More Vintage Emulations!!!

Man-Machine · Post by **Man-Machine** » Tue Feb 01, 2005 6:19 pm

The only reason why I bought a Creamware card was because of the outstanding vintage emulations alone. The Pro-Tone effectively replaced my old dead Pro-Tone which I used in almost every recording. I also bought all the other emulations after that. They all complement each ther real nicely with their unique filters and all. I don't know how hard it would be to legally emulate Rolands, Korgs and Oberheims but these synths have very demanding and unique filters. The Saturn freeware synth has Oberheim and Roland emulated filters in a generic VCO -> VCF -> VCA format, unfortunately it has not been updated in a while.

If these synths cannot be emulated for legal reasons how about a Solaris type of synth with even more filters (Roland, Oberheim, Korg) Basically a software synthesizer like Solaris with more filters. The Alesis ION/MICRON hardware virtual analog are doing this with a variety of different classic filters...

_________________
L8ter Oscill8ters!

[ This Message was edited by: Man-Machine on 2005-02-01 18:24 ]

wayne · Post by **wayne** » Tue Feb 01, 2005 11:21 pm

- that's a different point of view, M-M!

Most folks around here say "no more VA" and "more innovation a-la-Flexor"

Have you checked out Spacef's Blackbox, notably Synthetic Dx ?

If you want some new filters (and oscillators & hosts) this is worth a look - not an emulation of anything, but a very flexible architecture.

steffensen · Post by **steffensen** » Wed Feb 02, 2005 4:41 am

FleXorius!

Mr Arkadin · Post by **Mr Arkadin** » Wed Feb 02, 2005 6:50 am

i'm actually a bit of a vintage whore too - OB-Xa someone pleeeeease! i'm all for new synthesis, but here's what happens:

Minimax is released - evreyone seems to have that.

Python is released - about three people seem to own that.

Also, to be honest, when i hear these so-called 'different' synthesis techniques, well a lot of them (like granular) just sound like a regular synth sound chopped up a bit and put through a distortion box, ie. not that different at all - even the Hartmann Neuron thing i tried and it was like a vector/pad/sampling thing, but i never hear a sound on a record now and think - ooh, i have never heard a sound like that before in my entire life - must be a completely new synthesis technique on the market. Oh well.

Mr A

[ This Message was edited by: Mr Arkadin on 2005-02-02 06:50 ]

Man-Machine · Post by **Man-Machine** » Wed Feb 02, 2005 2:05 pm

He, he, funny you mention that. I was just researching and reviewing these additive/resynthesis soft and hardware synths and I don't find too much difference in their methods. It seems to me that you pretty much only need one of them to cover those sounds.

On the other hand, you get so much more variety in analog synthesis (not just subtractive synthesis) than any other sythesis method, IMO. The analog components it's what makes them unique, all those different types of oscillators and filters, the erratic way they behave, drifting, etc. Two of the same model never sound the same. You just don't get that with other synthesis methods because they mostly use digital components to do it and sound too precise (but they shouldn't have to necessarily...)

All these synthesizers are dying now and we should take advantage of current technology to try to emulate them the best we currently can. I think Creamware does the best job so far in accomplishing this so I hope they keep on crancking them out. I know I'll buy them all...

Frosty · Post by **Frosty** » Thu Feb 03, 2005 3:00 am

I'd love to see my Kitten again (Octave-Plateau )

sharc · Post by **sharc** » Sat Feb 19, 2005 4:02 pm

Tell you what ...If someone is willing to send me an SDK board I'll be only too happy to spend a few months recreating some classics. I've got a few here that I'd love to see - Oxford OSCar, Octave CAT(and Kitten!!), Transcendent 2000 to name but a few. I'd also like to see/do an emulation of my sci drumtraks.

Shayne White · Post by **Shayne White** » Sat Feb 19, 2005 5:52 pm

If you wanted to do anything REALLY complicated you'd need the atom-creating software....

at0m · Post by **at0m** » Sat Feb 19, 2005 9:45 pm

A couple months to get to know the SDK. Then a couple of months studying the synth's typical aspects (osc, filter, env, porta mode,..), then translating into SDK and debugging... there is huge possibilities in SDK. It's impossible to find all the tricks that you can see in various 3rd party releases, cos then you know as much as all of them.

It'll take longer than one might suspect - and then you have something that already exists.

For us mortals there's still the modular, that can mimic quite a few options :> I'd like to give Flexosaurus as example. It has a lot of the features of the Minimoog, but we don't have these osc or filters for example. Instead we inserted shapers, filters with osc modulation etc. It took about an afternoon of research and patching, and it's something reasonably new with prooven features of the Minimoog.

_________________
More has been done with less

[ This Message was edited by: at0m on 2005-04-07 14:22 ]

CroNiX · Post by **CroNiX** » Sun Feb 20, 2005 12:23 am

Also, even two of the exact same model of analog (true analog) synths will not sound *identical*. The reason being is that they are made up of resistors and diodes (etc) which all have tolerances, meaning they are within a range of values. So, two 2K Ohm resistors will not give exactly the same value. Throw a couple of hundred components with varying tolerances and you have a slightly different sound.

It would be kind of cool if when creating a digital synth, they would emulate the tolerances as well to give a unique sound like these old classics. Digital synths or recreations will always sound exactly the same because they are precise.

ChrisWerner · Post by **ChrisWerner** » Sun Feb 20, 2005 5:56 am

On 2005-02-01 23:21, wayne wrote:
- that's a different point of view, M-M!

Most folks around here say "no more VA" and "more innovation a-la-Flexor"

Ok planetarians, let us talk about a new synth, will we?
What kind of synthesis is in your mind, that could be worth to give it a try?
I prefer a newer synthesis, I ve got something in mind, based on the good old Waldorf (RIP) wavetable in a combination of something else, but the idea is not worked out yet. I think on a spectral wavetable synth, some kind of additive wavetable synthesis.

I would call out for a team worked Z synth, any interest? Tell your ideas and maybe we can come out with a patch first and then a sdk device. Should we go for it, I would like it.

astroman · Post by **astroman** » Sun Feb 20, 2005 8:45 am

On 2005-02-20 00:23, DementiaT wrote:
...when creating a digital synth, they would emulate the tolerances as well to give a unique sound like these old classics. Digital synths or recreations will always sound exactly the same because they are precise.

it's a waste of effort as it's only relevant to someone who also has the physical counterpart of an emulation

Afaik a lot of digital VA synths use random variations of parameters - that's not difficult at all to implement.

cheers, Tom

8-Bit · Post by **8-Bit** » Tue Mar 01, 2005 11:58 am

On 2005-02-20 00:23, DementiaT wrote:
Also, even two of the exact same model of analog (true analog) synths will not sound *identical*. The reason being is that they are made up of resistors and diodes (etc) which all have tolerances, meaning they are within a range of values. So, two 2K Ohm resistors will not give exactly the same value. Throw a couple of hundred components with varying tolerances and you have a slightly different sound.

It would be kind of cool if when creating a digital synth, they would emulate the tolerances as well to give a unique sound like these old classics. Digital synths or recreations will always sound exactly the same because they are precise.

Exactly0----ex-schmactly. Who cares if hardware analogs sound 'slightly' different due to all those factors. i believe the main idea here was for someone to digitally emulate the synth.

If you say "tuna sandwich" 300 times, i can almost bet you that it wont' come out the same. Does that change the fact that your voice is still the 'same' and the words are still the 'same'? Its not the little differences that they're talking about. Its the synth itself.

TUNA SANDWICH!
8

Man-Machine · Post by **Man-Machine** » Wed Mar 30, 2005 12:57 pm

I see what DementiaT is talking about. It happens on acoustic instruments like let's say my Martin D-28 guitar. People out there have the same instrument but they don't sound exactly like mine because of number of factors like wood, age, neck, etc. It gives the owner a uniqueness feel. That kind of stuff also happens with analogue synthesizers, specially on something like the Minimoog. They're all different sounding. Inspite of all these differences between same model analogue synths, I'm more interested in the small nuances within one analogue synth itself that make it feel more like an acoustic instrument. Take for example a snare on an 808, it never sound the same twice. I fortunately (or unfortunately) can tell a difference now. I like that. I makes things less precise on an already precise, sequenced, computerized music I make. That's why samples will never give full justice for me (or for that matter, digital synthesizer emulations in the current state...)

at0m · Post by **at0m** » Wed Mar 30, 2005 1:08 pm

I'm more interested in the small nuances within one analogue synth

In modular you can try with adding very small amounts of random modulation on various parameters, the patch will sound much more alive, less sampled or static. With prefab synths it's more difficult, too much MIDI CC# will eventually hog the driver.

BingoTheClowno · Post by **BingoTheClowno** » Wed Apr 06, 2005 10:42 am

On 2005-02-19 17:52, Shayne White wrote:
If you wanted to do anything REALLY complicated you'd need the atom-creating software....

Can you elaborate on this, is it available anywhere?

astroman · Post by **astroman** » Wed Apr 06, 2005 4:36 pm

generally not - but of course anything might be subject of negotiations... if you have a big name, a reputation, a lot of cash, a hyper brilliant idea or any combination of the afforementioned...

background: if you write some processing routines in SHARC DSP assembly and compile them with Analog's original tools, it doesn't help much.
The code will not run in an SFP system because each and every program module has to be passed through a special 'loader' and as such has to be encrypted, as ARE all existing building blocks (atoms) even in SDK.

It's the core part of protecting CWA's intellectual property, as otherwise it wouldn't need much more than the wink of aneye to disassemble the processing part of the complete system, make it relocatable and upload it on arbitrary SHARC based boards.

that encrypting tool is treated like a kind of holy grail...

cheers, Tom

Shayne White · Post by **Shayne White** » Wed Apr 06, 2005 6:35 pm

Yeah, I suppose handing out that encrypter to everybody would essentially make all the DSP modules open-source, and they don't want to do that. Of course, you can always do what Adern did, and that's build your custom algorithms with math modules....

Shayne

symbiote · Post by **symbiote** » Wed Apr 06, 2005 7:50 pm

I'm pretty sure they sell their dev system, or at least they used to, but in the same range as the other similar types of development system, in the tens of thousands of dollars, to companies who want to do exactly that, develop algorithms and run them on arbitrary SHARC boards, usually sitting in hardware units that cost more than a few scope boards together even =P.

For example, I've tinkered around a bit on a similar dev system made by a company called LyrTech, the hardware side was pretty nice, consisting of a TI DSP and Xilinx FPGA sitting in a box outside a computer, that you could program with MATLAB/Simulink (ugh gag choke die.) The TI DSP cranked out around 1-1.2GFLOPS, the FPGA is hard to measure theoritically but probably a bit faster. It was pretty neat, but it was worth 15000$, including the compiler and tools for MATLAB, but not MATLAB itself, so add another few grands to get some of the sigproc/comm toolbox.

Of course, once you have functionnal code, you can just buy 1000 TI DSPs and FPGAs, plug them on a PCB, feed them the code you've compiled, and you can get something nice going and sell a bunch, and that's the idea, but it's pretty expensive, that's why Creamware won't give it out to you for free. And lemme tell you, as awkward as the SDK is, it's a hell of a lot better than Simulink, both on the circuit design and interface design side.

Seriously tho, if people just want to code algos and process sound realtime with them, you can get Cycling '74's Max/MSP for PC and Mac for pretty cheap now, it runs natively and you can compile your own code in a few different languages for it.

BingoTheClowno · Post by **BingoTheClowno** » Thu Apr 07, 2005 10:58 am

Thanks Tom, Shayne, it makes sense to me now.

Symbiote, I hear the recently released Reaktor 5 is capable of compiling modules to VSTs too. This is a major change in Reaktor and it probably represents NI's response to some of the Max/MSP software competitive features.

Very exciting times in the music software industry!

[ This Message was edited by: BingoTheClowno on 2005-04-07 12:00 ]