a DX7 including all the weird preset algorithm choices and that's not really necessary if you aren't using DX7 presets. It's a DX7 emulator, not really jsut an FM synth for its own sake like Operator is
there are numerous good FM modules inside VCV Rack (and Cardinal), which have the benefit (or drawback) that you get full control over both upstream modulation and downstream processing. And they exist for all platforms that Rack (& Cardinal) run on, not just "windows and mac".
I have been out of audio for a long time but this sounds crazy.
Have owned FS1R and DX11, made my own FM synths in reaktor but I would still rate FM7/FM8 the greatest of all FM synths. There has just never been a better interface to program FM. In the same way I would probably rate the FS1R as one of the all time worst FM synths and I do enjoy a synth that is not easy to program for the uniqueness.
Aalto is great. It's more of a west-coast style Buchla complex oscillator clone than a traditional FM synth with multiple operators, but complex oscillators have at least one carrier and one modulator by design.
I use very few software synthesizers but Aalto and Operator are two of my favorites.
Had the good luck of attending an Henke (Monolake) concert in large-ish and cubical space fully rigged with speaker arrays doing wave-field-synthesis which was beyond anything I heard before or since. Loud is one thing, gigantic is something else.
https://roberthenke.com/concerts/wfs.html
You should have stepped over to the TU Berlin. A few month's before that concert, there was a performance/recreation of a concert that took place in Koln cathedral of various Messiaen organ works, played back on the WFS system installed at the TU. Much, much bigger than Henke's setup at Tresor, and it was a truly remarkable experience.
About a decade ago had the luck attending in-person his field recording and granular synthesis workshop in Amsterdam. Such an inspiring and humble individual! Still coming back from time to time to his timeless album Monolake - Hongkong.
In the 2009 upgrade I replaced the aliasing wavetables with bandlimited ones, generated using IFFT, one per octave. With 2x oversampling, it became aliasing-free as long as you didn’t use FM. When adding the IFFT, the feature of drawing harmonics also became obvious.
Fun fact: The four oscillators were calculated in parallel using SSE intrinsics. It’s the only time I’ve ever been able to improve the performance of something using that particular technology.
For me personally, Operator is a pinnacle of my engineering career - It is one of the most-used synthesizers in the world, though of course, there are much better ones out there.