How come "Hardware" VA's 16 years ago sounded so good like the AN1x and now....

[Haraldator]

Was this Lush? Please tell me this was Lush ...
(I know you're a beta, you lucky man)

No it was SynthMaster actually, but Lush will do something similar, though not quite the same:

http://soundcloud.com/aiyn-zahev/lush-101-bit-reduction

Woah, that's a LOT more than I use.

[Uncle E]

In the 90s the hardware companies had analog gear laying around to study, disassemble and with documentation! Plus, they bought up the analog companies that went out of business in the 80s (and their patents!).

Excellent point!

It's not like just anyone can make a good synth, you gotta have big brains to think it and good ears to evaluate it.

That sounds like u-he! GForce is another good examples, I don't think they do their own programming yet their synths are successful out of Chris's and Dave's unquestionable expertise with the original instruments.

Korg's still getting milage out of Dave Smith with updates to the wavestation in the Kronos and they're also still licensing some patents from Yamaha, I think!

Now I want to spend some time with a Kronos! Did those updates to the Wavestation make their way into the Krome?

Bottom line though, if you can't make modern software work for you, it's not the software (at least not the sound quality of it).

Well said!

[Uncle E]

All internet arguments sound pompous one paragraph at a time.

Beware the quoth!

[Uncle E]

No it was SynthMaster actually, but Lush will do something similar, though not quite the same:

http://soundcloud.com/aiyn-zahev/lush-101-bit-reduction

I quite enjoyed that!

D16's Decimort is wonderful for things like this. I'm glad to see it built-in to the LuSH-101.

[Uncle E]

Various people on KVR have reported liking the sound of their softsynths better simply by sending the audio out through an amp and back into their computer.

Eric Beam did a Nebula impulse of a Fender Deluxe Reverb and has audio examples of synths going through it:

http://rhythminmind.net/STN/?p=1765

Personally, I preferred the clean signals every time.

[grillo]

To me the Novation V-station vsti comes closest to that kind of VA sound. It has a certain kind of sharpness that reminds me of the classic sounds from the novas, the nord lead and even the virus.

I never found the virus to be that warm or fat honestly, especially the Ti or even the C imo, granted i never owned an older one.

There's something else that i'd like to point out here though, both the Virus and the Nord Lead (i chose those because i have direct experience with them) have pretty sophisticated synth engines, sure their osc and filter models might be outdated compared to today's softsynths, but they had a lot of extra bells and whistles that went beyond the old analog emulations.

For example the Nord had the morph system which made for very expressive patches. That sounds pretty easy to implement in software but not many virtual synths have it. Then it had layering, basic fm, different filters. The TI has basically multiple models for every stage, oscs, filters and envelopes, loads of modulations and effects.

There should be a big test with some very simple sounds (ie bare oscs, open filters, or classic moog bass with resonance, whatever) to understand if it was more a question of those layering tricks, fx and clever programming or it was just the bare sound.

One of my fave vsti is still the fab filter one because i love its filter and with the single osc the sound never gets muddy.

[Aiynzahev]

There should be a big test with some very simple sounds (ie bare oscs, open filters, or classic moog bass with resonance, whatever) to understand if it was more a question of those layering tricks, fx and clever programming or it was just the bare sound.

I've thought about this. It would have to be set up and controlled very well, and probably all by one person so its the same A/D.

Here is something though I did with some soft-synths:



A. Saurus
B. Saurus with "boost" on
C. Diva (Moog)
D. Diva (Dual VCO saw)
E. ElectraX on Linear
F. ElectraX on Analog
G. Sylenth
H. Strobe (8 time over-sampling)

I did not include the V-station or Largo there, but I have those snapshots too.

They look like Sylenths saw, but they have that extra spike you can see in Diva and Strobe, though maybe not quite as pronounced. Largo and V-station looked very similar in the saw-tooth.

[optofonik]

Using a fast modern multicore CPU, someone should post two versions of the same song. One made with new VST gear and another with old VST gear. The synths, old and new, would have to use the same patch parameters, same for FX. I would be willing to bet that if the song is a good one; one that is compelling, catchy, dope, sick, whatever, and all things being equal, it would be difficult if not impossible to tell the difference.

Imagine today's computer running yesterday's solid, good sounding, VST gear. No more CPU overloading, plenty of tracks to go around, plenty of synths to go around, seemingly endless effects and processing, low latency; it would be what we hoped for. Ten years ago we wanted more powerful computers to run the VST gear we had. We got the powerful computers but to what end? Writing more complex code that uses up all the power we've gained doesn't make our songwriting any better and it certainly doesn't make it easier. We've lost the plot entirely.

I want VST gear that sounds good in a mix and doesn't monopolize my CPU. I've upgraded quite a bit this past year and purchased a few newer well reviewed synths but I find myself using more older or free ones. DCAM, Harmor, and Circle are examples of synths I bought that I don't use much because I have to babysit their CPU usage. DCAM sounds amazing but in a mix I want it to sit and behave well.

Now, having said that, there are synths and FX today that can do things the old ones can't but I'm limiting my opinion to so called emulations of hardware synths and romplers.

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

As a music major I was required to be in an ensemble, a choir. I shared first and second tenor duties with my friend, Rick. We were amazing together, or so we thought. We could blend our voices so well that the volume of just the two of us could overpower not only our fellow tenors but the other sections. So, we were, in fact, quit amazing. Problem was, it got in the way of the overall performance to such a degree that our professor pulled us aside and told us that although he admired what we could do we were to, literally, tone it down.

[Uncle E]

Using a fast modern multicore CPU, someone should post two versions of the same song. One made with new VST gear and another with old VST gear. The synths, old and new, would have to use the same patch parameters, same for FX. I would be willing to bet that if the song is a good one; one that is compelling, catchy, dope, sick, whatever, and all things being equal, it would be difficult if not impossible to tell the difference.

LOL!

I think you're trying to say that a good song will be a good song whether it's on old VST's or new ones. However, if you really do mean "impossible to tell the difference", I will literally bet you any amount of money you have in your bank account that I can tell that difference.

Show me the money!

[pdxindy]

Various people on KVR have reported liking the sound of their softsynths better simply by sending the audio out through an amp and back into their computer.

Eric Beam did a Nebula impulse of a Fender Deluxe Reverb and has audio examples of synths going through it:

http://rhythminmind.net/STN/?p=1765

Personally, I preferred the clean signals every time.

I am entirely happy with ITB sound quality... todays softsynths sound gorgeous!!

[optofonik]

Using a fast modern multicore CPU, someone should post two versions of the same song. One made with new VST gear and another with old VST gear. The synths, old and new, would have to use the same patch parameters, same for FX. I would be willing to bet that if the song is a good one; one that is compelling, catchy, dope, sick, whatever, and all things being equal, it would be difficult if not impossible to tell the difference.

LOL!

I think you're trying to say that a good song will be a good song whether it's on old VST's or new ones. However, if you really do mean "impossible to tell the difference", I will literally bet you any amount of money you have in your bank account that I can tell that difference.

Show me the money!

In a good mix? LOL, right back atcha.

[tapper mike]

This is going to take awhile to explain some might just want to skip my post.

Ya know it's funny. When I started listening to mid's in the 90's with over quantized tracks and a crappy soundcard.... I loved it. I loved it for the sake of the music and the fact that someone put a lot of effort into trying to recreate the song as much as possible. It was corny sounding and cheesy and I found myself enjoying the .mid more then I would have the original. It beckoned me back to a bygone era where having fun and coming close and improvising were more important then accuracy. Midi recording got better, Sound Cards and audio got better. Transcriptions became more accurate and with the advent of Guitar Pro and RSE engine I was floored with the sound of being close as midi possible to the real thing.

I don't casually listen to mid's anymore. I don't casually listen to guitarpro songs. Though I do have a great deal of respect for those who compose in gp. When I want to listen to music I go to youtube.

Sure a new sound that's alive and fresh can be a source of inspiration to perform well but it is the performance not necessarily the quality of the instrument that matters the most. It's the performance you bring to it. And upon listening to a song the same holds true. If the song takes you to a special place it won't always be on account of the sound quality but more over your ability to enjoy the song for the sake of the song.

[osiris]

That's very interesting Aiyn. More comparisons please. Visual doesn't lie, as opposed to what people hear, or think they hear (or want to hear)
Out of curiosity, since people say AD conversion affects the sound, what did you use to generate the above pics? (I.E., DAW, etc....)
Do they look different in a different DAW? Do they look different coming through different soundcards??
It's some interesting questions.
Also, could you do some compares between VST's and hardware?

[gunnare]

The fashion at the moment is high CPU synths. People want high CPU. They will not use a low CPU synth because they will think that it sounds bad. The message to developers is clear.

I have been working on oversampling lately. I have big problems to hear any difference between oversampled and not oversampled. With 8 times oversampling you increase CPU w 800% and the sound becomes ca. 2% better. FM modulation is much smoother at 8X oversampling, but many times I prefer the raspy powerful FM sound from 1X oversampling over the smooth 8X oversampling sound. For simple patches it is very difficult to hear the difference between 8x oversampling and 1X in a blindtest. I think the biggest difference I hear is in the envelopes. The envelopes seem more powerful at 8X oversampling.

You can not see the waveform sound in an oscilloscope, because you might get fooled. A highpassfilter will change the shape of the wave, but it will still sound the same. The minimoog saw looks rounded and soft in the oscilloscope, yet sounds raspy and sharp. My old Roland JD990 has sampled minimoog saws, and sounds more minimoog than any softsynth I have tried. I believe it has got 44.1 samplingrate and 18 bit converters. If you imagine that that soundengine was oversampled 50X it would sound a bit smoother, but not much different in sound.

[osiris]

I don' think this is the fashion, I think it's dev's getting, or wanting, to get ahead of where technology is headed. This happened to me in SE. One synth I made was blowing up CPU. Bought a new computer, same VST clocked in @ 20%. So, I think it's all relative.
I also think most devs work on really high powered computers and forget about Joe Blow who has one core and 2 Gb of memory.