Quote:
Originally Posted by infraction
If a say twenty strong team of native plugin developers got together and decided to make a more professional product than no doubt it would easily compete or maybe surpass the virus sound.
But that isn't happening, even novation who seem to be very pro software keep their best algos in hardware these days. Maybe the fear of warez is to blame for the lack of really good quality softsynths that can match their hardware counterparts?
|
I would definately agree that piracy fear holds back VST progression, but I couldn't agree that there aren't some soft synths taking it to the next level with large teams of talent. Take a look at something like Omnisphere or Alchemy. Especially with Omnisphere, you can browse around it and immediately see that there were a lot of folks involved in it's making. Granted, we are talking $500 softsynths at this point, but they are out there.
If I could try to identify one item that contributes to the widespread mediocrity of so many "regular" soft synths, it may be the availability and use of open-source algorithms to do much of the processing or filter work. I think many of these VSTs are based off the same common library that provides much of of the math functionality so the developer can focus on the creation of the instrument. The instrument itself may turn out unique, but that uniqueness is defined by the UI, the range of parameters, the included patches, etc.
At this point, the VST developer has made the choice of not re-inventing the wheel... this is typically a good idea in software development, finding ways to re-use work others have done, but when creating virtual instruments or similar, it can be enormously stifling to creativity and uniqueness to have a common foundation of re-use across projects.
Some of the most famous synthesizers were created by one or two guys, and what often gave the synth their unique sound was really a flaw or bug in the design that manifested in ways that was recognized as character. If we saw more of that in soft-synth creation, the end result would be better, but there is a resistance to recreate all algorithms from scratch (first because it's really damn hard work and not everyone has the math skills to do it, second because its a means to an end and most synth creators would prefer to work at the end).
Back to the subject of piracy for a minute, although as a developer of business software I am not up to date on embedded systems / DSP programming, I am unaware of anything that would prevent the code that powers the Virus from being completely reverse-engineerable. In fact I'll step out on a limb and say that someone has surely done it already. Furthermore, the fact that they offer a pure software solution that requires only powercore tells me that there is nothing necessarily magical about the algorithms in the Virus, someone has surely already dissected them.
I think hardware has a bit of sound advantage simply due to power of dedicated processing. By that I mean the DSP in the virus is doing one thing and one thing only -- producing sounds. The actual power of the unit is rather pathetic compared to a modern CPU, but the difference is that it is not producing sounds for multiple, disparate instruments on top of all the executing code for the host software which is running atop a gazillion services, which is running atop the OS kernal itself which is of course way more general purpose than a synth DSP. Multi-threading / multi-core helps with this problem but it's only as beneficial as the software is designed to take advantage of it, and since multi-threaded code is hard and limited by the presence of any serial-execution tasks, there is a seriously limiting lowest common denominator here.
So by adding a hardware to the arsenal, you are guaranteed a certain sound output. It doesn't matter how many other pieces of kit make up the studio, each one always produces constant output. I can't say that for a VST environment on my current PC, even though its a nicely powered CPU. I can add a VST, whip up a sound, and compare that side by side with a Virus and be amazed that the VST sounds as good or in some cases better. But once I've built out the mix such that there is A LOT going on, the sounds coming from that VST can be muddied up or lagged by other things -- in some cases the entire DAW starts wigging out from just too many instruments, effects, etc all bogging down the same chip that is also running regular OS services. Meanwhile the Virus is able to do it's sole job, producing a consistent sound that doesn't change as you add more hardware synths (it could of course as you add more effects or somethign to the virus sound itself, which may affect polyphony etc).
But despite all that, and the elegant simplicity we should get from just plugging in another piece of hardware kit, it seems we are still plagued with timing / latency issues, integration problems in general. Now we are at the core of why I am remaining software based at the moment.