Okey some of what i'm about to say is personal experience and personal experience only, you can choose to agree or to disagree.

So i'm trying to make the jump from Vst 2 to VST3, tried something similair in the past but i took a step back, because at the time (few years back) VST3 wasn't as stable as now.

So from my past experiences here what i found, VST3 has actually better sound quality compared to their vst2 counterpart, some V3 actually took less time to load, while other took more, but what was the deal breaker for me was on some of my main plugins (the ones i usually use all the time) is consuming a lot more CPU, don't forget that some plugins do actually uses less (always comparing to their VST2 counterpart) but some of my main do considerably consume a lot more, and that's what i based my decision in the end.

so, what are your experiences and do you actually recommend it or not?
(i'm after your experiences not random links here and there)

 

Yes and no some are good. and to use ara you need vst3.

 

I doubt that. How do you measured the better quality ?

 

applying same FX on the same source, with the exact same setting, this is not just with a single plugin happend with a few actually, as you i had doubts i thought that it was a unique situation, and to my surprise it was not...

 

I prefer 2 over 3, this way I can install the dll where I want to.

 

I'm not saying it doesn't happen, but in theory that shouldn't be possible.
A bad implementation could explain the differences (VST3 new features aside) with the performance and loading speed, but better sound quality...

 

Cool thread! Although rigorously discussed before such as here.

I use VST3 for the following reasons.

In summary:
- plugins are able to shut down as soon as there's nothing to compute, thus saving CPU power (massively, under certain circumstances)
- they are always 64bit compatible
- they are able to scale to a multi-channel track-format automatically
- side-chain options (!!!)
- dynamically assignable outputs in case of VSTi's
- more that one internal MIDI-port in case of VSTi's (... something other formats are missing dreadfully)
- automation options are more consistent, the automation itself is sample-accurate

 

VST3 does have some basic advantages and it is probably worth it to transition soon or even now for the trailblazers among us.

The main advantages are the dynamic inputs and outputs, meaning no more mono, stereo, surround version of each plugin you use VST3 adapt itself to the channel it was inserted on (its not that big of a trick but making each channel its own process and expandable is good practice and abstracts away that decision from the plugin-developer to the DAW.) A little bit better memory management and processing efficiency in that a VST only actually runs complex calculations when a signal is present not is sleeping when no signal is present. Native plugin bussing + sidechains and the ability for sample accurate automation with XML and multiple midi ins and out)

With all that it's kind of funny that Cubase which has had VST3 before anyone else sometimes filter out some VST3 best features on different price points even though they are coded in by default as part of the standard

Meh

 

in that case, it's probably Image-line's fault, even thought i have nothing to base my claims on, i specifically mentioned that this is solely based on my own experiences.

 

I never really notice anything like that, mostly because when I switch from VST2 to 3 is with an update (I had issue with stability using vst3 before with FL11, but since the v20 no problem anymore - was probably fix with the 12 but didn't try to switch at that time).

The difference you ear with quality, is that on playback mode or when you compare 2 export of the same sound ? If it's with export, what's left when you do a phase inversion with the 2 version ?
In play mode it can be a difference because different CPU/ram usage, but plug should sound the same (when using same version) at the export

 

Using more CPU is not an indicator of quality. The devs usually wouldn't bother with different code paths for different deliverables.

 

while the other post is more like "on paper" this post will be about real life situation and real life experiences and advantages vs disadvantages, (the reasons you mentioned for exemple...)

 

in play mode, won't lie did not bother to do the phase inversion trick...

 

using more CPU has nothing to do with quality, i was talking (in general) while some plugin used less some used more, a total separate from the sound quality.

 

I'm partial to believe OP regarding sound quality difference. If you're using 24bit size headroom which is massive, and the differences are so subtle in sound quality, how can you be sure doing the reverse polarity trick would bring up an audible difference? The audio could be so low in the test that the 1's and 0's are microscopic no? There will be a wall of rebuttles but I understand the frustrations of being told what I hear is wrong lol.

 

I think a lot depends on your host but IMHO vst3 is now well imbedded in the industry. I run Cubase so if I have the option of vst3 I use it, as it has clear advantages over vst2 that Giggity has pointed out previously. If only NI, Soundtoys and Spectrasonic would get their act together I would be completely VST3. As for sound quality, I could not say I've noticed any difference.

I think, correct me if I'm wrong @No Avenger, but the reason No Avenger talks about choosing where to put a plugin is quite valid as some hosts don’t provide an effective plugin manager to sort them, you must rely on building a file structure that helps you organise the .dlls, therefore its quite important to place the .dlls in specific locations. vst3 just puts them all in one folder so in some hosts you have no way of sorting them, you just end up with a huge list of random plugins. Cubase has a fantastic plugin manager that you can build multiple organisational directories so not a problem in this particular host.

 

newer plugins works great some older have issues in my case, plugins builded before 2016
some plugins as u-he is still work in progress, but close.
Regarding sound quality did't spot any difference yet, my guess is same code is ported just to be compatible
with newer standard, second and I don't have expensive studio with uber-fine-tuned acoustic and my ears are damaged from clubs anyway... in short works for me

my only wish is to all daw someday maybe 2056 izolate vst plugins from daw, somekind sandboxing, so no plug-in is able to crash the whole project, or even stop the project’s audio.

so for now we are all beta testers, my guess next 2 years

 

I didn't know but this seems an insane Steinbeirg fault. Could you elaborate more please? Or links or whatever...

 

hhmm

 

Last edited: Apr 13, 2019

It's not a fault its just business. Cubase and Nuendo's mixer's and midi piano roll are native VST3 extension with backward compatibility for VST2 so when a lower tier version of Cubase doesn't offer sidechains, note expression or over 2 channels of audio per channel it is not just because those features have been switched off, they have been coded around to be switched off.

I am not begrudging them wanting to make money on more expensive versions, it's just strange that the way the standard works they have to compile a different version with those features turned-off which is more work writing routines not to allow something than just letting the mixer work as it was intended.

This idea of feature full product with deactivated portions isn't unique to Steinberg it just seems like more work than it is worth but being able to run a sidechain isn't that advanced of a feature but I guess it carries so much perceived weight that it's worth reserving for the well paying clients even it require more work to just lock out the little guy.

 

Last edited: Apr 14, 2019