Thanks for the comments! It has been fascinating to see the responses, even the, um, wackier ones. The generally positive feedback we've had from the wider audience of synth fans has been exceptional.
Plugin designers can't please everyone, of course; we've had nearly all positive feedback about the JPS Harmonic Synthesizer, and some of which has been utterly overwhelming to me, including from a couple of industry players (just this past Friday, a producer who has worked with the likes of Bjork, Editors and Tears for Fears wrote to me saying it's one of the best synth plugins he'd seen in years).
The fact is I can't argue with someone who actually owns an original machine—like Stephen above does—as to whether it does or doesn't sound exactly like his, as I know I'd lose that argument. However, that's simplistic and perhaps even naive argument to make anyway, imho. So I will defend it against some of the attacks above, because the truth is not so simple as an arbitrary "Does it sound exactly the same in a side-by-side comparison? No? Then it must be crap" comment. The reality is more complex, and it seems that a few people have become obsessed by notions of literal accuracy over usability and pure enjoyment. Whether the JPS HS or any other emulation softsynth doesn't sound exactly like the product they are based on doesn't change the fact it might produce a great sound worth having in its own right! And the truth is all such virtual softsynths invariably don't sound like exactly like their real-world counterparts, regardless of what their marketing guy writes.
It's important, then, to note I called it the JPS Harmonic Synthesizer rather than the RMI Harmonic Synthesizer. I'm aware it won't sound exactly like an original machine but I can confirm it was developed using the exact technical specifications as the starting point, unlike any previous attempt at an RMI HS.
The wave generators and presets of the JPS are, to the best of our knowledge, what the original machine was designed to have produced, disregarding any otherwise unintended artifacts created elsewhere in its electronics. I don't need to have one next to me to know what the intended output of the harmonics and presets was. There are compromises, some behavioural gaps like potentiometer curves we had to fill in with best guesses based on practicality or what we thought just sounded good, and some other deliberate choices made to suit a modern sensibility and others to extend its usefulness (like the release trigger mode) for commercial viability and getting an RoI. So while the Oxy IV/V patches do use the harmonic slider settings Jarre uses as seen in various video and photo resources, the RMI potentiometer curve might have been just different enough that a side-by-side comparison of the same settings for the DHG's and presets may not sound entirely alike; but result I don't believe will be that far off the mark in raw harmonic terms. Fortunately the sliders move so either can be tweaked to sound more like the other!
So the JPS HS is both technically modelled on, and more generally inspired by, Rocky Mount's. It's not a like-for-like clone of an actual HS, I have made no suggestion or claim that it is, nor do I apologise that it isn't. One can talk about "modelling the electronics" as suggested above like it's some necessary gold standard and anything less is substandard, but in reality such modelling is largely, again, just marketing BS.
It seems to me though, that some people would prefer to be upset that something that sounds pretty damn close to what they need for just $79 doesn't sound exactly like the something they want but they'll never have: the alternative is they can wait until one of the few still working pops up on eBay for £7,000+!
But unless it's one of perhaps as few as a dozen worldwide that are still actively used and serviced, it'll have an SNR so low you can drop the S entirely, and spend another 7k restoring it and praying the FedEx guy doesn't take too many speed humps on its journey home.
So it can be a tad disheartening that a few people are instantly dismissive of it simply because one demo doesn't sound exactly like the original recording of Oxygene part IV.
Because of course it doesn't. Jarre created that in a kitchen in 1976 using the fairly basic tools he had to hand, all of which influenced the sound you heard on the album. There's a certain amount of AC hum from the synth itself, there's the choice of tape it was recorded to, there's whatever compression and eq and particular peculiarities of his Electric Mistresses, and harmonic filtering though track bouncing etc, etc. Real synths sound different as virtual synthesizers. Unless they're sampled. It's also likely that no two Jarre performances used exactly the same settings anyway once those performances were actually live. But if you add models of his exact tools to the chain, I see no reason that one can't make it sound as close as you want. Or you can free yourself from those kinds of self-imposed, artificial limitations and use it to create something entirely new in whatever way you want.
When I first posted the O4 demo someone said it was rubbish purely because it didn't sound exactly like the original. Song cloning doesn't interest me, I don't have the patience for it. If that person wants to hear Oxygene IV sounding like the original then he should just play the original ... the original vinyl at that. (A cynic might argue that by the latest official remastered releases, Jarre's Oxygene Part IV sounds more fake than my cover!
So I'd ask that person, what is the "real" sound? Even for Jarre, it's changed with the times.)
For me, a cover version that sounds exactly like the original is not a cover version, but merely a rote exercise with no intrinsic artistic value (although it has a technical, experimental, and archaeological merit), so I opt to find play with the feeling of his music with my Jarre covers, rather than absolute recreation, whether it's my vintage-pop take on the Equinoxe 1968 album or RendezVous IV, or a cheesy sniff on the already Camembert-laced Last Rumba. Other people, like thedk or EditEd, are far better at the literal recreation stuff. Thus someone with more patience, and better mixing expertise than I, could get all the Jarre-related tones much closer I expect, but since my intention of creating the JPS was to bring the RMI to a wider audience, not merely to a group of Jarre fans, of which I remain one (for my sins, I used to write for COTM in the early 90s), covering the same two or three tracks, I chose not to get too hung up on the fact that I'll never get closer to a real RMI than I did at the 02. We produced a number of original demos in other styles to appeal to a wider demographic.
So the JPS HS isn't there merely for people to recreate one classic tune, it has to be usable in other contexts, hence why I went with options like ADSR rather than ASR and AD (although note I do go to great pains to highlight to users how to achieve the original RMI shapes, especially since I've included curve adjustments), and added numerous extra features, including the ability to change the waveform shape: some people might prefer to ramp up the Shape to max and lower the bit-rate from 5 to 4, others might prefer a more Kawai-like 80s sound by increasing the bit-rate and lowering the Shape.
I really did agonise over whether to include more than one LFO, I know Jarre had the second added (though I'm unclear about whether it's the entire Osc 1 chain, or just the Osc 1 Trem), but in the end the ease and flexibility of Reason's routing from external LFO's I opted to stick to just the one rather than risk cluttering the GUI.
So at the end of our journey, we enjoy what it is and what it can do and how it does sound, rather than worry too much about what it isn't. Even RMI stated different tonalities may exist between their own synths!
On a personal note, I'd always wanted a Harmonic Synthesizer; it's a fascinating curio, and surprisingly versatile. While I accept there will always be criticisms, some valid, some not, I am really proud of this one.