Trueno: the world's smallest analog synthesizer

axoman 2017-12-07 22:08:29 UTC #21

As far as I'm aware, there is no such thing as a digitally controlled VCO, it is either one or the other as the "C" in VCO stands for "Controlled".

It cannot be both Digitally controlled (Digital), and Voltage controlled (Analogue) at the same time:

DCO = Digitally Controlled Oscillator
VCO = Voltage Controlled Oscillator

Nice idea though, and assuming they're not lying, it would certainly be nice to have a real analogue synth to plug in and use with a DAW having on on-screen interface. I will not be buying one, though, because it doesn't have an analogue output.

It feels completely and pointlessly idiotic to do that.
On the plus side, though, at least they spell "Analogue" correctly!

Really baffled by the lack of anlogue oputput. Could be designed by an Apple fanboy who thinks everyone else is dumb enough to buy audio-capable devices without an analogue headphone/output socket

SmashedTransistors 2017-12-07 22:14:13 UTC #22

A Voltage Controlled Oscillator can by synched by a Digital trigger to ensure accurate pitch. A la Juno.

Or phase lock loops can also be used.

http://www.jacobwatters.com/blog/a-different-kind-of-dco/

SmashedTransistors 2017-12-07 22:16:29 UTC #23

Whatever the technology used is... it lacks the tweakability of analog synths.

axoman 2017-12-07 22:23:27 UTC #24

I stand corrected, and very true, shame about the lack of analogue output though.

I would likely have bought one if it turns out to be real analogue and it had an analogue output that completely bypassed any AD or DA conversion.

axoman 2017-12-07 22:26:45 UTC #25

Also, even taking the computer aspect of it into account, I reckon the sampling should have been at least 48k, so that's another bad decision.

They should have designed it so that the sampling was done by the DAW, not the device, and use that money to give it an analogue output instead.

It would have had the best of both worlds if they had done it that way.

ricard 2017-12-07 22:39:05 UTC #26

Well, the website says the ADC 44.1 kHz which is good enough. (There was something about 22.05 kHz but that seems to be for the sample playback oscillator).

As for the VCO vs. DCO debate, especially the term DCO doesn't really have a precise definition. If one is thinking of the type of oscillators in a Juno, the pitch is digitally controlled, but the waveform is controlled by a voltage. Roland have labeled it DCO on the front panel, and I'd agree, although one aspect of the waveform is controlled by a voltage (the slope of the sawtooth), the pitch is digitally controlled, the point being that pitch accuracy is way more important than waveform accuracy.

So I'd definitely say these are digitally controlled, as the most prominent parameter of an oscillator is its pitch, and it's controlled digitally.

However, if you hook a classic VCO to a DAC, you also have a digitally controlled oscillator ... so is it a DCO now ...? In that case, a Prophet 5 filled with CEM 3340 VCO chips would be considered DCO's as they are in fact controlled via a DAC from the instrument's microprocessor.

I would say that if the pitch is controlled in such a manner that it is precisely controlled by digital means (e.g.. a Juno oscillator) it's a DCO. If somewhere in the chain there's an intermediate stage with a control voltage that is susceptible to various forms of drift (temperature, supply voltage, etc), it's a VCO. And then queue something like the Oberheim Matrix-6, where the oscillators are essentially DCO's - the pitch is accurately controlled by dividing down a master frequency with a digitally set divisor - but the master oscillators are free running, and hence there's a certain randomness (and possibly modulation, I can't remember) invoked on an analog level. To my mind, they're still DCO's, but with a twist to emulate analogue behavior.

SmashedTransistors 2017-12-07 22:44:30 UTC #27

The master frequency/accuracy of the digital trigger (equivalent to a sampling rate) is in the MHz range for the junos. That's why there is almost no aliasing with these DCOs. The Deepmind12 use this technology.

SmashedTransistors 2017-12-07 22:50:39 UTC #28

As the master oscillator is modulated by an analog source, this is a voltage modulated digital oscillator.
(some nice Crumar Italian string machines used up to 3 LFO modulated master oscillators).

The only case of a totally voltage controlled digital oscillator i know is the RMI Harmonic Synthesizer (which used digital walsh functions, not sine waves, hence the specific overtones but without aliasing).

Maybe the wavetable/additive oscillator is derived from such a concept...

ricard 2017-12-08 05:34:42 UTC #29

Yes, that's a very accurate description ... although I've never seen it used before.

Could be, but I would have expected that Trueno would have used that heritage in their marketing in that case, as the RMI name would carry same weight as it was used on Jean-Michel Jarre's Oxygene.

Interesting about the RMI design. I've never known much about how it works. Although by that description, thinking about it, aren't there modules for modular synths which are sample playback engines or other forms of digital oscillators that are voltage controlled?

EDIT: Walsh functions - by the looks of it, it's a glorified (well, mathematical) sum of square waves at octave ratios. Similar to he way the Poly-800 or Bit One create their sawtooth waves. What comes to my mind though is the ultra rare Davolisint which had switches rather than sliders to enable the different square wave octaves. So that is essentially the extent of the 'Harmonic Synthesis' of the RMI then?

jaffasplaffa 2017-12-08 08:25:43 UTC #30

22.05 is the nyquist frquency of a sample rate at 44.1. All frequencies above 22.05 hz will create aliasing, that is the essence of sampling theorem. Which means ALL frequencies above 22.05 will be folded back into the audible signal range and they sound horrible. And these frequencies will sound horrible cause they have nothing in common with the original signal, because they are folded back. Again, this can be fixed by using oversampling, aka anti aliasing. Here is imo a proper explanation of aliasing and oversampling that makes sense to most people:

jaffasplaffa 2017-12-08 08:42:48 UTC #31

Regarding the DCO, VCO:

I dont understand that discussion. DCO's are emulated in digital world, VCO's are emulated in the digital word. Even though the terms might come from analogue world, it doesnt mean that they dont exist in the digital world. So I dont see it as a measure of the synth osc's might be analogue or not. Everything can be emulated.

axoman 2017-12-08 09:45:52 UTC #32

Well, at 44.1K I have to wonder what market they're aiming it at, cause it's generally accepted since as far back as I can remember, that 48K is the sweet spot beyond which you get diminishing returns. I've always been in agreement with it too, cause I've obseved it first-hand.

For a product such as this, 48K should have been the absolute minimum, after all, the job of the sampling on this device is to capture the "analogue" aspect as faithfully as possible. 44.1 is indeed very capable, but I happen to know first-hand that 48K is notably better at it. The sampling should have been left to the DAW, that way the device would only ever have been limited to the processing power of the computer running it.

Regards the VCO/DCO thing, personally, as long as the oscillators themselves are true analogue then I'm not too bothered about that. I love the Roland JX-8P, intend to own one some day, and that's a completely digitally controlled analogue.

So in a nutshell, for me, a synthesizer can only truly be said to be analogue if it's signal path is 100% analogue. And the unfortunate thing is, this device falls flat on it's back even if that were the case internally, because there's no way to get an analogue signal out of it without going through conversion.

It's like they're saying, here's an analogue synth, but sorry, it doesn't actually output analogue
So my own personal conclusion: I think it's a good overall idea that was poorly thought out.

ricard 2017-12-08 22:12:50 UTC #33

Yes, of course, but when they mentioned that the digital oscillators output an "endearingly lo-fi 22.05kHz in 8 bit" I interpreted it as being a sample rate of 22.05 kHz, since a sample rate of 44.1 kHz is hardly 'lo-ff'. It could very well be that it's just the "8 bit" thing which is "lo-fi", and the sample rate is the same as for the ADC.

As for the analogue signal path ... well, if it did output an analog signal, the first thing one would do would be to feed it back to a sound card so that it could be digitized and recorded in a DAW, So having the A/D converter built in makes sense, even though it does make the device less useful for some.

The main point for me though is that if it were just a pure software implementation, there'd be no point in having it on the end of a USB connection, one could just run it as a softsynth plugin and dispose of the hardware completely. So either it's a clever marketing gimmick, or there really are some key analog signal path components in there, the oscillators and filter being the most important ones to keep in that domain.

For me personally it falls flat for completely different reason: it has no user interface, so just as with a softsynth you're left with a virtual control panel on your computer screen, which for me doesn't cut it, whatever hardware its running on. But that's a very personal opinion. It'll be interesting to hear what people say about it after having heard it for real.

axoman 2017-12-09 11:52:47 UTC #34

Being analogue, I would have connected it to an analogue EQ and analogue recorder - not a soundcard.

Analogue equipment without analogue output has to be the most ridiculous thing I've heard in years (and that's saying something). It comes second only to Apple's feeble attempt to kill-off the analogue headphone jack (something it will never succeed in doing despite the effort and brain-washing they will put into doing so).

I can understand that their reasoning for this product could be that being designed for a computer (digital), there was no need for an analogue output. I could never agree with such reasoning though, it is completely ridiculous as it costs literally pence to add a analogue headphone out jack that bypasses any conversion and simply supplies it with a feed direct from the analogue circuits' output. That would have been no biggie on a product that is priced as this one is. It's every bit as ridiculous as it sounds. It must be the only "analogue" synthesizer on earth that doesn't output in analogue!

That said, I don't like to knock startup companies, I do respect the general idea despite me seeing this particualr product as a dud, and I do wish them well with their product and their venture. They screwed-up on the spec of this one, but if it sells well, who knows, perhaps they'll release a better one with a direct analogue output next time, one that bypasses any conversion and keeps it completely in the analogue domain.

thetechnobear 2017-12-09 13:20:10 UTC #35

I’m open to new ideas, and I applaud anyone trying something different, esp. when they put their money where their mouth is

I’d want to hear some independent hands on reviews before being quick to judge.

Apart from the sound, I’d also like to know some technical details (or perhaps i missed)
Particularly , what’s the protocol it talks to VST/app , is this going to be open sourced?
I’d have liked it to have appear as a usb class compliant audio and midi device, this would have meant it would be compatible without additional software, so something like Axoloti/rPI could have used it.
(Admittedly, for audio this would potentially necessitate an aggregate device but still)

Anyway, I’m not really in the market, but I hope it’s a great product, they find a market and it does well.

axoman 2017-12-09 14:23:29 UTC #36

I just realised, looking at the specifications, that I couldn't use one even if I wanted to
There's no Linux support, there's no excuse for that these days, so I wouldn't support the project anyway.

SmashedTransistors 2017-12-10 21:19:13 UTC #37

True for octave ratios 2, 4 ,8, 16,
but other "harmonics", called sequences, are not regular squares.

see CAL5 and CAL6 for example

Here is a patent from Allen organs, the company that owned RMI...
https://patents.google.com/patent/US3878749A/en?q=allen&q=organs&q=walsh

A bunch of counters and exclusive or can be used to generate them.

In the case of the RMI, the "digital" signal was clocked by a VCO, this design prevents aliasing, but not "imaging", i.e. overtones caused by the steppy character of the signals. This gives the RMI synth its particular sound quality.
I made an object syncAlias capable of creating such overtones.

It is probable that their wavetable oscillator is fully digital... Hope we will get more technical information about this.

jaffasplaffa 2017-12-23 11:06:17 UTC #38

Here are a few screenhots of the internals(not mine):

The VCO1 stuff down in the corner is a bit hard to see what is going on.

What does those who know about analog gear say about it?

SmashedTransistors 2017-12-23 13:46:32 UTC #39

074 are quad op amps. The 051B are analog multiplexers. Maybe it is more c8mmutation controlled than voltage controlled..

servandisco 2017-12-29 19:46:24 UTC #40

I hear aliasing in the pulse 50% demo from their page. Definitely it could be way cooler if it was more hackable and more stand-alone oriented. I think..