Since this thread started, I had been thinking a lot.
Before I bought my Solaris, I was looking at Prophet 12 and OB-6. Those two had possibility to give me hands on experience and sounds I needed. OB-6 in the front and P12 in the background. I had to chose one at the end of 2016 and maybe afford second right about now. Space and money were my main concerns. Then I found Solaris and it fit perfectly, but it was over my budget and in order to get it in one go ASAP, I had to stretch my financial possibilities. I ordered last or one before last unit from mentioned batch, begging John to hold one for me. Since it was one synth for a long time and only big "analog", and I don't go well with overdubbing, I was hoping for multitimbrality to come. I like to record and modulate everything in one go. I might automate via MIDI, to free my hands, but not overdub.
I'm looking at this synth as VA, with main purpose to be as close to analog as possible, and assume, when looking at implemented models, this was the design goal. So if this is my main machine, and only "proper poly <<analog>>", let it be the best one.
Furthermore, I am lusting over Jupiter (out of my reach! hence my previous request for this filter type), for its "organic" sound. And am trying to build some patch that sounds like best big Jupiter sounds I have heard. This is my target and my reference point. Everything else can be done with what I have right now: Nord (love it!), Blofeld (hate it!), Sub (love it!), Monologue (don't know yet), smaller stuff... Achieving this with Solaris would be my dream, without going to obtaining and maintaining real one.
So to the point. I am changing my priorities. Therefore, John, if you do some sort of statistical analysis, I might give you new input, as those are my priorities right now:
1) For $100. Some sort of instability for VCO, VCA, EG, LFO (the more, the higher resolution, the better), per patch. I already am using all LFOs and lag processors (S/H at 5 Hz -> lag, which I treat as an RC integrating circuit, giving it large time value -> generous modulation of parameter) toward my goal and don't know what I'm doing wrong, feeling I'm still missing something.
Some time ago I found this quote from Korg's engineer that speaks (quote not engineer) best, what I'm looking for:
"I think it's really telling of the age we live in when you get a knob like "SLOP" on the new Prophet that makes pitch inconsistencies a programmable parameter. On one hand, you think that the level of control is great, but on the other it feels weird to deliberately degrade something that's stable. Especially if you're a young engineer striving to design something to be close as possible to perfection, it can be hard to grasp. The best lesson about this came from Mieda – my hero at Korg. When he looked at my first synth schematic, he told me, “Takahashi-kun, your circuits are functional, but they are not musical. Musical instruments do not need perfect waveforms and correct operating points. You need to use the transistor for what it is. As long as it sounds good, it’s OK.”
My example would be looking at string instruments. When I was in university, I had to calculate string spectrum in frequency domain, when guitar string was plucked. When simulating and hearing this from my sound card this was nothing like real instrument. It's like listening something obviously pretending. I know, that goog programmers can overcome this, but I am not good programmer and slop would be very helpful for me.
2) For $200. Roland filter from Jupiter 8, 4 or 6, like stated before, but either non sacrificing polyphony or with ability to switch it off and return to 10 voices. If sacrifice was permanent, I could live with 8 polyphony, but 6 is to little.
3) For $200. Wavetable interpolation. I did some WT sweeps and it seemed fine, but when I tried to do it slower and in narrower range, I started to hear stepping. This thing here excludes some sounds I'm after.
3b) Another $200. To the point three I would add adding of own wavetables. That would be completely next level for this synth! I would stop lusting over Quantum!
4) For $50. More modulation sources for EG.
5) For $100. Microtuning. Recently found ultimate solution for my problem of imagining sounds that are not distributed like typical white and black keys do. Mind blown!
6) For $100. Phase modulation.
7) Now I'm guessing, since I haven't been near original synths, but there is something missing in lower end of my Solaris. Is it DAC or are they models - I can't tell. I feel like there is too little of earthquake in my headphones while Sub, going thru same mixer, is ripping my eardrums. If it's the latter, I am voting for some agressive set of models (VCO, VCF, VCA from Jupiter?) or correcting old ones. I'm not sound engineer so I might be wrong all the way and solution lays in mastering...
For $50-$100. Proper reverb. One model is enough if it stays in spacey realm. If it would be something based on true physical models (like impulse response researched), then I am willing to pay $100 total for some nice and subtle (compared to spacey verbs) room or hall AND spacey reverb (like Cloud from Big Sky).
9) I no longer vote for multitimbrality, but if you start collecting for this feature, I will chip in.
Most important. John, please, don't wait too long.