U-he hardware using FPGA?

[david.beholder]

You said FPGA is not made for tasks like u-he have. Kyra is a virtual analog synth... subtractive synth... which is what Diva, RePro, Ace are... Kyra is well spec'ed, has Osc FM, Ring Mod and 128 voices... seems like plenty of capability to handle a u-he synth.

And what analog synth Kyra is emulating?
What exactly a role of FPGA in soundgeneration of Kyra?

[pdxindy]

You said FPGA is not made for tasks like u-he have. Kyra is a virtual analog synth... subtractive synth... which is what Diva, RePro, Ace are... Kyra is well spec'ed, has Osc FM, Ring Mod and 128 voices... seems like plenty of capability to handle a u-he synth.

And what analog synth Kyra is emulating?
What exactly a role of FPGA in soundgeneration of Kyra?

Huh? It's not emulating any synth... Why does that matter?

In Kyra the FPGA is doing all the synthesis...

[Pictus]

The Kyra is very interesting!
Manuel Caballero is a genius!!
So the u-he guys!!!

[Elektronisch]

Uhhhhhhh this is super round-about. Creating a hardware synth that replicates your software synth that aims to be like a hardware synth

The idea itself is also backwards. You don't decide to make something with a hammer and nails because you just found out about them! You should remember that you have them so that in the future when a task requires them you can use them in that situation.

I am not interested in a hardware replication of Diva. Just make another hardware synth if anything, but I'm still not interested in that!

Well this is what Roland Boutiques, System 8, Virus A B C Ti, Roland JP8xxx ect. are.

[EvilDragon]

Yes, after years and years of developing you might get close. But somebody gotta spend that time, so it probably isn't happening for u-he.

[Urs]

probably not

[rod_zero]

I don't care if it is FPGA or not but I want Diva in hardware. I bought a System8 because it it's the closest thing avaible but I would sell it and get Diva if it every come true as HW.

[Elektronisch]

I don't care if it is FPGA or not but I want Diva in hardware. I bought a System8 because it it's the closest thing avaible but I would sell it and get Diva if it every come true as HW.

Doesnt basicly any subtractive hardware synth with 3 oscillators and 2 envelopes and 2 lfos comes close?

[sleepcircle]

Diva is more like several synths in one, because of all the different types of oscillators and filters there are. it would probably be fairly expensive to make

EDIT:

hahah!! imagine if it came in cartridge form, almost like a semi-modular thing, but just with different 'cartridges' like old mixer boards did, that you could only connect up in sequence.

EDIT EDIT: btw i am not actually speaking IN FAVOR of a hardware version of diva. the thought just occurred to me.

[Urs]

As apparently we need almost two years to bring our first Eurorack module into stores, I think a hardware version of Diva might become my retirement project (we have no garden to look after).

[Tj Shredder]

FPGA's strength is the ability to change the whole wiring of very basic components. If you have something like Diva with a fixed architecture this would be overkill... It could be interesting for modular approaches though...
There are FPGA based computers, you can have a FPGA to switch from an Atari into an Amiga including all peripherals...

I don't care if it is FPGA or not but I want Diva in hardware. I bought a System8 because it it's the closest thing avaible but I would sell it and get Diva if it every come true as HW.

Get a decent x86 linux computer, lock it into a cabinet with all the knobs connected to a Midi control board and let it boot into Diva standalone... Paint it nicely and you got it...
Way less hassle than reprogramming that into any other code...

[teilo]

What we really need is an FPGA standard architecture with a corresponding toolchain and OS support. This is happening in the high-end computation market, but it's nowhere close to coming to the consumer world.

The basic idea is:

• The FPGA lives as one or more co-processors in a standard x64 machine, either on PCIe cards or on the motherboard.
• The OS supports the FPGA with standard interfaces to reprogram them on an application-by-application basis.
• A standard development toolchain provides the ability to program FPGAs using higher-level primitives.
• Applications can bundle FPGA modules which the OS will automatically deploy and make available via standard APIs.
• The OS manages resource contention among applications.

I know a company (friends of mine) that is doing revolutionary work in this space. Their model is quite different and oriented toward transactional processing. But they also created the aforementioned toolchain which can bring FPGA programming to the masses.

Imagine what could be done in the sonic space with such tools. This would compete directly with Universal Audio's architecture, but would be open to anyone for much less money.

[Tj Shredder]

I don't see the musical necessity here. In the end, if I have an innovative idea for a new synthesis method, I can implement it in any Turing machine. Sometimes I just need to wait to get the processors which are capable to deliver the necessary computing power. Mostly that is faster than developing and learning for new platforms...
As far as I know, its not possible to implement analog circuits in FPGA, which could open some possibilities...
I suspect a FPGA is less efficient than a fixed processor with the same circuit. Its just more flexible...
Dedicated DSPs seem to develop slower than general purpose CPUs, but shaders in Graphics cards develop faster...
That would make much more sense to move the DSP code into graphics chips and I think some work is done already in that direction...
FPGA is the dream for digital archeology (recreate the old machines and switch from a Fairlight to a UPIC with the push of a button...)