Developer Jan Ostman has kicked off an IndieGoGo project to develope the DSP-G1 synthesizer – a DSP-based synth that’s designed as “a tribute to analog synthesizers of the 70’s and 80’s”.
Features:
- The dsp-G1 is a computer model of an analog synthesizer, running with a sample resolution of 44.1KHz and 16-bit. It reads the parameters from the front panel knobs and calculates the sound output. It is played through a MIDI-IN connection.
- It features a 3-DCO variable waveform oscillator, 18dB Lowpass filter with resonance, 2 Envelopes and 1 LFO. That adds up to 17 frontpanel realtime knobs.
- The waveform can be seamlessly morphed from Sine, Triangle, Pulse to Saw.
- There is also a Wrap setting for additional harmonics.
- The 18dB resonant lowpass filter sounds very realistic and selfoscillates at high resonance settings.
- It’s 5-Note paraphonic meaning that it can take 5-key chords but they are all routed through single filter and amp envelopes.
- All internal calculations are done at 32-bit integer resolution and output at 16-bits.
- The dsp-G1 is a standalone synthesizer and requires no computer to play.
- Just MIDI input from a keyboard or sequencer.
Here’s an audio demo:
“The great sounds from Roland, Sequential Circuits, Oberheim and Moog can once again live in the form of Analog Modeling on microcontrollers and dsp-chips,” says Ostman. “So I have written the analog algorithms to recreate the sound of the past.”
Here’s another audio demo:
The DSP-G1 is available to supporters of the project for $200. Details on the DSP-G1 are available at the project site.
I am intrigued by this offer.
Anyone going to jump on it?
Wow, a virtual analog synthesizer, what an innovative idea!
I know, right? Lol
“analog algorithms” … ok than.
Perhaps I’m being too sceptical but they are looking for $200 and already have a proto built? So why can’t we see it? All we can see is some resistors and hookup wire with an un-socketed IC and the wrong number of knobs on some proto board… Can we see the other side of the board or get a little more info on the DSP chip and microcontroller being used?
On the single available sound file there is a lot more than a single saw going on – I’d love to see it in action! Really hope I’m wrong and they get fully funded but I would need to see more before jumping aboard…
Hi
The board you are seeing in the picture is just the frontpanel scanboard, the knobs and an analog mux to scan them. There is nothing on the backside except a capacitor on each pot.
The DSP is a 72MHz variant from NXP.
The sound you hear is the dsp-G1 singelSaw.
It always run with 3-DCOs on a singel voice with range and detune knobs.
So it actually have 15 DCOs.
The microcontroller is an ATmega responsible for scanning the knobs, receiving MIDI and handling the patchmemory.
Hi Jan,
Thanks for posting! It does look like an intriguing product.
I’m no analog purist, and I think if oscillators & filters can sound good in the digital domain that’s fine with me.
Nice to see a box with knobs, and a somewhat limited function set. (Easily mastered.)
I’d be curious to hear where you think your product would differ from other software emulations.
Will the software evolve (i.e., is it updatable?) Is there a chance that a noise osc will be included at some point?
A noise osc is very possible as it hardly requires any processing power.
But it requires another knob and I already added one as someone noticed.
Via MIDI-CC its possible to add just about anything but I like one knob per parameter.
Much more creative that way and makes you forget its actually digital.
I have some analog inputs free as I wanted to add CV/trig jacks on the back.
The software is fully updatable over USB, no problem.
I’m interested but you need to do some audio and video demos if you want people to back this.
The price seems ok, but I want to have some confidence in the project.
Another way you might be able to boost the range of “takers” would be to have the USB be a class-compliant MIDI interface, (hear me out) so the knobs transmit CCs to VI’s via USB, AND it would merge the incoming MIDI input with the knobs. I think that would be unique.
The other nice feature about this is that it is a standalone synth, but with other useful functions.
I would not be bothered by a lack of presets, because I have other rigs from which I can call up presets (cue understatement banner). Having a rig that is purely set with knobs would be both fun and kind of cool.
Someone mentioned the resolution. The main goal would be to minimize aliasing and eliminate digital clipping. I’d say if you can make that happen with 16/44.1 then good. But it would seem that 24 bit would make the sound noticeably better (given the various ways it would be used).
I’d back this if there were more videos, an idea of what the final product would resemble, more info on how this synth will differ in sound from the others I own, where the added value is, more on the backstory, etc. I’d recommend more marketing if you want this thing to really take off.
I’m beginning to hate the word “paraphonic”
44.1KHz/16-bit seems kinda low by today’s standards. I’m sure this project is neat and all, but unless it’s 96Khz/24-bit, and sports a few unusual features, I don’t see how it’s supposed to compete with a number of software synthesizers.
Going that quality would be really nice!!
Jan can’t you do that?
44.1KHz/16-bit is the standard
I can’t speak for the developer, but while 44.1/16 is “A” standard, but not “THE” standard, in this millennium, anyways.
As a format, 44/16 leaves very little wiggle room for varying levels, no headroom, etc. It is fine for listening to masters, but for synths, where levels can be all over the place, 16 bits is perhaps cutting it close.
just want a new Korg DSS1 ( with same features ) …
Why limit a VA to being paraphonic?
Good question, but since the synth IS paraphonic, there’s a more interesting design choice (in my opinion) to be made. The synth could be made with a digital oscillator and an analog VCF. There’s a lot to be said for pairing a digital oscillator with an analog filter, sound-wise.
It would ease the processor load a bit and add to the component count a bit, but I’d be more likely to fund a mixed synth than a straight digital one.
I think this thing is cool. I like that it has lots of knobs, a small form factor, and is affordable price. A great way for people with limited budgets to get a convincing analogue sound (analogue purists can suck it for all I care). If I had $200 to spare, I’d buy one.
Good luck, I hope this thing gets funded.
It will get funded.
Why? Because it’s cheap, lots of knobs and sounds like the SH, Juno and early Roland synths.
It’s a great way to learn and mock up some analog sounds at your breakfast table.
The NXP DSP has plenty of power to even make it polyphonic, no problem.
I just wanted it to be and sound like a 70’s synth.
96KHz 24-bit? Sure, go ahead and do it.
I’m not. Do a fast calculation and calc where the price of the synth goes and you’ll see.
I’m not. I rather put the money on great feeling knobs and 44KHz 16-bit.
By the way, there is a new soundclip with a classic sample and hold sound.
/Jan
Why did I choose to just use range and detune and always 3 oscillators?
Supersaw 🙂
It doesn’t have to be used for that but it’s a great synth string sound.
Once upon a time, I paid for an interesting sounding synth that was “in development”. I never got my synth or a refund, so best of luck to you, but I’m never going to pay for something that doesn’t exist again.
$200?
Thank you very, very much for the answers Jan!
I hope you can see why I was asking – I’m interested in this indeed but as others have said, I’m hoping to hear and see more before committing!
Best of luck with the campaign and thanks for answering!
Ok, another VA, another Meeblip… and where’s the inovation?
aq> Sounds like you have the knowledge to build it your self?
If so then go ahead.
The dsp-G1 is for people who cant and dont want to spend $$$ for a VA.
I only see some similarity with Meeblip project, where the classic SE was available for about $150, the micro one was and should be available in the near future again for less than $50 (where you can make your own enclosure with full controls), and now the brand new Anode is for about $130. I think it’s really cheap, catchy project in the same range as your dsp-G1 is conceived. And sure you can make your own customizations in the Meeblip synth model if you want. Now I see the pleasant inovation of Anode using pretty nice analog filter which beefs up the sound and simplification in control.
There is nothing wrong with the MeeBlip and the analog filter sounds great.
But since it doesn’t run on dsp hardware (ATmega1284) it is very limited in bit resolution and the filter cant be upgraded without a hardware modification.
My filter exists as dsp software and changing it takes only seconds with a USB cable.
My DCOs are mathematical models of Sine, Tri, Pulse and Saw oscillators.
There is nothing stopping it from being a rompler or even processing external audio through the dsp.
Also just a software mod.
Sure, this is the same at Meeblip SE which is btw named “hackable digital synth” meaning that everything is modelled using ATMega processor and can be reprogrammed as you want. The only drawback is that this classic SE version with complete surface is now out of production and it’s partly surpassed with Anode version. Brand new “micro” version of classic SE should be produced in mid-2014. See https://www.synthtopia.com/content/tag/meeblip-se/ or http://meeblip.com/
It’s not a matter of being able of building one, but it is just yet another virtual analog synthesizer, the world is plenty of them (even in the DIY world), and nowadays it is even easier than a few years ago to buy one real analog synth.
Where’s the point to build another VA synth and even asking money to fund this project?
Because of some requests I got, the initial series will include a noise source, a HPF and an arpeggiator.
I also secured delivery of 1000 Alps potentiometers today.
That is a really good potentiometer.
I suppose you only need to sell 25 of these to meet your funding goal, but I have to say I think you’ll get there a lot quicker if you improve your sales pitch. Eg:
• A much much prettier picture/3D model of what it might look like.
• An engaging video of the prototype in action, talking through why it’ll help people’s musical process.
• Having a reasonably representative range of amazing sounding sounds/melodies, presented conveniently with embedded player, not links.
• Rewrite the copy with an emphasis on benefits not just lists of features. Why is it unique, exciting, essential?
One of the few analog synths I’ve used from that era was a Juno 106. I did enjoy the sounds I could get from it. Also, I’ve been to a couple of synth jams recently where I tried using the iPad and wished I had some knobs.
I’m coming around on this concept. I think Jan is right that it will be fully funded.
Another thing that would be good to hear in the demo material is how well it deals with aliasing and very very high note ranges.
I’ve put together the dsp-Gx, a more sketchbook oriented synth with a display and less knobs.
It still has the same dsp-G1 engine but a smaller package.
The best of all is the price, $120.
Today we released the dsp-Grack. A 19-inch 1HE rack version of the dsp-G1.
Besides the features of the dsp-G1 it has 12 realtime knobs, a 320×240 TFT display, true stereo output and onboard FX processor for delay, reverb etc.
Check it out.
http://igg.me/at/dsp-g1
Hi Jan,
I couldn’t resist the Gplug and just contributed 😉
It’s a bloody GREAT idea and I’m looking g forward to messing with it! lucky I have a BCR-2000 here to use with it!
Thanks Mckenic.
You will have bloody great fun playing with it.
/Jan