View Single Post
  #17  
Old 13.08.2015, 06:35 PM
MBTC MBTC is offline
This forum member lives here
This forum member lives here
 
Join Date: 16.04.2010
Posts: 1,082
Default

I too am a software engineer (director of engineering by title, but still very hands on in coding), started programming in assembly in 1980. As an odd coincidence, I also develop iOS apps (but it's only one type of project I work on, not really my primary focus. Sadly I also have to do Android too). From that perspective I could relate to Kemper's bias toward the Freescale chips -- with lower level programming, it's not as easy as just recompiling for a different platform, I suspect this is true with the Virus code base.

I do think that a specialized chip is absolutely going to be better than a general purpose chip for the simple reason that it is dedicated to a single function and specialization is everything. It does not have to deal with OS task scheduling and all of the other things that come into play in a general multithreaded operating sytem. A specialized chip can focus on the single task it was designed to do (but that of course is different than saying the DSP is more powerful, because typically they are not). Just as a general idea in computing, I find that when dealing with smaller, specialized applications, things like latency become less of an issue, so it could be that filter latency is a big advantage? That's only a guess.

My experience with polyphony / total voices is quite different than yours. Currently I only have a Snow but I did own a Ti2 at one time. The total amount of polyphony I could get out of it was wholly disappointing compared to what I can get out of a single softsynth instance (take your pick, Massive, Zebra whatever). But it's important to keep in mind that A LOT of factors come into play here that makes that particular discussion perhaps not even worth comparing notes between one individual experience and the next. By that I mean there are things like the specs of the PC itself, which services or background tasks are running on that PC, the type of audio interface in the PC, the DAW software, the plugin software, how everything is configured, which audio driver version and what the buffer size is set to, the actual complexity of the sound being compared and the impossibility of perfectly duplicating the resource usage scenario on two separate platforms.... so on and so forth. Those can all dramatically influence how many voices you get out of a particular softsynth. I will say that in general with my particular setup, I can usually get the polyphony equivalent of maybe 5 or 6 hardware Virus Snows with a single softsynth. With minor tweaks someone could get a very different result.

I try not to read too much into the sound comparison vids I see, for example the Sylenth one which I had seen a while back. Sylenth is a decent plugin, but I wouldn't call it current-gen. Also that particular example really only shows what happens when one guy sets out to create one sound a certain way. It only shows what occurs when a knob with one purpose is turned a certain amount on two different synths (but really the implementation behind that knob can differ greatly). To some it stands to reason that if you start with the same wave, then turn the same knobs the same amount you should get the same effect, but really if it were that simple it would just be proof of the over-reliance on open-source we talked about. It's not an apples to apples comparison unless every bit of code behind each knob is identical on the 2 synths being compared. I'm pretty sure if I sat down and created some patches on Dune2, then handed them off to him and said "here, recreate these sounds on the Virus", the results would be quite different. I've also seen some vids where Virus vs. softsynth results were pretty much indistinguishable, so I think a lot of that depends on who is making the vid and what point they are trying to establish in doing so.

I definitely agree on the point about the the Nord softsynth. It's really not to the advantage of any hardware synth maker to produce a perfect replica software version, only to watch it get pirated into oblivion. The profit margins are higher if you can sell the hardware and the software together, and to some extent support costs are lowered because (in the case of the Virus) you don't have to support the basic sound engine every time a new version of OS X or Windows comes out, you only have to support the editor (Virus control)... or in the case of most hardware synths you just say screw it and don't have an editor at all. lol. Support costs saved.
But even beyond that, hardware has a sexiness and a musical allure to it that is hard to replicate, so honestly if I were Kemper and I had created a musical instrument like the Virus which not only was so influential to electronic music over the last 15 years or so, but has still maintained a certain mystique and respect, I certainly wouldn't look to toss all of that in order to produce a cheaper software-only equivalent, it would be like throwing out the baby with the bathwater. Would not be a wise business choice.

Also, I've mentioned this before in other threads but I actually heard a representative from Moog say in an interview that there is nothing you can do with a Moog synth that can't already be done with software, but the real reason to buy their products was the inspiration to music creation they provide. I thought considering the source that was saying a great deal (and I thought he was right on both points, about the capability of software but also the inspiration hardware provides).
Reply With Quote