Korg also makes their digital and analog ecosystem fairly open at the lower end; analog devices are designed to be easily hackable and they facilitate sideloading of DSP applications; code that runs on this device can also be loaded onto several of their consumer models: https://www.korg.com/us/products/dj/nts_1/
This has a lot to do with why they've been eating Roland's lunch for the last several years, while that latter company keeps trying to convince buyers that completely locked down circuit simulation of classic devices at fancy prices is what the people want.
At the cheaper end of the scale, companies like Sonicware are exploiting commodity-priced microcontrollers for their 12-bit synthesis and processing capabilities. While their own offerings are not open-source, abundant resources exist for anyone interested in processing audio on crunchy chips and you can get into it for about $20 in hardware costs (years of your life not refundable; synthesizers are nerd crack).
Roland is doing just fine selling digital emulations of their classic gear. But they aren't exactly amazing innovative instruments. I look at Roland today as the Gen-X Harley-Davidson: Selling nostalgia gear to aging dudes hitting their mid-life crisis.
Korg has some missteps (the Electribe 2 could be so much better), but seems to still want to make unique, interesting instruments. I'm disappointed in the build quality of the wavestate/modwave/opsix, but their other gear is top notch.
> I look at Roland today as the Gen-X Harley-Davidson: Selling nostalgia gear to aging dudes hitting their mid-life crisis.
So true, and it huuurts. :D
However, one very valid selling point for gear (old or new) is that it is still, by far, a better user/musical experience than software emulation, even with multiple innovative controllers. Either because a screen lack multiple physical points of control, or because software/plugins rights managements are _always_ getting in the way.
Of course, only if you do something with it and it doesn't pile up in a too small room.
What’s an innovative instrument and does that matter when making music? There’s quite a fetishization of gear happening nowadays, making and buying gear for gear’s sake.
By now one can see that subtractive is king, because it’s easy to understand, sounds great and one can hear it everywhere. And subtractive’s pretty much done, manufacturers are just tinkering with filters, modulation and effects at this point. Popular VSTs nowadays are mixing several types of engines (Pigments) or expanding the set of oscillator waves (Serum). In HW we see more of the same or subtractives.
Subtractive will never get old and Roland classics like the Juno, Jupiter or SH-101 just sound good. The Roland drum machines sound good. I never owned any classics and don’t revere that “Roland sound”, but at the end of the day they sound good, so they’re good - and more importantly they’re enough.
All the innovation happening in sound synthesis seems like over engineering and developing things for the sake of novelty.
An innovative instrument is one that's trying to sound like 2022 and not 1982.
I love the hardware Roland sound. Especially the Juno 60 which is one of the sweetest sounding synths ever made. The Jupiters and Junos are musical in a way that other synths never quite matched.
But the Roland Cloud and Zencore versions are close, but... eh. And Roland Cloud Manager is an absolute clusterfuck of a product. It seems like Roland are all about the money now, and they don't even like their customers all that much.
Meanwhile... I bought ModWave Native yesterday and it's the most fun I've had with a synth for a long time. It sounds nothing like a 1982 throwback. The wonderfully bonkers gravity-well based KAOSS modulation makes the sounds organic and unpredictable, far away from any traditional ADSR/LFO combination.
I don't think it's gratuitous or over engineered at all. It's in a nice spot where there's some interesting added complexity on top of a recognisable wavetable engine.
It also works much better as a VST than as hardware because there's more polyphony, it's easier to edit, and it doesn't take up any space.
There’s nothing inherent to the Juno 60 (to take something with the simplest architecture) that makes it sound like the 80s vs 2022. For electronic music it’s the drum beats, melodies, arrangement, the modulation, filter use and effects that have changed with the times.
Wavetable’s popular right now, so credit due to Korg for putting that out in HW format. On the other hand there’s also Waldorf which built their company over decades around this type of synthesis both in HW and plug-in format. ;-)
The thing with wavetable is that it works much better as a VST. One look at the front panel of the modwave lets me know that fiddling with the dozens of buttons and knobs would not be fun for me.
Anyway, there’s hardly anything revolutionary happening in synths. My theory’s that Roland recognized that and are focusing on workflow, availability, interoperability, etc.
Personally I don’t get it why there’s so many new synths being built…
Korg is really amazing; they're also doing faithful classic analog synth reproductions like the ARP instruments and their own MS-20 line: https://www.korg.com/us/products/synthesizers/
Open source or open synths is a pretty arbitrary category that few are rushing to compete in because most musicians don’t care how hackable their instrument is.
According to their financials, Roland are doing fine: profit’s down for 2022, but this seems partly related to a very profitable 2021. Can’t find any financials for Korg on their website, oddly enough, but when it comes to synths at least they seem to be competing only in the mid (Minilogue vs e.g. Boutiques) and low (volca vs new Aira trio) range. Korg doesn’t have any flagships any more and have few products in the DJ/production category. They’re apparently also 1/10th of Roland’s size…
Roland was founded by a hacker who was rejected from the Japanese university system, but has a little more of a top-down culture. Yamaha is the oldest Japanese music company, having started selling pianos in the 1880s, but focuses more on the pro audio & education markets nowadays. Also the only musical product company with a subsidiary that does everything from racing motorcycles to industrial robotics.
I like Korg, have a couple of their products and think they’re okay. I just never had the impression that they’re “winning” against Roland or even that they’re direct competitors.
The former are focusing on analog machines with some wave/fm detours. The Drumlogue didn’t seem that great from the demo, but the other logues are solid.
The latter make digital emulations of their classics, sell them in VST/cloud format and also in flagship, boutique and now small Aira formats. It seems to be working for them.
Do you know how they're making the synth sounds on the pi? Like are they going full pure-data/MaxMSP/SuperCollider with it or are they still using discrete sound chips?
They may still use discrete circuitry on their high end workstations, but I'm not sure. Korg embraced all-DSP synthesis early while their Japanese competitors (Roland, Yamaha, Akai) maxed out on sample playback and high polyphony. They have always had a narrow focus on synthesis, with minor incursions into the guitar utility/space, whereas Roland has a massive product catalog of electronic instruments (but a highly proprietary ethos) and Yamaha hasn't done anything in years (a pity, as they made great hardware sequencers. Akai is a special case as they basically own the market for hip-hop and its derivatives but are also unable to break out of it. I believe their gear runs on Linux, but like the other two their hardware/software stack is heavily locked down.
Regarding Yamaha and synthesizers I just wanted to mention that I consistently hear rave reviews of the Reface CS. I am more into vintage electric piano sounds than synths and currently own a Reface CP and the sounds on that are legit.
Akai has really been disappointing since they were purchased by inMusic brands. They have released a bunch of half baked products with serious bugs. (Looking at you, Akai Force).
They're using the Pi for the synth sounds, but it's not pure-data/MaxMSP/SuperCollider. They're likely just writing their own DSP and synthesis libraries for it. If you're interested check out https://www.electro-smith.com/daisy/daisy — this uses a lower powered Cortex processor, but you can see how the something like this could work: https://electro-smith.github.io/DaisySP/index.html
Small businesses can't access them - they are sold out everywhere and waiting lists are years long, while giant companies like Korg can afford to buy bulk and then flood the market with cheap devices.
That is probably their commercial secret, but given that for their price it makes sense for smaller businesses making niche products to buy them up and desolder the required components and they are easily available on Amazon, I'd guess the volume is large.
The shortage of Raspberry Pi's is going to be their undoing. Everyone in the circles I travel (3D printing, Hobby CNC, Self-hosting, cyberdeck, plotters, etc.) are looking for replacements. There are quite a few contenders, and more recently lots of folks are grabbing thin clients (then installing Linux).
My thought process; In a silo, the raspberry pi isn't all that spectacular and needs a lot of do-dads/thing-i-majigs for even basic functionality. Power supplies and real HDD/SSD's for Rpi's can be a pain point. There are cheaper and better alternatives imo. What makes the Raspberry Pi so cool is the community and ease of use due to said community effort. If that community coalesce around another platform, the rpi foundation will be in big trouble.
At the moment some of those competitors may have spare supply because people weren’t interested in them before. But once that runs out, I don’t see a reason they won’t all run into the same problems as the RPi Foundation as they get in line behind them for the limited-available time at the fabs.
* Performance garbage, single core, very cardy to use
* Some software porting risc-v has wonderful bugs
* No GPU, very laggy
and really nothing exciting on arm. I'm expecting these to disappear soon as well. really hoping some fab time opens soon to get even the allwinner devices back on the market at reasonable price/quantity.
Why do you think they will be supply constrained? Does Allwinner make Rockchips?
I see the competitors being a bit more flexible in regard to SoC used. There seems to be a plethora to choose from with different chips, but I am speaking from a very lay person POV.
I just bought a thin client and replaced three of my Pis and added a lot of spare room to host a Plex server and a Pleroma instance. The only remaining Pi in use drives my 3DPrinter. Thin clients seem really great for most typical server-type use cases with low power needs.
I don't see why an RPi is needed for a thing like that.
If you need a cheap Linux desktop, or a cheap Linux server, there is a ton of x86-based options.
It's when you need all these numerous GPIO pins is when an RPi starts making serious sense. (But an ESP32 could do, too, if running Linux is not a hard requirement.)
I read a comment recommending used thin clients on HN like a week ago and did the same thing. I set up one as an octoprint server, one for the hardware lab, and one for a retro gaming station. At ~$30 a piece, it was a no brainer.
I reused the rPis into a k3s cluster for experimentation so they aren't going to waste.
At this point, I think I'm going to rely on esp32 + thin clients for most of my projects. The only uses for rPis that I can still think of is a PiKVM, rPi Weather Station, and maybe a cyberdeck (mostly for the low power usage).
That sounds great, I don't know why I didn't think of that. Any models that you would recommend? Also, are there any independent retailers that gather and resell these that you'd suggest looking at?
I did a bunch of research on thin clients before buying one and I found a site[0] that compiles data on a bunch of different machines.
I ended up going with a Wyse 5060, it's a quad-core machine at 2.4ghz + 4gb of ram. It was pretty widely available for ~$35 on ebay so I figured even if it didn't work out, I wouldn't be out too much money.
As far as other retailers go, every discussion I found online basically told me to look at the usual places. eBay, FB marketplace, maybe craigslist. Unfortunately I didn't find any independent retailers.
Also, depending on your use case, you might want to look into AIO thin clients. You can pick up a full machine with a screen for under $50. Unfortunately, the specs on those machines were a little underpowered for what I needed, but it might be a better fit for other projects.
Thanks, these should be fun to dig in to. Most of the applications I look for involved 'detuning' the machine until I can run it fanlessly. I use these for servers for various things, small test hosts, etc.
Yep, and there are thousands of them around the $30 dollar mark. The case and power supply included!!!!! Governments (Local and State in the US) auction them off cheap if you really want to buy in bulk.
People have adapted all manner of thin, thick and virtual clients to run Klipper. I had it running on a VM in VirtualBox before I moved it to a Jetson Nano.
We were an industrial customer; comparatively small volumes (~100/quarter) but scaling up. The shortage has meant we've pivoted to something more guaranteed, so no more raspberry pi for us. Which is a shame as it's a nice platform and community.
> and more recently lots of folks are grabbing thin clients
But thin clients are typically much bigger than a Pi and consume much more power. Sure if it's something that never moves and is used as a server in an attic somewhere it might work, but if it's something small that you want to be able to carry around and throw in a backpack then the Pi is irreplaceable?
It's true that availability is a big problem though.
They are not that much bigger and the power draw isn't much more either, what would you say? The DELL WYSE Dx0Q has 15w printed on the case. How many people need to print or buy a case for the rpi? The dell comes with one... and is easily interfaced with a HDD/SSD and super easy to connect a cheap microcontroller to get GPIO if you need that.
I have been trying to find a Raspberry Pi Zero 2 W for months now. They are unobtainable, and I can’t find any communication from Raspberry Pi about it.
Don't know the backstory, but if the comapny that makes Raspberry Pi can make them enough to fulfil demand, how are those replacement going to come by? If it was easy, Raspberry's company would do it themselves...
I've read that the production is favouring industrial/commercial users, and that would include Korg. I was wondering where the large-scale users were (enough to create a consumer drought), but this would go some way to explaining it.
Raspberry Pi's supply chain & distribution were their biggest strength IMO, I can get their products on day 1 in India from number of vendors which I've not witnessed for any other HW.
Optionally, buy a PCM5102/PCM5122 or ES9018K2M I2S DAC to improve the output sound quality since the DAC built into the Raspberry Pi is not that great:
Alternatively you can run a DAW like SunVox on just about any smart device made in the last two decades from Windows CE and Palm to modern Android and iOS devices, as well as Linux, Mac, and Windows.
In most cases this is probably way more flexible, fully featured, easy to use, and cheaper than involving a Pi.
My current set up is a MIDI controller connected to an otherwise junk and unloved 2-in-1 laptop/tablet running Ubuntu Studio. Pretty low latency and SunVox is literally made for touchscreens and slow hardware. I love it!
I have not used Sunvox but I make music as a hobby using trackers like Renoise, and also various DAWs. I would still say that there is an advantage to having a specially designed hardware synth with intuitive knobs and faders. I'm glad that KORG can make reasonably affordable and professional synths while using off the shelf hardware like the Pi. In my opinion the presets and ease of use of paid synths (both software and hardware) are still quite a bit ahead of free options.
I don't want to make SunVox or its other fans look bad so I just encourage anyone reading this to give it a whirl.
It's jam packed with features and stellar patches/examples despite its simple and clean interface. It really is a joy to work with, especially on a tablet sized screen. It's so good when scaled properly that I prefer it over physical controls sometimes.
You may or may not like the pattern editor because it's a tracker interface, but the rest of the tool is so good live that you can initially ignore it and then slowly incorporate it into your workflow.
This could be .... kind of exciting because Korg makes some great synths (I am partial to the Minilogue, Prologue, and some of the Volcas) and now know that they might be hackable?!!
> 1988’s iconic M1 keyboard sold for $2,749, which would be about $6,800 in today’s money
They have great M1 apps for iOS (and Nintendo 3DS?!) as well as plugins for DAWs on macOS and Windows (and of course Gadget on mobile, desktop and ... Nintendo Switch - though I think the Switch doesn't have an M1 app or plugin.) Party like it's 1991.
Korg's iOS apps are all phenomenal. I particularly like the Gadget DAW and associated plugin collection, and the software versions of the Wavestation, ARP Odyssey, Electribe, and Polysix.
I also had a lot of fun with the DS-10/DS-10 plus app for the Nintendo DS/DSi.
I've had the M1 VST for a long time, and I've been really impressed that with every tech bottleneck (64bit OS, and just now Apple Silicon CPU) Korg has continued to update the software to work on the new hardware, even though this piece of software must be 15+ years old now. I love companies that behave like this, Korg has my loyalty.
Korg always seem to have been a bit pragmatic with using ready-made hardware where other manufacturers go at great lengths to come up with their own solution.
They should really be working on their own solutions as they can afford it rather than buying from pool of hardware small businesses are struggling to secure.
Korg themselves now struggle delivering those synthesizers. In many shops they are sold out for months. They are in exactly the same situation as those small businesses. It's quite disingenuous to blame Korg for a shortage of parts that had no shortage when they designed and first sold those synthesizers (more than two years ago).
Also, even big brand names in the synthesizer world are often comparatively small businesses themselves. Korg is probably a lot smaller than you imagine right now. Their name is much bigger than the actual company.
Everything but the keys, right? I've got my opsix right behind me, and I just can't believe how bad the keys are. TBF, it's right above my Moog Matriarch, so that might be a bit of an unfair comparison.
This is more or less a meme on music gear forums, and in my view it's a bit overstated. It's no Fatar to be sure, but it's still in line with what I'd expect from a sub-$1k keyboard. The critical thing they got right is the velocity sensitivity: Especially on an FM synth, you need that expressive range, and the Opsix keys deliver there.
I got a bit triggered by that "sub-$1k keyboard" of yours. I mean, you are totally right, this is a typical product for that price, i.e. this is a "normal" market price for this kind of product… But, man, I just can't get over the fact how ridiculously overpriced these things are! These are such simple devices, so much easier to produce than "classical" music instruments. Yet because of small market and the product being somewhat of a "professional tool" we still get these pretty basic keyboards for $1k+ and behave as if it's a totally fair price because of hammer action and stuff (and take it for granted that $500 devices should have shitty keyboards).
Sure, good mechanics are worth something, but (especially when talking about midi-keyboards) it's the only thing that's complicated in these devices, and it isn't that expensive to produce.
>I just can't get over the fact how ridiculously overpriced these things are!
Surely you understand that you're paying for value, and what's valued here is the combination of the software engineering, hardware design, and the cost of manufacturing, assembling, packaging, warehousing, and shipping the final product.
Korg were blowing out the Opsix on Reverb four months ago for $329.
That's about $50 more than a set of nice cello strings.
I agree that the velocity sensitive is good, but the keybed is by far the worst one in my studio, regardless of price. I have plenty of cheaper keybeds that feel significantly better.
"In early 2020, Korg R&D announced the wavestate, a successor to its 30-year-old Wavestation and its first instrument to use Raspberry Pi Compute Module 3."
I came across the LMN 3: there’s a lot to unpack but it might be the most mature project in terms of both hardware and software out there. It’s as close to having the right tools, PCBs, etc. for an open source Teenage Engineering OP-1:
There's a bunch of people doing some pretty amazing synth builds with the Raspberry Pi -- the Zynthian crew [0] springs to mind.
Basically bring your own USB midi keyboard / controller - these tend to be cheap, but also engender very strong opinions, so there's some distinct advantages to having them as separate components, but with the synth box being much more portable than a laptop or desktop.
As to the Korg Wavestate - on this side of the pond (AU) it has an RRP of A$1500, though street pricing is around A$1000.
I disagree, the success story has details that reveal lots on what's going on.
> The team uses software to prototype their instruments before implementing hardware designs. With the basic software platform already functional, developing the wavestate using Compute Module 3 took a fairly modest year
From that I would assume that software is developed on regular PCs and it took a year for them to get the software running on CM3 and hook up the CM3 to the two circuit boards and various systems. Which is super fast.
> The setup has two circuit boards. The main panel board contains all of the user interface elements, including display, buttons, knobs, wheels, and other synth-specific controls, along with MCU microprocessors to support them and communicate with the CM3.
Main board has all the physical buttons, knobs wheels, displays etc and MCUs are used to communicate with CM3
> The other circuit board has subsystems for audio, MIDI, the musical keyboard, and power, plus the socket for the CM3
The 2nd circuit board has D/A converters, midi connectors, keys and power and the CM3.
The CM3 is basically responsible for all the computations on the device. It gets inputs from various sources and outputs constantly digital audio to the DAC, midi to the midi out, data to the display etc.
I guess this would be the part where details would have been nice, but there's probably lots going on. How they ensure low latency function of the synth platform, how does the development process go, how does the CM3 integrate with the various systems. Each of those would be very indepth stuff, but imo the HN relevant part how they sped up the overall development of the platform and that is covered by the article.
Anyways, the cool part is how the three devices use the same hardware, so Korg can basically recycle both hardware designs, components and software from synth to synth. This speeds up development and reduces costs. Super cool.
Comparing that to how synths were made in the 80s, where you had to have a separate board per voice and replicate all the analog components between voices and keep their power usage and heat in control.
Thomann has nice pictures of the synths. The reuse is very obvious:
Yeah, I went looking for one about two months ago and decided it wasn't worth it.
The value calculation has always been a bit iffy vs "let's see what small-factor desktop machines people are dumping in the local classifieds" if what you want is a cheap PC to mess with. That's probably very much true for software instrument / sequencing stuff.
Because of the ecosystem, the value calculation has been better for the Pi if you're building projects that are a small computer built into a larger electronic setup. But at the current price & availability I'm definitely trying to figure out what other single board options are easy enough or considering learning more about the microcontroller space.
… they are perpetually out of stock, even in brick and mortar retail.
Which makes it all the much odder for RPi to write what amounts to a press piece about their hardware. You don't need more marketing, you need the price to not be $∞ due to supply issues.
Recently, though? This worked earlier this year, but one vendor had at least 90x RPI4-8B in stock yesterday and they were gone in seconds. rpilocator didn't even see it.
I've bought 2 or 3 Pi's this year at the regular price, thanks to rpilocator. The most recent was a couple of months ago.
You do have to kinda watch it a bit obsessively if you're really serious about finding one, but it can be done. Or you could also just constantly refresh the page(s) of a couple of vendors that you know to get stock on some regular / semi-regular basis. Adafruit is one, and Elektor seems to get stock now and then as well.
You'd be surprised how much of your medical data is generated by pis! Many medical OEMs I work with encourage their clients to integrate such chips into their machines. I just can't wait until they're cycled thru so I can play quake 3 on a centrifuge
I wonder if the Raspi uses OS in this use case. I imagine using the hardware directly like a microcontroller is also a possibility. I also wonder if they use some kind of linux distribution with their own software on top of it.
Korg also makes their digital and analog ecosystem fairly open at the lower end; analog devices are designed to be easily hackable and they facilitate sideloading of DSP applications; code that runs on this device can also be loaded onto several of their consumer models: https://www.korg.com/us/products/dj/nts_1/
This has a lot to do with why they've been eating Roland's lunch for the last several years, while that latter company keeps trying to convince buyers that completely locked down circuit simulation of classic devices at fancy prices is what the people want.
At the cheaper end of the scale, companies like Sonicware are exploiting commodity-priced microcontrollers for their 12-bit synthesis and processing capabilities. While their own offerings are not open-source, abundant resources exist for anyone interested in processing audio on crunchy chips and you can get into it for about $20 in hardware costs (years of your life not refundable; synthesizers are nerd crack).
Roland is doing just fine selling digital emulations of their classic gear. But they aren't exactly amazing innovative instruments. I look at Roland today as the Gen-X Harley-Davidson: Selling nostalgia gear to aging dudes hitting their mid-life crisis.
Korg has some missteps (the Electribe 2 could be so much better), but seems to still want to make unique, interesting instruments. I'm disappointed in the build quality of the wavestate/modwave/opsix, but their other gear is top notch.
> I look at Roland today as the Gen-X Harley-Davidson: Selling nostalgia gear to aging dudes hitting their mid-life crisis.
So true, and it huuurts. :D
However, one very valid selling point for gear (old or new) is that it is still, by far, a better user/musical experience than software emulation, even with multiple innovative controllers. Either because a screen lack multiple physical points of control, or because software/plugins rights managements are _always_ getting in the way.
Of course, only if you do something with it and it doesn't pile up in a too small room.
Oh, yes, I am definitely on the hardware synthesizer UX train.
What’s an innovative instrument and does that matter when making music? There’s quite a fetishization of gear happening nowadays, making and buying gear for gear’s sake.
By now one can see that subtractive is king, because it’s easy to understand, sounds great and one can hear it everywhere. And subtractive’s pretty much done, manufacturers are just tinkering with filters, modulation and effects at this point. Popular VSTs nowadays are mixing several types of engines (Pigments) or expanding the set of oscillator waves (Serum). In HW we see more of the same or subtractives.
Subtractive will never get old and Roland classics like the Juno, Jupiter or SH-101 just sound good. The Roland drum machines sound good. I never owned any classics and don’t revere that “Roland sound”, but at the end of the day they sound good, so they’re good - and more importantly they’re enough.
All the innovation happening in sound synthesis seems like over engineering and developing things for the sake of novelty.
An innovative instrument is one that's trying to sound like 2022 and not 1982.
I love the hardware Roland sound. Especially the Juno 60 which is one of the sweetest sounding synths ever made. The Jupiters and Junos are musical in a way that other synths never quite matched.
But the Roland Cloud and Zencore versions are close, but... eh. And Roland Cloud Manager is an absolute clusterfuck of a product. It seems like Roland are all about the money now, and they don't even like their customers all that much.
Meanwhile... I bought ModWave Native yesterday and it's the most fun I've had with a synth for a long time. It sounds nothing like a 1982 throwback. The wonderfully bonkers gravity-well based KAOSS modulation makes the sounds organic and unpredictable, far away from any traditional ADSR/LFO combination.
I don't think it's gratuitous or over engineered at all. It's in a nice spot where there's some interesting added complexity on top of a recognisable wavetable engine.
It also works much better as a VST than as hardware because there's more polyphony, it's easier to edit, and it doesn't take up any space.
There’s nothing inherent to the Juno 60 (to take something with the simplest architecture) that makes it sound like the 80s vs 2022. For electronic music it’s the drum beats, melodies, arrangement, the modulation, filter use and effects that have changed with the times.
Wavetable’s popular right now, so credit due to Korg for putting that out in HW format. On the other hand there’s also Waldorf which built their company over decades around this type of synthesis both in HW and plug-in format. ;-) The thing with wavetable is that it works much better as a VST. One look at the front panel of the modwave lets me know that fiddling with the dozens of buttons and knobs would not be fun for me.
Anyway, there’s hardly anything revolutionary happening in synths. My theory’s that Roland recognized that and are focusing on workflow, availability, interoperability, etc.
Personally I don’t get it why there’s so many new synths being built…
Korg is really amazing; they're also doing faithful classic analog synth reproductions like the ARP instruments and their own MS-20 line: https://www.korg.com/us/products/synthesizers/
Open source or open synths is a pretty arbitrary category that few are rushing to compete in because most musicians don’t care how hackable their instrument is.
According to their financials, Roland are doing fine: profit’s down for 2022, but this seems partly related to a very profitable 2021. Can’t find any financials for Korg on their website, oddly enough, but when it comes to synths at least they seem to be competing only in the mid (Minilogue vs e.g. Boutiques) and low (volca vs new Aira trio) range. Korg doesn’t have any flagships any more and have few products in the DJ/production category. They’re apparently also 1/10th of Roland’s size…
Korg is and always has been a labor of love that honors the craftsmanship of individual engineers.
https://www.soundonsound.com/music-business/history-korg-par...
https://www.redbullmusicacademy.com/lectures/korg-hiroaki-ni...
Roland was founded by a hacker who was rejected from the Japanese university system, but has a little more of a top-down culture. Yamaha is the oldest Japanese music company, having started selling pianos in the 1880s, but focuses more on the pro audio & education markets nowadays. Also the only musical product company with a subsidiary that does everything from racing motorcycles to industrial robotics.
I like Korg, have a couple of their products and think they’re okay. I just never had the impression that they’re “winning” against Roland or even that they’re direct competitors.
The former are focusing on analog machines with some wave/fm detours. The Drumlogue didn’t seem that great from the demo, but the other logues are solid.
The latter make digital emulations of their classics, sell them in VST/cloud format and also in flagship, boutique and now small Aira formats. It seems to be working for them.
The rule of thumb in game making is you can /either/ make and awesome game /or/ an awesome game engine... Unless you're John Carmack.
Likewise, in music you can either make cool music or make cool instruments.
That's so cool! Thanks for sharing that!
Do you know how they're making the synth sounds on the pi? Like are they going full pure-data/MaxMSP/SuperCollider with it or are they still using discrete sound chips?
Mostly in C; check out the Logue-SDK repo: https://github.com/korginc
You might also like this (though I don't know if it's using the R-Pi or is user-moddable): https://www.korg.com/us/products/dj/nts_2_pt/
They may still use discrete circuitry on their high end workstations, but I'm not sure. Korg embraced all-DSP synthesis early while their Japanese competitors (Roland, Yamaha, Akai) maxed out on sample playback and high polyphony. They have always had a narrow focus on synthesis, with minor incursions into the guitar utility/space, whereas Roland has a massive product catalog of electronic instruments (but a highly proprietary ethos) and Yamaha hasn't done anything in years (a pity, as they made great hardware sequencers. Akai is a special case as they basically own the market for hip-hop and its derivatives but are also unable to break out of it. I believe their gear runs on Linux, but like the other two their hardware/software stack is heavily locked down.
Regarding Yamaha and synthesizers I just wanted to mention that I consistently hear rave reviews of the Reface CS. I am more into vintage electric piano sounds than synths and currently own a Reface CP and the sounds on that are legit.
Roland say that their ACB machines are DSP-based: https://www.rolandus.com/blog/2014/02/14/analog-circuit-beha.... That would include the System 8 and Boutiques. I imagine that the ZenCore stuff runs on some SoC, since it’s all software.
The Korgs I like are analog although that’s typically all controlled digitally. Is their purely digital stuff all that successful?
Akai has really been disappointing since they were purchased by inMusic brands. They have released a bunch of half baked products with serious bugs. (Looking at you, Akai Force).
They're using the Pi for the synth sounds, but it's not pure-data/MaxMSP/SuperCollider. They're likely just writing their own DSP and synthesis libraries for it. If you're interested check out https://www.electro-smith.com/daisy/daisy — this uses a lower powered Cortex processor, but you can see how the something like this could work: https://electro-smith.github.io/DaisySP/index.html
They are cannibalizing on 303-808-909. If they only stop with that green leds.
So there are going all those STM32F4...
Small businesses can't access them - they are sold out everywhere and waiting lists are years long, while giant companies like Korg can afford to buy bulk and then flood the market with cheap devices.
This is not cool.
How many devices do you think Korg is putting out?
That is probably their commercial secret, but given that for their price it makes sense for smaller businesses making niche products to buy them up and desolder the required components and they are easily available on Amazon, I'd guess the volume is large.
The shortage of Raspberry Pi's is going to be their undoing. Everyone in the circles I travel (3D printing, Hobby CNC, Self-hosting, cyberdeck, plotters, etc.) are looking for replacements. There are quite a few contenders, and more recently lots of folks are grabbing thin clients (then installing Linux).
My thought process; In a silo, the raspberry pi isn't all that spectacular and needs a lot of do-dads/thing-i-majigs for even basic functionality. Power supplies and real HDD/SSD's for Rpi's can be a pain point. There are cheaper and better alternatives imo. What makes the Raspberry Pi so cool is the community and ease of use due to said community effort. If that community coalesce around another platform, the rpi foundation will be in big trouble.
At the moment some of those competitors may have spare supply because people weren’t interested in them before. But once that runs out, I don’t see a reason they won’t all run into the same problems as the RPi Foundation as they get in line behind them for the limited-available time at the fabs.
I've been trying to find a pi zero 2w for quite a while, and have started looking at whatever else I can import at a decent price.
so far, I've run into things like https://www.aliexpress.us/item/3256804186541934.html with such interesting features as:
and really nothing exciting on arm. I'm expecting these to disappear soon as well. really hoping some fab time opens soon to get even the allwinner devices back on the market at reasonable price/quantity.I think RPi's problem is specifically Broadcom because the Pi Pico is available and the only one that doesn't use a Broadcom chip.
This level of supply issues isn't affecting competitors.
Why do you think they will be supply constrained? Does Allwinner make Rockchips?
I see the competitors being a bit more flexible in regard to SoC used. There seems to be a plethora to choose from with different chips, but I am speaking from a very lay person POV.
I just bought a thin client and replaced three of my Pis and added a lot of spare room to host a Plex server and a Pleroma instance. The only remaining Pi in use drives my 3DPrinter. Thin clients seem really great for most typical server-type use cases with low power needs.
I don't see why an RPi is needed for a thing like that.
If you need a cheap Linux desktop, or a cheap Linux server, there is a ton of x86-based options.
It's when you need all these numerous GPIO pins is when an RPi starts making serious sense. (But an ESP32 could do, too, if running Linux is not a hard requirement.)
The power draw and size is substantially lower.
I read a comment recommending used thin clients on HN like a week ago and did the same thing. I set up one as an octoprint server, one for the hardware lab, and one for a retro gaming station. At ~$30 a piece, it was a no brainer.
I reused the rPis into a k3s cluster for experimentation so they aren't going to waste.
At this point, I think I'm going to rely on esp32 + thin clients for most of my projects. The only uses for rPis that I can still think of is a PiKVM, rPi Weather Station, and maybe a cyberdeck (mostly for the low power usage).
That sounds great, I don't know why I didn't think of that. Any models that you would recommend? Also, are there any independent retailers that gather and resell these that you'd suggest looking at?
I did a bunch of research on thin clients before buying one and I found a site[0] that compiles data on a bunch of different machines.
I ended up going with a Wyse 5060, it's a quad-core machine at 2.4ghz + 4gb of ram. It was pretty widely available for ~$35 on ebay so I figured even if it didn't work out, I wouldn't be out too much money.
As far as other retailers go, every discussion I found online basically told me to look at the usual places. eBay, FB marketplace, maybe craigslist. Unfortunately I didn't find any independent retailers.
Also, depending on your use case, you might want to look into AIO thin clients. You can pick up a full machine with a screen for under $50. Unfortunately, the specs on those machines were a little underpowered for what I needed, but it might be a better fit for other projects.
[0]: https://www.parkytowers.me.uk/thin/hware/hardware.shtml
Interestingly, those Wyse 5060 (at least the Dell ones) have 2x SO-DIMM slots so the ram can be upgraded (to 16GB):
https://www.youtube.com/watch?v=Pvg7hOXpHFU
Thanks, these should be fun to dig in to. Most of the applications I look for involved 'detuning' the machine until I can run it fanlessly. I use these for servers for various things, small test hosts, etc.
Yep, and there are thousands of them around the $30 dollar mark. The case and power supply included!!!!! Governments (Local and State in the US) auction them off cheap if you really want to buy in bulk.
People have adapted all manner of thin, thick and virtual clients to run Klipper. I had it running on a VM in VirtualBox before I moved it to a Jetson Nano.
We were an industrial customer; comparatively small volumes (~100/quarter) but scaling up. The shortage has meant we've pivoted to something more guaranteed, so no more raspberry pi for us. Which is a shame as it's a nice platform and community.
> and more recently lots of folks are grabbing thin clients
But thin clients are typically much bigger than a Pi and consume much more power. Sure if it's something that never moves and is used as a server in an attic somewhere it might work, but if it's something small that you want to be able to carry around and throw in a backpack then the Pi is irreplaceable?
It's true that availability is a big problem though.
They are not that much bigger and the power draw isn't much more either, what would you say? The DELL WYSE Dx0Q has 15w printed on the case. How many people need to print or buy a case for the rpi? The dell comes with one... and is easily interfaced with a HDD/SSD and super easy to connect a cheap microcontroller to get GPIO if you need that.
I have been trying to find a Raspberry Pi Zero 2 W for months now. They are unobtainable, and I can’t find any communication from Raspberry Pi about it.
If you are in the US and have a MicroCenter near you, they usually carry them. I see them every time I visit the MicroCenter near me.
Are you sure you aren't seeing the pico? Those are still easy to find.
I see both, I am sure.
I was able to order a pi 4 B+ this morning by watching this site. They sell very quickly once listed so it took couple weeks before I could land one.
https://rpilocator.com/
Don't know the backstory, but if the comapny that makes Raspberry Pi can make them enough to fulfil demand, how are those replacement going to come by? If it was easy, Raspberry's company would do it themselves...
I've read that the production is favouring industrial/commercial users, and that would include Korg. I was wondering where the large-scale users were (enough to create a consumer drought), but this would go some way to explaining it.
Raspberry Pi's supply chain & distribution were their biggest strength IMO, I can get their products on day 1 in India from number of vendors which I've not witnessed for any other HW.
For people interested in this, there are several open source projects which turn Raspberry Pis into a MIDI synthesizers:
MiniDexed: 8 DX7 Tone Generators
- https://github.com/probonopd/MiniDexed
- https://www.youtube.com/watch?v=Z3t94ceMHJo
MT32-Pi: Roland MT-32 emulator
- https://github.com/dwhinham/mt32-pi
- https://www.youtube.com/watch?v=vaSD_wbzJRw
Optionally, buy a PCM5102/PCM5122 or ES9018K2M I2S DAC to improve the output sound quality since the DAC built into the Raspberry Pi is not that great:
https://raspberrypi.stackexchange.com/questions/76188/how-to...
PCM5102 (~$7): https://www.aliexpress.com/wholesale?SearchText=PCM5102+DAC
ES9018K2M (~$12): https://www.aliexpress.com/wholesale?SearchText=es9018k2m+i2...
And likely you can port these to systems you can buy right now like the Teensy 4.x - or they are already ported.
Like MicroDexed https://codeberg.org/positionhigh/MicroDexed-touch
https://www.parasitstudio.de/
Alternatively you can run a DAW like SunVox on just about any smart device made in the last two decades from Windows CE and Palm to modern Android and iOS devices, as well as Linux, Mac, and Windows.
In most cases this is probably way more flexible, fully featured, easy to use, and cheaper than involving a Pi.
My current set up is a MIDI controller connected to an otherwise junk and unloved 2-in-1 laptop/tablet running Ubuntu Studio. Pretty low latency and SunVox is literally made for touchscreens and slow hardware. I love it!
I have not used Sunvox but I make music as a hobby using trackers like Renoise, and also various DAWs. I would still say that there is an advantage to having a specially designed hardware synth with intuitive knobs and faders. I'm glad that KORG can make reasonably affordable and professional synths while using off the shelf hardware like the Pi. In my opinion the presets and ease of use of paid synths (both software and hardware) are still quite a bit ahead of free options.
Oof that last sentence :)
I don't want to make SunVox or its other fans look bad so I just encourage anyone reading this to give it a whirl.
It's jam packed with features and stellar patches/examples despite its simple and clean interface. It really is a joy to work with, especially on a tablet sized screen. It's so good when scaled properly that I prefer it over physical controls sometimes.
You may or may not like the pattern editor because it's a tracker interface, but the rest of the tool is so good live that you can initially ignore it and then slowly incorporate it into your workflow.
SunVox is indeed amazing.
This could be .... kind of exciting because Korg makes some great synths (I am partial to the Minilogue, Prologue, and some of the Volcas) and now know that they might be hackable?!!
> 1988’s iconic M1 keyboard sold for $2,749, which would be about $6,800 in today’s money
They have great M1 apps for iOS (and Nintendo 3DS?!) as well as plugins for DAWs on macOS and Windows (and of course Gadget on mobile, desktop and ... Nintendo Switch - though I think the Switch doesn't have an M1 app or plugin.) Party like it's 1991.
Korg's iOS apps are all phenomenal. I particularly like the Gadget DAW and associated plugin collection, and the software versions of the Wavestation, ARP Odyssey, Electribe, and Polysix.
I also had a lot of fun with the DS-10/DS-10 plus app for the Nintendo DS/DSi.
I've had the M1 VST for a long time, and I've been really impressed that with every tech bottleneck (64bit OS, and just now Apple Silicon CPU) Korg has continued to update the software to work on the new hardware, even though this piece of software must be 15+ years old now. I love companies that behave like this, Korg has my loyalty.
Aside: the poly-800 is a pile of fun!
Korg always seem to have been a bit pragmatic with using ready-made hardware where other manufacturers go at great lengths to come up with their own solution.
Look at the inside of this Korg M3: https://www.personal-view.com/talks/uploads/FileUpload/37/98...
It seems like the operating system comes on an SD card (makes upgrades easy, I guess), and the power supply looks like a standard laptop power supply.
They should really be working on their own solutions as they can afford it rather than buying from pool of hardware small businesses are struggling to secure.
Korg themselves now struggle delivering those synthesizers. In many shops they are sold out for months. They are in exactly the same situation as those small businesses. It's quite disingenuous to blame Korg for a shortage of parts that had no shortage when they designed and first sold those synthesizers (more than two years ago).
Also, even big brand names in the synthesizer world are often comparatively small businesses themselves. Korg is probably a lot smaller than you imagine right now. Their name is much bigger than the actual company.
It would be cool if they actually mentioned which synths use the Raspberry Pi, but it appears they don’t.
It's the modwave, wavestate and the opsix.
They all share the same core hardware and physical layout and enclosure but have customised front panels.
Actually really good products, well designed, I love the opsix :)
Everything but the keys, right? I've got my opsix right behind me, and I just can't believe how bad the keys are. TBF, it's right above my Moog Matriarch, so that might be a bit of an unfair comparison.
This is more or less a meme on music gear forums, and in my view it's a bit overstated. It's no Fatar to be sure, but it's still in line with what I'd expect from a sub-$1k keyboard. The critical thing they got right is the velocity sensitivity: Especially on an FM synth, you need that expressive range, and the Opsix keys deliver there.
I got a bit triggered by that "sub-$1k keyboard" of yours. I mean, you are totally right, this is a typical product for that price, i.e. this is a "normal" market price for this kind of product… But, man, I just can't get over the fact how ridiculously overpriced these things are! These are such simple devices, so much easier to produce than "classical" music instruments. Yet because of small market and the product being somewhat of a "professional tool" we still get these pretty basic keyboards for $1k+ and behave as if it's a totally fair price because of hammer action and stuff (and take it for granted that $500 devices should have shitty keyboards).
Sure, good mechanics are worth something, but (especially when talking about midi-keyboards) it's the only thing that's complicated in these devices, and it isn't that expensive to produce.
>I just can't get over the fact how ridiculously overpriced these things are!
Surely you understand that you're paying for value, and what's valued here is the combination of the software engineering, hardware design, and the cost of manufacturing, assembling, packaging, warehousing, and shipping the final product.
Korg were blowing out the Opsix on Reverb four months ago for $329. That's about $50 more than a set of nice cello strings.
You're paying for the innovation.
I agree that the velocity sensitive is good, but the keybed is by far the worst one in my studio, regardless of price. I have plenty of cheaper keybeds that feel significantly better.
"In early 2020, Korg R&D announced the wavestate, a successor to its 30-year-old Wavestation and its first instrument to use Raspberry Pi Compute Module 3."
so https://www.korg.com/products/synthesizers/wavestate/
How do you break into the sound world? It looks like a really fun space to be a programmer. I love hardware and tactile stuff.
InMusic/Akai/Alesis/Numark are hiring. Just learn C/C++ and JUCE and read a few digital signal processing book, you're good to go.
The market for synthesizers/grooveboxes has literally exploded in the last 10 years.
I came across the LMN 3: there’s a lot to unpack but it might be the most mature project in terms of both hardware and software out there. It’s as close to having the right tools, PCBs, etc. for an open source Teenage Engineering OP-1:
https://www.synthtopia.com/content/2022/06/25/the-lmn-3-an-o...
Look at Bela and Electro-smith Daisy, there are a bunch of simple diy synths made with teensy and some minimal components
https://bela.io/ https://www.electro-smith.com/daisy
There's a bunch of people doing some pretty amazing synth builds with the Raspberry Pi -- the Zynthian crew [0] springs to mind.
Basically bring your own USB midi keyboard / controller - these tend to be cheap, but also engender very strong opinions, so there's some distinct advantages to having them as separate components, but with the synth box being much more portable than a laptop or desktop.
As to the Korg Wavestate - on this side of the pond (AU) it has an RRP of A$1500, though street pricing is around A$1000.
[0] https://zynthian.org/
Also unusual music related things: Moog
https://en.wikipedia.org/wiki/Moog_synthesizer
https://en.wikipedia.org/wiki/Moog_Inc.
The cousin of the guy who made the Moog synthesizer created that company, which makes all sorts of technical/industrial/defense industry things.
Some details about the insides would be useful. This is barely a marketing blurb.
Edit: the success story is also thin on details.
I disagree, the success story has details that reveal lots on what's going on.
> The team uses software to prototype their instruments before implementing hardware designs. With the basic software platform already functional, developing the wavestate using Compute Module 3 took a fairly modest year
From that I would assume that software is developed on regular PCs and it took a year for them to get the software running on CM3 and hook up the CM3 to the two circuit boards and various systems. Which is super fast.
> The setup has two circuit boards. The main panel board contains all of the user interface elements, including display, buttons, knobs, wheels, and other synth-specific controls, along with MCU microprocessors to support them and communicate with the CM3.
Main board has all the physical buttons, knobs wheels, displays etc and MCUs are used to communicate with CM3
> The other circuit board has subsystems for audio, MIDI, the musical keyboard, and power, plus the socket for the CM3
The 2nd circuit board has D/A converters, midi connectors, keys and power and the CM3.
The CM3 is basically responsible for all the computations on the device. It gets inputs from various sources and outputs constantly digital audio to the DAC, midi to the midi out, data to the display etc.
I guess this would be the part where details would have been nice, but there's probably lots going on. How they ensure low latency function of the synth platform, how does the development process go, how does the CM3 integrate with the various systems. Each of those would be very indepth stuff, but imo the HN relevant part how they sped up the overall development of the platform and that is covered by the article.
Anyways, the cool part is how the three devices use the same hardware, so Korg can basically recycle both hardware designs, components and software from synth to synth. This speeds up development and reduces costs. Super cool.
Comparing that to how synths were made in the 80s, where you had to have a separate board per voice and replicate all the analog components between voices and keep their power usage and heat in control.
Thomann has nice pictures of the synths. The reuse is very obvious:
https://www.thomann.de/fi/korg_wavestate.htm
https://www.thomann.de/fi/korg_opsix.htm
https://www.thomann.de/fi/korg_modwave.htm
The insides of Yamaha CS-80 from 1977 (weight: 82kg) look a bit different
https://www.youtube.com/watch?v=_poihkLM5Go
Of course, with modern components a CS-80 clone (Black Corporation Deckard's Dream mk2) fits in to rack format and weighs only 4.5kg:
https://www.youtube.com/watch?v=BNf0kpidGc4
but the assembly of the DIY kit looks pretty painful:
https://www.youtube.com/watch?v=fk-pM2OBU1o
the average pi costs about $200 these days (if you can find one)
Yeah, I went looking for one about two months ago and decided it wasn't worth it.
The value calculation has always been a bit iffy vs "let's see what small-factor desktop machines people are dumping in the local classifieds" if what you want is a cheap PC to mess with. That's probably very much true for software instrument / sequencing stuff.
Because of the ecosystem, the value calculation has been better for the Pi if you're building projects that are a small computer built into a larger electronic setup. But at the current price & availability I'm definitely trying to figure out what other single board options are easy enough or considering learning more about the microcontroller space.
… they are perpetually out of stock, even in brick and mortar retail.
Which makes it all the much odder for RPi to write what amounts to a press piece about their hardware. You don't need more marketing, you need the price to not be $∞ due to supply issues.
They're only out of stock because they only sell in retail channels what is left over after industrial customers have bought everything they need.
So if you look at this as an advert about industrial usage aimed at industrial users, it makes sense.
Check out rpilocator.com. I was able to get 4 different pis at MSRP over the course of this year.
I'm jealous. Every time I see adafruit get restocked, it's all gone less than 10 minutes later. I've basically given up on Raspberry Pi at this point.
That mirrors my experience with Adafruit. I had better luck with digi-key and pishop.
Honestly I lost count how many times I perked up from the alert on rpilocator only for the stock to disappear while I was going through the checkout.
Recently, though? This worked earlier this year, but one vendor had at least 90x RPI4-8B in stock yesterday and they were gone in seconds. rpilocator didn't even see it.
I've bought 2 or 3 Pi's this year at the regular price, thanks to rpilocator. The most recent was a couple of months ago.
You do have to kinda watch it a bit obsessively if you're really serious about finding one, but it can be done. Or you could also just constantly refresh the page(s) of a couple of vendors that you know to get stock on some regular / semi-regular basis. Adafruit is one, and Elektor seems to get stock now and then as well.
I think the latest one I got was in July and it was a CM4 8GB/32GB
Raspberry Pi - Enterprise Edition
You'd be surprised how much of your medical data is generated by pis! Many medical OEMs I work with encourage their clients to integrate such chips into their machines. I just can't wait until they're cycled thru so I can play quake 3 on a centrifuge
I wonder if the Raspi uses OS in this use case. I imagine using the hardware directly like a microcontroller is also a possibility. I also wonder if they use some kind of linux distribution with their own software on top of it.
I am curious if there are any great plans to make your own synth with a pi?