Owners of the big DSP cards such as Pulsar/Scope almost always say that their synths sound better than VSTi. I own a PulsarII and to a degree i think that's true: the Creamware Primsa, Lightwave and Minimax have a indefinable richness that few VSTi that I've tried come close to matching. I'm not talking oscilloscopes here, just the human ear. Nor am I desperate to defend my "investment" in DSP - I've recently moved to a predominantly native solution. But, well, the synths just don't quite seem to sound as good.

Against that, synths like Absynth are more versatile and innovative than anything available for Pulsar, and I find the lack of VSTi-style rendering of Pulsar synths to audio very frustrating.

But what I'm really wondering is this: What are the quality limits in VSTi ? What determines the richness and depth and character ? Is it something fundamental in the specification (I wouldn't have a clue here); is it that developers trade off too much quality to conserve CPU, or is it just not possible to match DSP algorhythms ? Would a VSTi that burned half your 2MHz CPU with just a couple of voices sound beautifully rich and harmonically complex, or is raw computing power not the issue ?

Alternatively you may think that there is no quality problem. Maybe I just know the "Creamware sound" and am waiting to hear it in VSTi, and all is illusion ?

Either way I'm extremely interested to know where "quality" comes from and whether VSTi can - or even need to - improve.

Well I think many oscillators on VSTi are sample-based and I imagine there is quite a bit of optimisations in building a VSTi for a native solution that developers for DSP solutions don't bother with.

Some generalisations there.

Mind you, perhaps my ears aren't up to scratch, but although I can sort of hear a difference, I'm not so sure it's a quality difference or just a character difference. But then again, I don't necessarily trust my ears to make that sort of value judgement on a synth.

Sound is sound to me, so I get lost as to describe what an "oscillator" is supposed to sound like or which sounds better.

It makes a noise, the noise work in my track and sounds good to me...voila!

Even if it were true that the DSP synths in Pulsar sounded "better" than VSTi on a native solution the fact is you pay a surcharge for it.

Oh well, this is not really helping your question at all. Maybe someone who really does have the ears to make quality distinctions in A-B testing can contribute.

But my personal opinion is that for a lot of us the huge convenience and monetary and physical economy of using VSTi is probably going to bias us in a topic such as this one. Nevermind that we are all VSTi fanatics anyway.

spirit wrote:

I find the lack of VSTi-style rendering of Pulsar synths to audio very frustrating.

Is this planned for a future release?Sounds like an important thing to have nowadays!

Concearning your topic, I think it just depends on plain cpu-horsepower and genial masterminds who develop great sounding instruments for any Hard/Soft-Environment.

peace, moritz

Everything is relative. Off course when you play solo a 32Meg piano, it will sound better than a 4Meg piano. But on the whole MIX, within a song, sometimes even a simple 2Meg piano do the work. So you have to do songs using VSTs and compare with a song made using a DSP system. My own experience shows that is not the actual Hardware/Software that counts, but how you use it. So, even if VSTs are not as good as a DSP system, you can make it sound good if you select the right synths. I have a friend here who did some songs (http://wellington.dashsynthesis.com) using some very old retro synths back on the 80's. He selected the best patches on each synth, so the song actually sounds great, even today.

Just my 2 cents about this topic...

Regards, WilliamK

Hi Spirit,

DSP is "Digital signal Processing" which is done in every softsynth found anywhere: Virus, Waldorf Q, Waldorf A1, FM7, DX7, NordLead, Virsyn TERA, Reaktor, etc.
So there is maybe a power issue that makes a difference between some DSP chip based solutions like Virus, Pulsar, PowerCore, etc. and a VSTi.

The question is: How skilled and how clever is the synth developer? How good is he/she in bringing the processor usage down to a reasonable amount?

I definitely would like to hear some examples of the Pulsar instruments you meantioned to hear their quality. Really.

Spirit, the following isn't a rebuff to you. You've just brought up something that has been naging at me...

I get so weary of this argument. VSTis, hardware, whatever, are ONLY as good as the notes you play with them. Listen to Klaus Schulze's "Picture Music". Done on a Teac 4-track with simple, limited synths. It is so good that it pops up OVER AND OVER in samples by current acts with much more expensive stuff ("Totem" was sampled by the Orb for "Fluffy Little Clouds" to name an example).

I have a ROOMFUL of "classic" hardware that people "in-the-know" assure me sounds superior to my software. I don't buy into this. This pursuit of the Holy Grail of "sound quality" is a red-herring that keeps us pouring money into the latest hype, and prevents us from getting down to writing our songs. We have such miracles at our disposal in the software world that exceed the quality of my wildest dreams back in the eighties.

I'm not saying that the Creamware stuff doesn't sound great, it does. And some of the synths sound better than some similar VSTi designs. But my point is that it DOESN'T MATTER. What matters is getting your ideas and visions across, and that miniscule difference in quality between one synth and another is not likely to affect your ability to do this.

And, has anyone else seen the contradiction in the trends and advertisements lately? I'm speaking of companies advertising that they are producing "such-and-such quality", and then turning around and trying to sell us "tape saturation/tube distortion" to make it sound LESS HIGH QUALITY?

Moritz Morpheus MkIII wrote:

Is this planned for a future release?Sounds like an important thing to have nowadays!

Pulsar does sport an XTC mode where you can use your Pulsar synths as part of your VSTi host. Spirit has probably got more experience with that then me as I've never bothered even trying it. Maybe they can render then.

I don't know really - doesn't interest me.

I totally agree - in fact I think it's self-evident - that it's inspiration and perspiration that will make the difference (and probably even more important are your contacts and schmoozing skills ). With a PC you can put together a FREE setup these days which blows away what I spent thousands on not so many years ago. So the question about audio quality is from the technical side rather than the musical.

For examples of Creamware sounds here are Planet-Z members' MP3s:
http://www.planetz.com/forums/viewforum.php?forum=17&2062

And Creamware's own demo sounds of the Minimax moog emulation:
http://www.creamware.de/en/Products/Software/CreamWare/Minimax/example s.asp

Hi Caleb. XTC mode is a steaming hillock of cow dung - and no it can't render - which makes it worse than useless.

Scott and William you have hit it on the head on a quit a few things!!!!

First... I get caught up in trying to find that Right tone or sound... and what happens his I end up spending too much money

I need to focus more on what something is going to sound like in a Mix rather then playing a few notes and going Bahhhh.... I have learned this lesson the hard way lately with Guitar in some tunes that I have been working on... I take the Guitar sound by its self.. and I go this sucks.. its not meaty enough... but then throw it in with some bass and Drums and a little FX and hmmmmmmmm yummie.

As far as what sounds better..... well I put it this way.... when I listen to a "song on the radio"... and it has me tapping my fingers... I dont go hmmmmmmm I wonder if that bassline is Harware or software or what plaform it was done on or if a Monkey played it ..... it doesnt matter because I am enjoying it!

now sopme of this might not have made sense to you all but it helped me work through some problems that my therapist has been wanting me to voice out

ner0z wrote:

As far as what sounds better..... well I put it this way.... when I listen to a "song on the radio"... and it has me tapping my fingers... I dont go hmmmmmmm I wonder if that bassline is Harware or software or what plaform it was done on or if a Monkey played it ..... it doesnt matter because I am enjoying it!

You enjoy things played on the radio...
You're not from round here are you?

choochcat wrote:

ner0z wrote: As far as what sounds better..... well I put it this way.... when I listen to a "song on the radio"... and it has me tapping my fingers... I dont go hmmmmmmm I wonder if that bassline is Harware or software or what plaform it was done on or if a Monkey played it ..... it doesnt matter because I am enjoying it! You enjoy things played on the radio... You're not from round here are you?

I knew that I should have made it clear that I meant.... not your Pop shite radio or top 40.... the only time I listne to the Radio is on the internet and it is speciality stations.. like Ambient Industrial Prog.... clasicall.... what ever

I Think that many native VSTi's and effects sounds good enough as long as you run them in 96KHz+ to get away from software aliasing as much as possible.

To put another perspective to it, even sound-quality in itself is a very relative value. Remember that the DX7 has always been looked down upon as sounding thin and uninspiering, but it's still the synth with the longest lifespan in music production because of how well it sits in the mix. Mostly sound-quality of synths is judged on how they sound on their own with lovely, lush multi-layerd sounds that are completely unusable in a serious mix

My stand on this is always this: there are incredible sounding professional productions that are allready made with only or a lot VSTi's. That's how good they are sounding

Hm, maybe you can judge a synth's quality just by playing a few bleeps and bloops. OK.

Years ago (about WWII) I read an interesting story about Eddie Van Halen doing a gig somewhere in the pampa... don't know why but he didn't have his usual guitars there so he played on a ultra-cheap mexican guitar with the result that it still was the original Eddie Van Halen sound.

Talking about VSTi's, heh, just do some arpeggio lines using the cheapest Neon sound and add just a bit of reverb and a long delay. So much about sound quality.

Bottom line, really is: what you use isn't as important as what YOU do with it.

Whether or not it sounds satisfying to the ear is the important thing: the only people who will ultimately give a damn about what was actually used to make it, what has more DSP horse-power, whether you used the latest available update to your VSTi host app, etc ...will be other musicians and gear-heads.
And if you only care about impressing other musicians and gear-heads....well, you're really limiting your audience.

Har wrote:

Bottom line, really is: what you use isn't as important as what YOU do with it. Whether or not it sounds satisfying to the ear is the important thing: the only people who will ultimately give a damn about what was actually used to make it, what has more DSP horse-power, whether you used the latest available update to your VSTi host app, etc ...will be other musicians and gear-heads. And if you only care about impressing other musicians and gear-heads....well, you're really limiting your audience.

can I hear an AMEN!!!! come give it up for the Rev HAR!!!!!

Well put!

Some people have a belief system and are tied to it. It colors their perceptions. Every time a new technology comes out it is challenged and called inferior to older, established instruments. Juno's were hammered by purists because it had a DCO so it wasn't really good. DX7s caused many a musician to sell off equipment that is cherished today. And DX7s were hated for a good deal of the 90s. You couldn't give one away. But now it's hoter than VA analog.

Some people are used to certain sounds and "how" that timbre should sound as far as they're concerned. Whether it's colored or pure, good or bad isn't the issue. Niether is quality. The issue is what sounds good to you and as William and Peter point out how well sounds work in mixes.

I've used enough equipment to last a lifetime. I like the sounds of VSTis enough to replace hardware. They fit well in my work. That's the only thing that matters, that and making sure the customer or consumer enjoys the work.

And yes, it's a programming thing, William can attest to that. In Reaktor some ensembles are high on CPU useage. To get them to reasonable levels involves removing or redsigning components.

For VSTi makers I'd imagine it's the same issue.

Whatever inspires you is great. Whatever you use to make the music you dream is great. Household appliances and a sampler can make for some of the best instrument sounds you've never heard. It's all relative...

Har wrote:

Bottom line, really is: what you use isn't as important as what YOU do with it.

GOOD! Totally true! Some time ago I was doing music just with the crappy sbLive and I was happy with it. Now I'm starting to use CubaseSX with a few SoftSamplers and Reaktor. I'm on a new level now, but is harder to make music this way, too much things to tweak.

Wk

[quote="x_bruce"]Some people have a belief system and are tied to it. It colors their perceptions. quote]

Too true. Remember when the 303 was released nobody wanted it because the sound sucked according to the 'specialists', for a bassline emulation, but also as synthesizer. But it got popular like no other synth after a few years and suddenley 'everybody' thought it sounded great. No not even with inspired compositions, but with the most basic music heard until then. Aciiid..!
And now you buy a 303 for 1000 dollars....That's how things turn around...

x_bruce wrote:

... Some people are used to certain sounds and "how" that timbre should sound as far as they're concerned. Whether it's colored or pure, good or bad isn't the issue. Niether is quality. The issue is what sounds good to you and as William and Peter point out how well sounds work in mixes. ... Whatever inspires you is great. Whatever you use to make the music you dream is great. Household appliances and a sampler can make for some of the best instrument sounds you've never heard. It's all relative...

Well said. Taste changes over time. I recently used Ninja in a few tunes because I now like the sounds it does and yesterday I checked out Synthetic again to find out that I really like it now.

And I took some of my voice to create some interesting drum sounds in Kontakt. Cool stuff. I should check my household appliances again...

Of particular interest to this thread, our latest techtalk feature deals with sample rate and bit depth, and gives some concrete reasons why some synths and systems will inevitably sound better, or richer than others from a scientific perspective...


http://www.applied-acoustics.com/techtalk2.htm

........ mmh, what the specs have to do with good sound? (PPG is 8bit, 909 hats are 6bit, linn is 8bit, emu sp12 is 12bit etc.)

Fascinating article Scott - answers most of the questions about the technical aspects of soft-synth quality. I especially thought it interesting that professionals couldn't really tell the difference between 16 and 24 bit. Perhaps for users this "shift" is also just part of the never-ending upgrade path we're all encouraged to follow?

[quote="Illusionist"]

x_bruce wrote:

Some people have a belief system and are tied to it. It colors their perceptions. quote] Too true. Remember when the 303 was released nobody wanted it because the sound sucked according to the 'specialists', for a bassline emulation, but also as synthesizer. But it got popular like no other synth after a few years and suddenley 'everybody' thought it sounded great. No not even with inspired compositions, but with the most basic music heard until then. Aciiid..! And now you buy a 303 for 1000 dollars....That's how things turn around...

Now TB-303s suck again! hehe They are boring...

Reg

Quote:

Perhaps for users this "shift" is also just part of the never-ending upgrade path we're all encouraged to follow?

i completely agree!! as a professional, working sometimes in big studios i can see that nobody cares of these "bit and rate" things, the signal jumps many times from digital to analog and back without problems, and, if it sounds ok nobody cares for some digital clippings when mixing....... and, i do not care to repeat, these aspects are totally out of context when we speak about instruments[/code]

The problem with the whol thing is that to compare $50 vsti to $2000 moog is daft . Of course the moog sounds better. So what. Id rather listen to good music played on a $1 kazzo than some dullard on a $2000 moog. Actualy as far as digital(FM/Samplers) synthesis goes i think VSTi have the edge. Most VA still lacks some of the warmth or edge but there are a few realy ones out there. And if the listener cares more about the sound quality than the musical content then there is a bucket load of shit chart music they can listen to.

chris

To get back to the original thrust of the topic, I think that any perceived difference probably lies in the hardware side of your set-up. Your Pulsar card probably has much better AD/DA converters and a good pre-amp when compared to the soundcard you listen to your VSTi through, so naturally it will sound better through your speakers. I certainly don't hear any difference in "quality" [whatever that means] when I plug my Trinity or Prophecy into the back of my soundcard and play it alongside my software stuff.

well, here is the algorhythm take on the issue:

I use both Pulsar and VSTi's regularly.

~ I do agree with Scott. S. and others that which you use makes or breaks tracks. The VSTi only tracks here at KVR attest to the fact that great tracks can be made with such a setup. The tracks of Pulsar people @ PlanetZ can similarly attest to the fact that great tracks are being made by such a system.

~ I also agree that CODE IS CODE. Hence, it is irrelevant whether a Synth is processed on DSP hardware, or on a computers' CPU.

THAT BEING SAID:
~ I think there is a difference in quality between Pulsar synths and VSTi's, generally speaking. Pulsar is a system with dedicated hardware - it has always been designed with quality over voice count. VSTi developers have to consider the bottom line, processor wise, which seems to be around a 400MHZ Pentium II. In other words, it seems that compatability seems to be put before sound quality in a lot of cases. Also, consider the programmers, most anyone can program VSTi's without special hardware, and the dev kit is free from steinberg. What is the price of SCOPE DP for pulsar these days - $$7,000 or so? This difference in accessibility to programming ensures that the quality bar will tend to be higher on the more expensive platform, in this case.

~ I agreed with Scot S. and others that it doesn't matter in the end with the context of tracks. But for me, I think I enjoy sound design as much, if not more than, writing "songs." To me, a song is a consumable for the public in and through which I can showcase what I have been doing with my sound design skills. So for me, the sound of the synth is very important. Sometimes a VSTi fits this bill, USUALLY NOT.

~ Feel free to disagree. Chances are, if your reading this on KVR, you can't compare anyways because you don't have Pulsar. And more to the point, I don't care if you do disagree, because I have stated my case, and am sticking to it. Of course, as Spirit has said, I may be blinded by the pricetag - I have never done A/B comparisons of Pulsar synths to feature-comparable VSTi's. To me though, the phenomenal experience tells me that "pulsar sounds better" and that is all I have to go with. Finally, CPU's get faster as we speak, and I presume that VSTi's will improve in quality because there will be a rising "bare minimum" system spec that the synths are designed for, and the VST-programming knowledge base will continue to grow.

At present, I mostly use VSTi SAMPLERS cuz they aren't synths. Of course, there are nice VST synths: like CRYSTAL!

And that is about all I have to say about that. -

algorhythm wrote:

What is the price of SCOPE DP for pulsar these days - $$7,000 or so? This difference in accessibility to programming ensures that the quality bar will tend to be higher on the more expensive platform, in this case. ~ Feel free to disagree.

Thankyou, I will. I think that this argument is flawed. There are some mighty talented programmers out there and I'm sure Steinberg or NI can afford to pay them more than Creamware, given their much larger user base. If anything, I think that the pool of programmers interested in coding for Pulsar would be very small and not necessarily people with good coding skills. It's more likely that they would be musicians first and programmers second, working more out of neccessity than desire.

of course the "song", the "performance" and the overall vision make way more difference than the specific sounds used. and of course the sound-making tools you use don't matter was much as your input into the sound design process.

but you still have to pick sounds and sound sources, so doing that becomes another part of the process, another way you express who you are and the texture of this moment. some people pick up the same accoustic instrument every day for most of their lives, and some shop for VSTIs like snacks.

if a synth is "hardware" based, that just means it requires and (hopefully) benefits from dedicated DSP resources devoted to it. but it's still software that makes the sound, and even in your pc, hardware runs the software anyway.

I guess what I'm getting at is simple: use what you like using . none of these choices are inherently "better", hardware just brings more horsepower to bear, probably in return for a smaller market and therefore more limited company resources. as computers get faster, they're able to do more an more of what it took dedicated hardware to do a few years earlier. when the DX7 came out, you *couldn't* do that in real time on *any* consumer computer.

Interesting thread.

As a Creamware Scope user I would have to agree that the quality of most VSTi instruments come up short by comparison. Creamware DSP is, frankly, outstanding. That being said, the Scope synths are something I use very infrequently. There is simply no comparison between virtual and hardware in terms of what makes music viceral and alive...feel.

Virtual instruments don't have it when you play them. They don't give anything back to you when you dig into them. To me they have a lifeless response to them that makes you feel unattached to the actual instrument, not the kind of thing a player likes to feel. When you try blowing a solo on a VSTi it feels flat and hollow.

I can certainly appreciate that for alot of folks a synthesizer *is* software so how could they know anything different. I can also appreciate that you could own dozens of softsynths for the price of one (reasonably lame) hardware synth. Likewise the convenience of having loads of instruments to hand in the small footprint of a PC is a nice tidy setup that almost anyone can afford.

But as I look around at all the lovely instruments I have collected over the years I can tell you that nothing even remotely compares to the vibe you get wanking on the real thing. The sound is right there.

My favorites;

Oberheim Matrix 12, Roland Jupiter 8, Mini Moog, Memory Moog, Prophet 5 and VS, Juno 106, Roland D-50, WaveStation, Kawai K-5000s, MKS-20, DX-7, Whurlitzer, Fender Rhodes Suitcase, Honer Clavinet, Hammond B-3.

It is sometimes a pain to keep some of this stuff working properly and you royally curse the tuning stability of the old analogs, but when you play them and hear how they sound in a track you don't even care. All the hassle is amply rewarded, then some.

A bunch of us old guys were talking about this whole virtual versus hardware instrument issue at a studio one time and my friend summed it up this way, "When given the choice between using a hard instrument or a soft one, I'd prefer the hard one any time."

The reason is simple...it feels better.

"Hardware just feels better"

OK - here is a thought experiment (us philosophers love these). Suppose that I sneak into your house, open up your Mini Moog, rip out the analog guts, install a small computer inside with a Luna II with the MINIMAX and a potentiometer -> MIDI controller kit so the knobs work, patch it all up to the power module and analog outs, and start up the OS, SFP, and then leave without telling you. Other than the fact that it would stay in tune, I don't think that you would ever be able to tell, and would still say "Hardware just feels better" but would be mistaken.

I will take your point that the interface with the sound engine is VERY important. But where the sound engine physically resides is irrelevant. For instance, consider a knobless rack-mount synth box - How could you love the "feel" of one of these? I see no difference between one of these from a players viewpoint to a VSTi or other softsynth.

And I will also say that, of all of my synths (including hardware!), two of my favorites are Crystal VSTi (v1.4!!!), Saturn, DubSub2, Transplanet, and Modular2 on Pulsar.

algorhythm wrote:

"Hardware just feels better" OK - here is a thought experiment (us philosophers love these). Suppose that I sneak into your house, open up your Mini Moog, rip out the analog guts, install a small computer inside with a Luna II with the MINIMAX and a potentiometer -> MIDI controller kit so the knobs work, patch it all up to the power module and analog outs, and start up the OS, SFP, and then leave without telling you. Other than the fact that it would stay in tune, I don't think that you would ever be able to tell, and would still say "Hardware just feels better" but would be mistaken.

Ohhh...

I would say "Hey, why did my MiniMoog sound better yesterday?" Because the MiniMax is maybe a good model of ONE MiniMoog but not a model of one that I like. I heard different MiniMoogs and they are all so different. Some where not as good as others that really ROCKED.

BTW, I own no hardware synths anymore...

..true, bajongo!

I still have lots of hardware and to give my two cents to this topic:

Certainly there are things that hardware can´t do because it didn´t take that drastic creative development like software over the last years..

..and there are emulations of classic hardware-instruments which sometimes would be very close to the original soundwise...but...what makes the original muc more organic/dynamic/whatever, are the mechanics, imho.

Of course I´m rather talking ´bout fenders, wurlis and clavis, pianets and mellotrons...

To sum it up for my part..my key is to take the best of both worlds and of course there are certain things that you love so much that you won´t get rid of them.

peace, moritz

Moritz Morpheus MkIII wrote:

Of course I´m rather talking ´bout fenders, wurlis and clavis, pianets and mellotrons...

Well said, Moritz!
I think it's rather about analogue vs digital than hardware or software:

Any instrument from an accoustic guitar to a tube amp to a 303 are analogue synthesizers (as in "tone generators") imo. No software will ever be like them, by its very nature.

The thing with hardware samplers, workstations, modern synths, grooveboxes etc. is that they are digital to begin with. In the way of these, computers will outperform them before to long.

My two eurocents

Cheerio,
Marco!

Funny thing is on the (restricted to owners) Creamware forum there's currently a thread called: "The end of DSP systems" which starts with someone describing the amazing power of a friend's P4 2.4GHz running native.

I added a comment about the lack of VSTi integration being a major problem for the platform and drew an unpretty response along the lines of "you amateur". Even more heretical it seemed was to suggest that Creamware incorporate a FUN style wav-export ability akin to Fruity. Gasp! Horror! Fun? This is a serious platform.

Of course there's no need to divide into camps, and I do use both side-by-side, but it certainly gets some people very excited. Another problem is that once you're in the VSTi world and start getting used to its workflow, then anything that doesn't easily integrate starts to feel "difficult".

Ah, well, I'll just go play some music.

PS: Hi Algo I'm still a big Prisma fan (and the ModV2 naturally!)

thx, Sir Bonte!

I would almost blasphemicly state that we're in a time where you're not going to be able to tell the (objective) difference anymore, and soon the software will even sound better.(!)

For the subjective users amongst us, a solution, Just do it like me;
Play your vst synths with an older synth, like (mine) Quasimidi Raven and connect the knobs via "cc-rider" to your host (Logic in this case, haleluja the environment, haha cubase ) and voilà!!! The perfect "retro-synth" emulation! Feels like the real deal, sounds even better....

I think the biggest problem VSTI still suffer from is the developers still have to take into account people running 533mhz celerons.
Ive never been able to get even close to maxing out my xp 1800, and now we have the xp 2600.... its time developers start eating more cpu and make something that sounds better than hardware

telxos

one some points vsti instruments are better to me ...on some points not :

why they are better to me :

less heavy to carry
easy swappable / plug able
no wiring spaghetti when you just want to assign some effects
no paging in very little screens to tweak on some different oscillators or lfo filters

why vsti instruments are worse

for some strange reason the sounds I make in my kurzweil (virtual architecture synthesizing technique) sound better then any vsti out there
the only reason I use vsti instruments and synths is I have better control over effects and I can render it to wave quick and easy without having to wait 3 minutes on recording a track each time. but I must admit I haven't got virtual beauty's like Absynth
I do use the freebie from smart electronix often ...It sounds so dark

One thing .... getting into vsti`s in definitely more fun
you get to try out new free stuff on a monthly basis and the prices on synths are lower as "real" ones

for me a super plus ... my ex girlfriend ones destroyed my hardware studio with a big hammer
the software I have pursued and the freeware I have downloaded is all burned on different cd`s and stalled in a save place on the net that only I can access
the worst thing that can happen is someone steeling my pc ... after buying a new pc I can install everything again
if my studio was packed with synths I wouldn't be able to get it all back so quick.

so lot's of advantages

telxosss wrote:

Ive never been able to get even close to maxing out my xp 1800, telxos

Really? What are you using so far, how many voices on each VST, could you make a list? (I just got myself a XP1800+ with DDR memory, so I'm curious on how things will be with CubaseSX)

Wk

How about the fact that hardware tends to have professional sound designers polishing away at the sound for months, whereas for most of us the flexibility inherent in soft synths means that we have a fairly unprocessed sound to start with and then use our skills (or lack of them) to try and make them replicate what's in our heads.

Although I am all software, I love hardware too and in the long run, as money permits I will have a hydrid studio. Of course there is something about playing the B3, by far my favorite instrument, that no software (or hardware emulator) can ever provide, as good as the B4 sounds, but that said the sound quality of much of the software today is on par with hardware, as for playability, a good controller (imagine a studio setup with a PC2X controlling a number of softsynth) does just that. Add to that the flexibility of software and one is dangerous enough to make music, with feel, love and passion. Software does lack in expressiveness, the DSP etc based features that makes playing leads on the Fantom so much fun, but software is growing rapidly and as the new generation of processors comes out, that gap will go as well.

prog

BONES wrote:

There are some mighty talented programmers out there and I'm sure Steinberg or NI can afford to pay them more than Creamware, given their much larger user base. If anything, I think that the pool of programmers interested in coding for Pulsar would be very small and not necessarily people with good coding skills. It's more likely that they would be musicians first and programmers second, working more out of neccessity than desire.

well, partially true, Bones.

Creamware is a small company, yes. But so is Native Instruments. and Steinberg hasn't made very many VSTi's have they? And there new audio-synth was developed in conjunction with people outside of Steinberg .. . .

I certainly hope that the coders are both musicians AND programmers. We all work out of necessity no?

some examples that are fit here:

The best first example is MiniMax - the best analog emulation ever. Nothing in VST is purported to be as good.

Better yet, lets take your example of NI - the NI Pro-52 was designed, at least in part by John Bowen (If you don't know who he is, ). Now, John also happens to be a Pulsar developer (helped design the Modular, and has bomb as synths @ www.zargmusic.com ), and developed the Sequential Circuits emulation on Pulsar. With the help of Creamware dsp coders, they actually modelled the OSCs, filter, and Env. of the Pro hardware. John had said that NI was looking to lower the CPU use and ended up doing some compromising. No compromise on Pulsar.

So yep, there are good coders for both ( the same in some circumstances ). If you pay a hefty tag, you get the nice VSTi's. If you buy a Pulsar, you get nice synths for free . . .

Poweruser wrote:

for me a super plus ... my ex girlfriend ones destroyed my hardware studio with a big hammer the software I have pursued and the freeware I have downloaded is all burned on different cd`s and stalled in a save place on the net that only I can access

OMG!

@ the guys that say that hardware synths are better and oh all I got is a free little vst-i...try Absynth, Fm7, ES2 etc.. and will talk again....

If I could find someone to buy my KORG TRINITY, I would sell it in a heartbeat. Hardware synths are just a waste of money. I guarantee that no-one who hears any of our music could tell whether we'd used VSTi or hardware V/A or classic analogue gear and if you think otherwise you are having yourself on. You may enjoy using one thing over the other but, at the end of the day, that doesn't mean one sounds one bit better than the other [pun intended].

I still say what sounds better is whatever gets the job done. I have loads of hardware and software, and I gotta tell you, it doesn't matter if you think an analog beastie has some elusive sound qualities when, say, the cables are humming from radio signals coming from the nearby airforce base (most hardare synths STILL don't use balanced cables!), or the volume pot you've cleaned a thousand times and replaced twice STILL introduces crackling to your killer solo! VirSyn Tera NEVER crackles or hums unless I tell it to. I never worry about faulty cables with Pentagon 1 or Delta III. Say what you want, but the sound that gets my inspiration down before it dissapates is the one that sounds the best to my ears. Nothing else counts.

I'm with Scot on this. I remember the first time I was in a studio. The engineer kept telling me not to use my "cheap" Electroharmonix delay and chorus and wouldn't allow me to use my Orange amp and 4x10 cabinet. Instead he insisted I used his tweaked Fender Twin which totally sucked for what I was doing.

We argued for a long time and finally agreed to do some takes with my setup and his. The setup with my so called crappy equipment made the mix.

Sound is too subjective to state what is quality and what isn't. Some of my favorite artists record lofi and exploit the lack of sound quality to enhance their artistic aesthetic. And it sounds good to me.

Bruce

I own both Absynth and a Kawai K5000s (a VERY underated additive synth). Editing wise there is no argument that Absynth is much easier to use, BUT i always end up going back to the K5000s before i've completed the track. On the other hand i always end up going for FM7 over the DX7. It comes down to individual sounds as always.

Another thing to note: soft synths are nice in the studio but i'm not prepared to trust my pc on stage yet.

What gives a particular synth it's expresiveness, or lack there of, is the resolution of it's envelope generators and response of it's modulation matrix. In particular the degree of resolution, in the so called attack stage of the envelope proper, is largely responsible for the very critical initial feedback the player gets from the instrument. It is what ultimately determines whether or not he\she enjoys interfacing with that instument from a performance stand point.

This is a weak spot in many softsynths, for very obvious reasons...the host cpu is still way too wimpy to do an adequate job of rendering a superb *sounding* ocillator, built in FX, and a comprehensive modulation matrix, while also providing advanced high resolution envelope generators to adequately shape the sound in a moving, musically expressive fashion.

DSP systems may be better but they need to be huge...very, very huge, to really nail the *total* vibe. When you start firing 16 multi ocillator notes, all distributed to complex, evolving modulation matrix assignments, all having fine enough degrees of resolution to interpret the human nuances of expression, you need very large scale processor technology in very large quantities. We routinely bring a *very* well specified Scope DSP system (floating point Sharc) grovelling to it's knees, and though it is a quite impressive sound in general, it cannot match the utter mass, smoothness and expressiveness of something like the Oberheim Matrix 12 when playing back complex modulation routings. It is also immediately apparent that there is much more openness and depth, more kick and cut if you will , to the sound. There is no reason for this not to be so, analog VCAs are not emasculated by the impositions of sample rate limitations.

When systems get bigger, which they will, we can expect to see better playability from software instruments. For some of us, especially the old codgers, this will be none too early coming because we want the sound to seem like it's coming from our hearts and fingers. The B3 is a marvelous instrument for that, you're attached to it by the ass, just the way you should be. The old modulars were great. You literally ran around shoving patches in here and there, twisting these big lovely knobs, caressing ribbon controllers...often all at once...while singing. The range of sounds you could mold was amazing. It was a fantastic thing. It made the action so much hotter on stage and in the studio because you never really knew what was going to happen. Sometimes you would accidently patch something in totally wrong and the whole synth would start shrieking and complaining...going totally wild. The engineers would be falling all over themselves for the fader to the amps, the monitor tweeters would go into protect mode, or simply blow. We loved it.

If all you have ever played with is virtual intruments you can't possibly know what you are missing.

[kilroy] If all you have ever played with is virtual intruments you can't possibly know what you are missing

no offence intended at all kilroy . . . but if you cream over your hardware so much . . . what are you doing at k-v-r ?

. . . just curious . . .

. . . cheers rob

xghoti, I also have a K5000 (I love that thing), and in this case (and the case of the Wavestation), I have not found a suitable VSTi replacement. This is largely down to the fact that no-one has developed an additive VSTi with the kind of power and quirkiness of the K5000. Certainly, a subtractive synth such as Absynth won't pull off the subtle delicacies of the K5000. There are some good additive ensembles for the VST modulars, but they are not quite there with the K5000. I'm afraid that history has shown time and again that there is little profit in additives. They require too much from the user to become popular, I suspect. I think it COULD be done, and easily, but I fear no-one would want to put the effort into it. It's a shame. It's been a dream of mine to have an additive VSTi with good FFT resynthesis, sort of like an expanded version of what you get combining the K5000 with Sound Diver...

As I have stated, I have a roomful of "classic" hardware stuff, and I am VERY aware of what I am missing when I use VSTis. I'm missing:

1. the 3 dB signal loss at each end of a cable
2. the noise from the mixer/bad components
3. oscillator drift
4. crackling pots
5. ugly cable spaghetti
6. the lack of total recall
7. poor integration into my host environment
8. a polyphony ceiling
9. nicad and lithium batteries going dead
10. dusting around the pots and sliders
11. limited waveform ROM

. . . but oh so true . . .

. . . cheers rob

"no offence intended at all kilroy . . . but if you cream over your hardware so much . . . what are you doing at k-v-r ?"

Fair enough question, pHz. In addition to having been fortunate enough to acquire many wonderful hardware instruments over the years, I also have a good smattering of VSTi and DXi stuff in addition to a Creamware Scope system. I basically drop by forums occasionally to troll for helpful information, just as you do, I would suppose.

For example, I never realized that there were so many free VSTis out there, some of them quite good. Crystal certainly cannot be called cheap sounding and the developers actually seem to appreciate the importance of well designed EGs. It sounds better than some I paid money for and never use. Finding this one VSTi made the stop by worth while, I figure.

I do most of my composition with the aid of computers, though I still have not found anything faster or more intuative for actual scoring than a pencil, stack of staff paper and a empty table...oh, and lots and lots of coffee, Irish Cream, and when it gets really late, the darkest Columbian I can find.

So you see, I *do* have an interest in things virtual, I do not have a closed mind in this regard, I just have an affection for real physical gear because that is what I cut my teeth on live for so long and I find to sound most pleasing to my ears.

Quote:

OMG

Yeah Moritz,

You probably have heard some guys at KVR talk about my craziness
Well at least you now found out where part of it comes from.

The software we use now will be overtaken by faster / better sounding and renovated stuff. But without thinking about that, the software I use at the moment will still be on a cd, 100 years from now, when time killed my hardware.

This is a point that none can deny

staff paper

(making mental note for self..)

true!

BONES wrote:

To get back to the original thrust of the topic, I think that any perceived difference probably lies in the hardware side of your set-up. Your Pulsar card probably has much better AD/DA converters and a good pre-amp when compared to the soundcard you listen to your VSTi through, so naturally it will sound better through your speakers.

What a very odd statement! Don't you think Pulsar users are listening to their VSTis through their Pulsar card?

I would find it really strange if someone went to the trouble and expense of buying a Pulsar card so that they can listen to their VSTi out of a SoundBlaster Live.

Maybe I just don't understand what you're saying here.

Caleb wrote:

Maybe I just don't understand what you're saying here.

Bones, like me, does not have a Pulsar system and might not have any idea how you use VSTi's together with your Pulsar system (I didn't untill you explained it to me in the "how do you work...." thread, remember ).

Quote:

The software we use now will be overtaken by faster / better sounding and renovated stuff. But without thinking about that, the software I use at the moment will still be on a cd, 100 years from now, when time killed my hardware.

Don't forget that CDs also have a limited lifetime

. . . and you'll need some *hardware* to play it back on

Caleb wrote:

BONES wrote: To get back to the original thrust of the topic, I think that any perceived difference probably lies in the hardware side of your set-up. Your Pulsar card probably has much better AD/DA converters and a good pre-amp when compared to the soundcard you listen to your VSTi through, so naturally it will sound better through your speakers. What a very odd statement! Don't you think Pulsar users are listening to their VSTis through their Pulsar card? I would find it really strange if someone went to the trouble and expense of buying a Pulsar card so that they can listen to their VSTi out of a SoundBlaster Live. Maybe I just don't understand what you're saying here.

Actually, it was I who had no idea that you could play your VSTi out through your Pulsar card. My assumption, based on vague memories, was that you couldn't route other internal things through it.

BONES wrote:

Actually, it was I who had no idea that you could play your VSTi out through your Pulsar card. My assumption, based on vague memories, was that you couldn't route other internal things through it.

That's one of the best things about the Pulsar. The ability to integrate different software programs as well as intergrate external hardware.

kilroy quoth What gives a particular synth it's expresiveness, or lack there of, is the resolution of it's envelope generators and response of it's modulation matrix. In particular the degree of resolution, in the so called attack stage of the envelope proper, is largely responsible for the very critical initial feedback the player gets from the instrument. It is what ultimately determines whether or not he\she enjoys interfacing with that instument from a performance stand point.

I have to take issue with that. To define 'expressiveness' solely in terms of resolution of modulation sources, let alone give significance to the attack stage, seems to be one of the strangest attributions I have seen in a long time. Firstly, there seems to be a blanket assumption on your part that for some reason hardware synths use higher internal resolution for their envelope generators than modern VST's. That's an unfounded situation, and I would actually expect the hardware synths, (which are often a combination of dedicated DSP or custom gate array doing the audio rate calculation, and an embedded controller doing the controller rate calculations) to have a lower resolution than an all-floating-point caculation in a VST. Resolution is range of data by the way, not rate of data, perhaps thats what you mean? Perhaps you do, although assuming that the control rate is faster in hardware seems spurious as well.

As for your secondary comment about resolution of the attack stage, again that doesnt make sense to me. Is a violinist's 'expression' bound up in their first contact with the string? Nope. A guitarist with the 'strum' or 'pluck'. Nope. Availability and controllability of modulation sources and destinations, are what make 'expression', not the initial attack. In many instruments the initial attack is indistinguishable from a noise burst anyway, and its the 'resonations' with the rest of the physical instrument that make the sound. And modification of those resonators that causes expression.

This is a weak spot in many softsynths, for very obvious reasons...the host cpu is still way too wimpy to do an adequate job of rendering a superb *sounding* ocillator, built in FX, and a comprehensive modulation matrix, while also providing advanced high resolution envelope generators to adequately shape the sound in a moving, musically expressive fashion.

What makes you think that the single DSP in many hardware synths is more powerful than a modern CPU? They're not. A 2 Ghz processor is capable of far more processing than many dedicated DSP chips. Even the new VST DSP boards like the Powecore and UAD1 run 6 or 8 processors and claim to get the equivalent speed of a 2 Ghz G4 (their own advertising) on the whole board. Im sorry, but that means that a single 1Ghz G4 is equivalent to 2 or 3 at least of those DSP's. The fact is, though, that they also have to run the operating system, the sequencer host, half a dozen synths and fifteen effects simultaneously, so VST writers typically dont write software thats going to cripple the rest of the machine. Thats nothing to do with the CPU being 'wimpy'.


DSP systems may be better but they need to be huge...very, very huge, to really nail the *total* vibe. When you start firing 16 multi ocillator notes, all distributed to complex, evolving modulation matrix assignments, all having fine enough degrees of resolution to interpret the human nuances of expression

Hold on, wait a minute. This is another fallacy, I suspect, or at least misguided. What actual 'resolution of nuance' do you think an 'expressive' human is actually capable of expressing, or even detecting? The vibrato of a string under a violinist's finger actually falls withing a remarkably small range, there are physical and mechanical limits to what can be achieved. But those limits define where we 'look' for expression. However there's nothing in those ranges which is too 'high-resolution' for a modern CPU to achieve, and most of the modulation is single source anyway.
Another obvious point here is that polyphony and 'fine' expression are effectively contradictory, or at the least, self-defeating. Soloists express, but they dont use 16 notes to do so, because that expression would be lost. Note that 'complex timbre' is not a synonym for 'expression'.

, you need very large scale processor technology in very large quantities. We routinely bring a *very* well specified Scope DSP system (floating point Sharc) grovelling to it's knees, and though it is a quite impressive sound in general, it cannot match the utter mass, smoothness and expressiveness of something like the Oberheim Matrix 12 when playing back complex modulation routings. It is also immediately apparent that there is much more openness and depth, more kick and cut if you will , to the sound. There is no reason for this not to be so, analog VCAs are not emasculated by the impositions of sample rate limitations.

I feel that here you are seeing an effect, and assuming the cause. That may not be the case. All sorts of other factors may be involved from memory limits, memory-bandwidth limits, the board architecture, limits in the

Do your comparison across three or more DSP platforms, and a general-purpose CPU. Only then can you make any reasonable judgement as to the cause.