New hardware?

[Berni]

The reason why I am so glad non of my early band gigs where never on video Thanks for sharing!

[TweakHead]

@MBTC

I'm actually glad someone put on the "geek" hat and made things so clear.

I was using Reaktor more as a metaphor, well aware of its limitations and that it doesn't qualify as a computer language, let alone a standard.

I'm not even remotely as educated as you are when it comes to computer languages. What I was trying to say - if it even makes any sense at all - is that I would like the differences between formats to be handled at the DAW level.

Why? So that developers wouldn't have to port their creations a couple of times.

I don't even know if such a thing would be possible or not. Why do I think this has to do with Access? Simply because - based on what users say - the performance changes when you switch platforms and the host software. And I imagine it's no easy to task to keep up with all the changes. I mean, Waves Audio and all the others had to put up some work for their plug-ins to be Lion compatible, right? Even though we're talking about the exact same format here which is "Audio Unit" (I think even the VST ones had to be updated to on the mac).

I imagine that programming a synthesizer such as the Virus is no easy task and the integration could be easier if only things would look the same everywhere.

Now, it's true that there's many people producing music with just their laptops. I started like that myself, even though I'm more into desktop computers myself

So I actually come from that background myself. I know all about "completely inside the box" music making. I don't even have a traditional music education at all.

But going the all way and say: who needs physical interfaces or instruments these days? You'd have to be nuts, seriously. We were talking about MIDI controllers the other day, right? There's plenty of them out there, but almost all of them feel cheap and require either very specific assignments to be made and saved, or you simply can't get the level of control you get with a dedicated instrument - and that's a fact.

@Berni

I to have a computer that can handle pretty much anything. But wouldn't you like to have a Prophet 12 next to it? Come on dude... You know you would, even though it's 2013!

[TweakHead]

And I honestly feel that there's some wrong assumptions being made here:

first, namnibor's approach to music making is just as modern as the "inside-the-box" one. There's a reason that "analogue" is making a huge come back these days. Most of the guys that have grown in a software environment such as ourselves used some kind of "emulation" of classic hardware one time or another and developed a lust for those instruments and are willing to try the real thing now.

second, take a look at second hand market even for "virtual analogues". Do you honestly think that Discovery DSP Pro sounds like a real Nord Lead? I don't. And to be honest, such a collection of synthesizers like namnibor owns puts almost any plug-in collection to shame in terms of sound quality - and I'm experienced enough to know that, and so should you. I don't know the technicalities behind it, I just know it sounds better.

third, it feels a lot better to. having a dedicated interface makes you take the time to learn the thing inside out and there's plenty of creativity involved in just combining the powers of multiple machines together, let alone the modular analogue "euro rack style" stuff.

4th, analogue modular is also very much alive and doing well, and there's a reason for it. it sounds better, it offers possibilities that software can't even dream off so far. Because, let's face it: running a new high end software filter with zero feedback latency is very demanding on the CPU, now imagine coupling that with crazy emulations of weird circuitry that just messes up the CV and being able to assign that anywhere you want. Think about it. Diva sounds wonderful but brings any CPU to its knees in "divine" quality. So, even though our computers can handle it, it's still a world apart from a real heavy duty analogue setup in terms of sound and options.

I think it has some ins and outs, just like anything out there. Being able to automate everything, to process sound quickly and with just about anything, was a revolution. But if you take pride in your sources, there's simply no argument here: hardware still sounds better but it's way more expensive, specially when you're talking about modular stuff.

Take a look at Make Noise stuff, for example

[MBTC]

Originally Posted by TweakHead

I was using Reaktor more as a metaphor, well aware of its limitations and that it doesn't qualify as a computer language, let alone a standard.

I'm not even remotely as educated as you are when it comes to computer languages. What I was trying to say - if it even makes any sense at all - is that I would like the differences between formats to be handled at the DAW level.

Why? So that developers wouldn't have to port their creations a couple of times.

In some ways my comparison of a synth-dev language versus a traditional down-to-the-metal language isn't completely ideal, or maybe difficult to put into context.

But believe it or not the challenges would be the same. For example, Java and the virtual machine technology created when Sun owned it was designed to do just that -- allow code to be written once that could run anywhere without a special port to every OS flavor. Write-once, run-anywhere was the promise. What it would require, however is what's called the JavaVM to be present on that OS. Microsoft has something similar that is known as the .Net Framework.

In the scenario you've described, each DAW would have to have something like the JavaVM or .Net Framework embedded in it, let's call it the CrossPlatformSynthHost or CPSH for purposes of this thread .. Anyway, someone would have to design, implement and own the rights to the CPSH implementation itself. Of course, nobody wants Steinberg, or Apple, or Cakewalk or Ableton or whoever to own the core technology, it would create a monopolistic scenario. This means they would need to start by forming a vendor-neutral committee (haha!! Now we are talking red tape, a big slow machine than gets very little done). Anyway each committee would need representatives. They would spend all their time in meetings every day, arguing about the way things should be done as each of them tried to sway the others in some way that is beneficial to their organization (at the end of the day, these organizations are there to make money, right? And that means staying competitive. They MUST compete, by definition).

Politics aside for a minute, and getting back to the cross-platform challenge. Java as a technology has been around for something like 20 years. Yet, most of us own very little software that's written in it. Why? Mostly because it performs badly compared to specialized languages that allow better optimizations. With a cross-platform language, you tend to end up with a least common denominator effect, where the language itself is limited by the fact that it cannot do device-specific things and take full advantage of the OS. There's a reason DAW software is not written in Java. The performance sucks. There is a saying in software engineering: "Portability is for canoes"... it just means whenever you make a technology portable across systems, there are going to be limits to that technology that are always begging to be broken through.

Then, back to corporate politics for example. Android devices are largely based on what's called the Android SDK, which is Java-based. It probably seemed like the logical language of choice at the time, what being cross-platform and all. Android is a highly fragmented OS with lots of devices that are incompatible, so a write-once language probably seemed appealing. Long-term, what happened, is that Oracle Corporation bought out Sun Micrososystems (the original creator of Java). They perceive Google (original creators of Android) as a hostile competitor, so once they owned the Java technology they decided to sue Google for using the technology that they *NOW* owned (operative word being NOW). Long-story short, big headache for Google, that choice to use Java.

So yes, this is well into the geek-speak arena and I hope that doesn't sound like I'm reaching out of scope to illustrate a point, because I truly believe it's all relevant to the cross-platform synth host idea. It's a good idea and would be great if truly viable but it would be plagued with performance problems and political problems by very definition. That's the reason it hasn't happened, is all I really wanted to say.

Originally Posted by TweakHead

I don't even know if such a thing would be possible or not. Why do I think this has to do with Access? Simply because - based on what users say - the performance changes when you switch platforms and the host software. And I imagine it's no easy to task to keep up with all the changes. I mean, Waves Audio and all the others had to put up some work for their plug-ins to be Lion compatible, right? Even though we're talking about the exact same format here which is "Audio Unit" (I think even the VST ones had to be updated to on the mac).

I could be wrong, but I think the performance differences people report are not really problematic due to the plugin standard used (VST, AU, etc) but are more due to the vast number of variable factors between the hardware and the plugin. Things like other devices sharing the same USB bus, creating latency challenges. Things like the ASIO driver and audio interface type... we didn't talk about ASIO but again its just a published standard like VST (although in the case of the Virus that's no excuse because they include their own audio interface!). Even something like the type of USB cable or whether the ports are directly on the motherboard versus a dedicated card can make a big difference.

The vast number of systems out there, running vast numbers of OS flavors, with vast number of combinations of other devices connected to them can add some real challenges to hardware testing. I'm certainly not going to make excuses for Access, because I think if they properly allocated development and QA resources to the project, proper integration could be achieved, no problem. The UltraNova is proof it can be done, I don't hear lots of complaints about the integration with the plug-in from anyone regardless of platform with that product. And the price point of the UltraNova proves that it's not some monumental task that's going to break their bank to simply develop and test their software properly.

In my opinion it simply reveals a management problem. Not enough resources of the right type being allocated to the right task. Its not a high enough priority for them if people are still willing to pay $3k for a synth that performs great as a live instrument but is poorly integrated.

So, technical solutions to a management problem almost always fail. There's only one way to solve the problem and in my opinion, new connectivity standards or plugin standards etc. are not the issue.

Originally Posted by TweakHead

I imagine that programming a synthesizer such as the Virus is no easy task and the integration could be easier if only things would look the same everywhere.

I do understand what you're saying, and to some extent consistency is great. I actually wish that all these vendors didn't have to republish their software in a gazillion different formats (VST, AU, RST etc) and the world would just accept one of them and that's that. But even with VST as the dominate standard, that standard evolves. Steinberg has the VST3 standard, yet most vendors still publish to the 2.4 standard, because theirs not really a revenue motive to implement special features of VST3 (I wish they did because hardware controllers work better with them among other things).

It's just that, as other vendors show that integration is perfectly viable and the technology is already there to do it properly, how do we convince Access to stop dawdling, acknowledge and fix the issues that exist, and do whatever it takes to man up, have some responsibility for delivering the Total Integration they promised and that many people paid for, and make things right? If Chris Kemper wants to go and play with his guitar amps for now I don't know what we can do to pull him back into the synth world to give the Virus product line proper attention unless he sees a direct negative impact to sales of the lackluster integration.

Wish I had the answer, but I don't. I have lots of opinions but no real solution to offer for that particular issue

[TweakHead]

That's a very good answer. Hats off!

Originally Posted by MBTC

If Chris Kemper wants to go and play with his guitar amps for now I don't know what we can do to pull him back into the synth world to give the Virus product line proper attention unless he sees a direct negative impact to sales of the lackluster integration.

This last part is funny! And that's because it's totally true. I mean, it's not a natural thing to leave a product that has problems to be solved in the shelf. Most specially if this product has such a high price as it does, making people expect better support for it.

And at least for some people, they're doing things wrong. I think the new specs on the Virus ti are very appealing. We've debated this over and over here, all right. But the fact is, seems like most of us are just waiting for them to fix this issues and even a new product that not only solves the problems but makes the Virus a more competitive product in a market growing rich in digital instruments that deliver good sound at reasonable prices. This is the issue, I think.

I thought the UltraNova only had a software editor that worked in stand-alone mode. Didn't know you could select it just like any other plug-in like you do with the Virus and that it could stream the audio through USB. If they do that and it works fine, even though it's a monophonic synthesizer, it puts Access to shame. I mean, it's called the TI = total integration. Not like it's only a detail for them.

Always thought that the TI thing was a big challenge and that people were a bit unfair. But if there's other products doing exactly that and with no issues at all, it's a different story. What I think still holds it for them is that the synthesizer's sound is plain gorgeous. But the specs alone don't really cut it for me, owning the C and all, to spend another big buck for the next Virus - unless 100% sure all bugs are fixed. Still, I'd rather wait to see the new offspring - whenever it comes - because a company can't be relying on the same product for so long without putting another thing on the market.

It's kind of crazy this whole story. I mean, for me it's still the best Virtual Analogue out there when it comes to sound and synthesizer specs. The synthesizer's engine itself is really good and bug free. It's a real charm to program and a very deep machine, otherwise there wouldn't be such a thing like this forum. Why screw up on this integration thing? Beats the hell out of me, and your answer just made things more clear. You and Berni were right all along about this. I'm an atheist now

[MBTC]

Originally Posted by TweakHead

That's a very good answer. Hats off!



This last part is funny! And that's because it's totally true. I mean, it's not a natural thing to leave a product that has problems to be solved in the shelf. Most specially if this product has such a high price as it does, making people expect better support for it.

And at least for some people, they're doing things wrong. I think the new specs on the Virus ti are very appealing. We've debated this over and over here, all right. But the fact is, seems like most of us are just waiting for them to fix this issues and even a new product that not only solves the problems but makes the Virus a more competitive product in a market growing rich in digital instruments that deliver good sound at reasonable prices. This is the issue, I think.

I thought the UltraNova only had a software editor that worked in stand-alone mode. Didn't know you could select it just like any other plug-in like you do with the Virus and that it could stream the audio through USB. If they do that and it works fine, even though it's a monophonic synthesizer, it puts Access to shame. I mean, it's called the TI = total integration. Not like it's only a detail for them.

Always thought that the TI thing was a big challenge and that people were a bit unfair. But if there's other products doing exactly that and with no issues at all, it's a different story. What I think still holds it for them is that the synthesizer's sound is plain gorgeous. But the specs alone don't really cut it for me, owning the C and all, to spend another big buck for the next Virus - unless 100% sure all bugs are fixed. Still, I'd rather wait to see the new offspring - whenever it comes - because a company can't be relying on the same product for so long without putting another thing on the market.

It's kind of crazy this whole story. I mean, for me it's still the best Virtual Analogue out there when it comes to sound and synthesizer specs. The synthesizer's engine itself is really good and bug free. It's a real charm to program and a very deep machine, otherwise there wouldn't be such a thing like this forum. Why screw up on this integration thing? Beats the hell out of me, and your answer just made things more clear. You and Berni were right all along about this. I'm an atheist now

Yeah the UltraNova editor can only run as VST/AU inside a DAW, you cannot launch it separately. Sometimes you hear people say "no standalone mode" and that gets misunderstood as thinking that the synth is not a standalone instrument and requires a computer... its not, its actually designed for live play in the true synth sense, you don't have to ever use the USB cord if you just want to gig with it, same as Virus. But some people don't like the idea of loading up a DAW to edit sounds, they want it to work like Maschine where you can just launch the Maschine editor. To me its not a problem, I've always got a DAW loaded anyway.

Even without the UltraNova/MiniNova as examples, the fact that there are many audio interfaces out there that stream audio from regular audio inputs to the computer via USB goes to show that USB is perfectly capable. Now granted the more instruments you have, the sheer amount of sound data being converted from analog to digital might saturate the bandwidth capability of USB (this is why higher-end audio interfaces tend to use Firewire or in some cases Thunderbolt), but I believe USB should be able to handle audio streams of the typical home musician. The higher bandwidth need is more for multiple instruments recording at one time.

That makes me wonder though... I think the TI2 only offers 3 stereo audio outs to the DAW if I recall correctly? How many multi-timbral parts could you have? It seems like only 3, at least as separate audio streams.

That could be one reason people run into issues with the Virus and not the UltraNova / MiniNova, because the UN is mono-timbral, thus it never needs to send more than 1/3 the max bandwidth stream of the Virus through USB. A lot of people have pointed out that there's nothing really stopping Novation from making the UltraNova multi-timbral, since its purely a software feature with a virtual analog synth like UN or Virus. But maybe they decided to err on the side of reliability rather than feature set? Who knows, maybe they tried to make it multi-timbral and realized the wonky nature of USB is going to put them in a dilemma like Access currently is, so they just made it mono-timbral and kept the cost low.

I wonder if the Ti2 was used with only one audio out would so many people still have so many problems with it? Maybe those that don't are using only one part at a time? I don't know, I seem to remember having problems even with only one audio out with the one I briefly owned. Besides, it would be a damn expensive mono-timbral synth up against the Ultranova as competition.

[TweakHead]

I'm not much interested in using the multi mode myself. Unless we're talking about preview purposes, of course. For recording the audio I'd rather have all the voices available for each patch. And surely wouldn't mind if they cut that feature all together if only things would work in a very reliable way. I think the integration has obvious benefits: being able to automate parameters is an obvious one, saving the patches used inside the projects another. Editing on the screen is useful for synthesizer with complex structures and many options like the Virus ti. Don't really miss it for my C - and if I did there's some options out there, so that's not really the bonus here. The other factor is the one you pointed out yourself: having a great sounding synthesizer that doesn't hit the main cpu that shares the best of both worlds: it behaves as an instrument and it's still easily configurable inside the software environment we use to make music with. So we're talking high end quality that spares our main cpu and we're talking functionality and convenience.

Unless Novation has some wild genius that managed to make at first attempt what Access has been struggling with for so long without success (which I find hard to believe), I'd say it's shocking that other developers haven't implemented this on their own products.

About the other thing I talked about. It would be possible to have a card similar to Universal Audio but open to third party developers, right? That would be amazing. Maybe a good incentive for developers to implement more demanding code without worrying about hitting our main cpu to hard - like it's happened with Diva. Noticed that their first update was mainly focused on that, adding the capacity to use more then one cpu core I think (but not sure). This is a big factor when it comes to software, right? I have no doubt that if this wasn't the case, software would easily compete with just about anything when it comes to quality. I've seen some texts about current DSP theory (again, not nearly as versed as you are in such matters), so I have a general idea of where we stand today. I mentioned Discovery DSP pro and failed to mention they've just implemented "zero feedback latency filters" on their last update - which brings it closer to the level of more recent offers out there. There's a lot of them doing that: Diva, Monark - even though this one is a Reaktor instrument, if you take the time to dig through it, they've locked access to the filter (sad, very sad but true) -; Madrona Labs's Aalto, Sonic Academy's ANA synth, Waves Audio's Element, Lush 101, etc. All of these have modern DSP (demanding) code into them and all of them present better representations of self oscillating filters that we wouldn't even dream about a few years ago. Funny how the good old Moog inspired filters on the Virus (from C onwards I think) still hold it compared to even these new offers. But not by much I'd say.

Which one of these you guys like best? (interesting subject, no?) very cool thread btw

[MBTC]

About the card that's open to third party developers... that was what the TCE Powercore cards were supposed to be. You might remember there is a Virus plug-in for it, that runs off the DSP on the PCI card or FireWire unit. But TCE stopped production on the product, not sure why. The Virus plug-in was based on the features on an early model (Virus A or B I guess), and supposedly for all practical purposes was just like having Virus integration that actually worked. Anyone have any info on why progress on this ceased? It would be really unfortunate for those who invested heavily in them (as I recall they weren't cheap, and you had to pay for each Virus instance you wanted loaded, for example four instances were the equivalent of having four Virus keyboards but the license cost was x4 as well, in addition to the hardware Powercore itself).

What's funny is that throughout all this, GPUs have made insane progress over the past five years, achieving the type of performance gains that CPUs used to, but what's more is they excel at parallel operations, the one thing that CPUs can do, obviously, just apparently not as well as GPUs do. Toolkits like CUDA SDK for NVidia chipsets allow writing pretty much any sort of application that runs on the GPU and utilizes that instead of the CPU. I'd hazard a guess that the average low-end gaming GPU (think $100 graphics card) has the computational power to run circles around the DSPs used in devices like the Virus (although it's a little hard to compare them on paper).

I really don't know why the GPU is so under-utilized in audio processing. It is an amazing hardware resource that all but the lowest-end desktops and most laptops already have inside the box. It could be because the GPU market is split (unevenly I might add) between offerings from NVidia and AMD, although NVidia has been the clear leader for some time and I believe has greater market share. I know there were a few plug-ins available that ran on CUDA, I actually downloaded and tried a convolution reverb plug-in that never seemed to take off.

You can see some of the non-gaming related audio & video apps that run on CUDA rather than the CPU here:

http://www.nvidia.com/object/cuda-ap...=Video & Audio

[TweakHead]

http://gpuimpulsereverb.de/

there is this one which is available already! For this type of reverb it's actually pretty cool that it doesn't hit the CPU. Shows what's being wasted

[MBTC]

Originally Posted by TweakHead

http://gpuimpulsereverb.de/

there is this one which is available already! For this type of reverb it's actually pretty cool that it doesn't hit the CPU. Shows what's being wasted

I remember trying that one a couple of years back. Seems like I ran into some issues with it, but I don't remember what they were. It looks like an actively maintained project, so maybe I need to give it another try.

There was another convolution reverb I tried, but I don't recall the name of it. It was a pure CUDA implementation (whereas this one seems to work on both Nvidia and AMD GPUs).