kortezzzz wrote:You can see that in the recommended system requirements at their site (top right corner).
That's obviously a mistake. Probably a leftover from a sample package. Why I am so sure? This is what the text says:
• 8 GB RAM or more is recommended. Plus at least 8 GB hard disk space for installation.
The installation uses approx 200 MB on the hard disk, and even the before metioned absolute maximum of what you can do with the synth never exceeds a few dozen MBs in RAM. I don't even have 8 GB RAM!
Trust me, neither does it need 8 GB RAM (or 4 GB), nor does it use 8 GB of hard disk space. Nor does it overload you CPU at any time. I used the same preset I mentioned before for the ultimate test. Everything to the max and then I played 32 notes simultanously (32 polyphony is the max). The CPU load raised to about 35% on my 3rd generation i5 with this totally unnatural crazyness (nobody would use a synth in that way). And I expected it, because it also says (and that's way more likely to be true):
• Intel Core 2 Duo, AMD Athlon 64 X2 or newer.
You see, it's much ado about nothing.
kortezzzz wrote:Other than that, only god know what is the "real world" performance gap between FS and and juce. Not only the macro differences, but also the micro ones, as I've stated about the voice management when playing chords. If spog reached to the top of FS's abilities with this synth, we may have a problem here, because new projects only push the recent limits toward for being able to compete with the behemoth firms in the market. I for instance work very hard to invent formats that would allow me to stay in the safe zone of the performance limits because some of projects already failed due to their size. So my conclusion is we have 2 paths to choose from; the first is being optimization masters with asm, ruby and dsp or be very selective with the projects we choose to develop by always considering the prices. Spog went here a little too far and you gave him one stringent comment. Then how would a client that payed money would react to such a situation?
I always give honest feedback, and Spogg knows me for that and likes it. Nobody's helped, if I'd hold back on what I witness, or just say "good job". Everybody here who creates a synth does a good job, so that goes without saying. And it'S the feedback that can help you improve your creations. Or you can choose to ignore the feedback.
About commercial plugins I can't say much, since I never sold any of my stuff. From Adam however I know that his customers are overwhelmingly positive, and rather ask for more features than complaining about the sound.
Flowstone will always be slower in processing than any C++ engine. That'S the price you pay for all the comfort, Flowstone offers in creating synths. No need to stare at rows of complicated code that will only work if you set up a programming file AND a header file, forcing you to switch back and forth between them when correcting stuff. You also get no visual feedback regarding your GUI. You have to build it, then compile your whole plugin, then load it in your DAW and finally you can test it. Compare that to the easeness of Flowstone's "views"!
To be honest, if you want to compete with the behemoths of the market, you have to use the same tools as they use. Which is C++ with a specialized library like JUCE or WDL-OL/iPlug.
That said, the 64 bit version of FS4 is far ahead of FS3, so there is yet another bit of headroom to explore, before it reaches its limits. At one point though, you have to switch to C++, if you're seriously considering to compete with Rob Papen, U-HE and the like.
