Diva Vs. Real Analog

[leggie]

Diva has given me sounds that I could only dream of years ago when as I could not afford the hardware required apart from one lucky time I brought a korg mono/poly and sold it a year later.

Im totally inspired to make and enjoy those electronic sounds from my younger years and loving every second of it. Since i got Diva my other VST's are collecting virtual-digital dust

Diva is a brilliant work of engineering imo and how exciting it will be to see it evolve.

Best throw in a problem i have with it to not sound like a fanboih, erm.....erm.... i dont like red synths, but that's sorted now thanks to IrionDaRonin

http://www.kvraudio.com/forum/viewtopic ... sc&start=0

[Gamma-UT]

If it were the case, then I would definitely be very interested to find out who formulated it and for what reason. I always thought mathematics worked in the opposite direction.

Indeed. I'd tend to see mathematics as a tool of analysis (with prediction and description both falling under 'analysis'). I don't necessarily see it as a tool of invention ("hey, I'm going to invent a replacement for the gas lamp, pass me my maths textbooks and i'll calculate up something based on hot bits of wire")

It can work either way. Sometimes pure maths does influence engineering. A lot of wireless communication systems rely on the maths of imaginary numbers. I could mention another more famous example but that would probably send the entire thread to HPC.

But a lot the analogue synths we know and apparently love were intended as emulations of physical things not abstract notions. When Bob Moog first described his approach in the late 1950s, his thinking was largely geared to modelling the timbres of real-world sounds and instruments. His problem was that the people with the money were the electroacoustic, musique concrète crowd who basically laughed at him and told him not to let the door hit him on the way out for proposing something so obviously primitive when they had the magic of tape. From a technical perspective, although voltage control was the breakthrough for Moog, there wasn't a whole lot of mathematical theory involved in the development.

Twenty years earlier, the Hammond Novachord was trying to emulate real instruments - and got nowhere because the sounds couldn't compare. Buchla got more success when he started out by going completely leftfield.

[spaceman]

It can work either way. Sometimes pure maths does influence engineering. A lot of wireless communication systems rely on the maths of imaginary numbers. I could mention another more famous example but that would probably send the entire thread to HPC.

It may influence engineering but I can't think of many(or any) cases where it was the model for an invention. It's not as if the engineer looked at imaginary numbers and then decided to do build a communications network that models them.

[hakey]

Emulating something has an intrinsic notion of intent.

Intent = "nebulous shite"?

Do you need it rephrased or something?

Emulation is specifically deliberate, copying can be accidental. Does that help with your problem?

No problem. I agree with your definition - emulation obviously implies intent.

I guess you didn't see it, but when I made the same argument a while back, I was told that intent and purpose were "nebulous shite" - the implication being that any definition of emulation that relied upon intent was itself nebulous.

[SJ_Digriz]

..It's just an example of how the idea that digital synthesis amounts to putting analog circuits into code is false.

Both digital and analog synthesis are ultimately "emulations" of "real" synthesis, which is abstract (mathematical).

I think I see where we were not communicating. You are parsing the words down far to low a level. This is a discussion of ANALOG vs DIVA. It is also, a bit tongue in cheek.

You somehow leaped to the idea that we are saying "Digital Synthesis = Software Modeling". Or, that all synthesizer software development is based on circuit modeling. Or, some combination of these concepts. I don't believe anyone was saying that. I know I wasn't.

[jupiter8]

Analog and digital synthesizers are both imperfect "emulations" of abstract mathematical synthesis.

What 'abstract mathematical' synthesis does a PPG-style scanned wavetable 'emulate,' then, for example?

The original idea was to get rid of the filter IIRC. Analog filters are expensive and if you can replace that with scanning thru a table of increasingly filtered waveforms you could build a cheaper synth. Not that PPGs were ever cheap but i believe that was the general idea to begin with. Turns out it wasn't such a good idea as it initially seemed as they added filters to later models..

Have no idea what abstract mathematical synthesis that would be. The original idea was just a cheaper way to do a basic subtractive synth i guess.

[Gamma-UT]

It can work either way. Sometimes pure maths does influence engineering. A lot of wireless communication systems rely on the maths of imaginary numbers. I could mention another more famous example but that would probably send the entire thread to HPC.

It may influence engineering but I can't think of many(or any) cases where it was the model for an invention. It's not as if the engineer looked at imaginary numbers and then decided to do build a communications network that models them.

When Hancock and Lucky and others developed quadrature amplitude modulation the hardware didn't exist that could implement it - it was about a decade later that anyone started building on their entirely theoretical work. The Viterbi algorithm - used by every cellphone in production and in DSL modems - was developed in 1967 to "help prove an asymptotically optimum upper bound on the error probability of convolutional codes" (The Viterbi Algorithm: A Personal History, G David Forney, http://arxiv.org/pdf/cs/0504020v2.pdf). Again, it took years before it was use in real-world systems. Andrew Viterbi did later cofound Qualcomm and make a ton of money out of it.

[declassified]

Wow, Urs must be really glad about all this commotion! It's really good for making a product popular. When I saw this thread "Diva Vs. Real Analog" with 65 pages, my first thought was: Wow, it must sound pretty darn analog if it takes people 65 pages of discussion to figure out how close it is

[whyterabbyt]

The Viterbi algorithm - used by every cellphone in production and in DSL modems - was developed in 1967 to "help prove an asymptotically optimum upper bound on the error probability of convolutional codes" (The Viterbi Algorithm: A Personal History, G David Forney, http://arxiv.org/pdf/cs/0504020v2.pdf). Again, it took years before it was use in real-world systems. Andrew Viterbi did later cofound Qualcomm and make a ton of money out of it.

But doesnt that paper state the opposite in places; that the applications came after the theory, that the theory wasn't developed with the purpose of creating a new invention.

the Viterbi algorithm for convolution codes... came out of my teaching... I found information theory difficult to teach so I started developing my tools...

Nobody though that it had any potential for practical value

Andy has always said that Jerry heller was the first person to realize that the VA might be practical.

[trimph1]

The Viterbi algorithm - used by every cellphone in production and in DSL modems - was developed in 1967 to "help prove an asymptotically optimum upper bound on the error probability of convolutional codes" (The Viterbi Algorithm: A Personal History, G David Forney, http://arxiv.org/pdf/cs/0504020v2.pdf). Again, it took years before it was use in real-world systems. Andrew Viterbi did later cofound Qualcomm and make a ton of money out of it.

But doesnt that paper state the opposite in places; that the applications came after the theory, that the theory wasn't developed with the purpose of creating a new invention.

"the Viterbi algorithm for convolution codes... came out of my teaching... I found information theory difficult to teach so I started developing my tools"

"Andy has always said that Jerry heller was the first person to realize that the VA might be practical."

It does appear as if he did say both things in that paper. I'm wondering if he actually did intend that though or was it some kind of mistake though. Either way it is a good read.

Back to Floor Wars.....

[Gamma-UT]

The Viterbi algorithm - used by every cellphone in production and in DSL modems - was developed in 1967 to "help prove an asymptotically optimum upper bound on the error probability of convolutional codes" (The Viterbi Algorithm: A Personal History, G David Forney, http://arxiv.org/pdf/cs/0504020v2.pdf). Again, it took years before it was use in real-world systems. Andrew Viterbi did later cofound Qualcomm and make a ton of money out of it.

But doesnt that paper state the opposite in places; that the applications came after the theory, that the theory wasn't developed with the purpose of creating a new invention.

the Viterbi algorithm for convolution codes... came out of my teaching... I found information theory difficult to teach so I started developing my tools...

Nobody though that it had any potential for practical value

Andy has always said that Jerry heller was the first person to realize that the VA might be practical.

Actually, that was the point I was making: that the application came out of the maths rather than someone using maths to work out whether some real-world approach was plausible. The Viterbi one is arguably a funny example as the same person devised the theory and then managed to apply it.

Hedy Lamarr and frequency hopping is arguably an example of the application driving the maths. Or encryption systems such as RSA and AES.

[whyterabbyt]

Actually, that was the point I was making: that the application came out of the maths rather than someone using maths to work out whether some real-world approach was plausible.

ah, sorry. too much multitasking on my part, losing track.

[zerocrossing]

I think felt would probably be a more appropriate material here than either the original foam, or leather

Felt is wool... wool is static electricity-o-rama. Might not be the best choice for electronics?

[spaceman]

Actually, that was the point I was making: that the application came out of the maths rather than someone using maths to work out whether some real-world approach was plausible. The Viterbi one is arguably a funny example as the same person devised the theory and then managed to apply it.

But if I'm allowed to be a bit pedantic here, the algorithms in question serve a bigger system. No one ever built a quadrature amplitude modulator for the sole purpose of modulating quadrature amplitudes. Again, its processing serves a bigger system and a more complex problem. Whereas someone here tried to imply that sound synthesis was entirely based on maths conceived with the sole purpose of synthesising sound.

[Meffy]

I think felt would probably be a more appropriate material here than either the original foam, or leather

Felt is wool... wool is static electricity-o-rama. Might not be the best choice for electronics?

Wool itself tends not to accumulate a static charge as it's strongly hygroscopic -- it absorbs water from the air, and being made of fibers wool has a huge surface area. The thin coating of moisture on wool fibers conducts away and dissipates static.

What wool does is to raise a static charge on amber, hard rubber, and similar materials when rubbed against them (not sure whether the plastics used in synths and MIDI controllers behave the same way). When not moving against such a material it won't generate static electricity. Any tiny charge that plastic might accumulate from small wiping motions would most likely dissipate in the damp wool, so I think it would be safe. To be really sure I'd want to do a few experiments at various real-world humidity levels.