Here's a quick review: The 24-bit VGS samples are clear as a bell and have the characteristic VSL sound. Running these samples through GigaPulse make getting a good sound ridiculously easy. I haven't had any desire to add even a hint of EQ or compression. I mess with the volume a bit - but mostly just to compensate for the instruments that I have quieted by playing the expression controller.
The articulations are limited - just staccato and sustain for most instruments. I haven't given much thought to the number of velocity layers, so no comment here. The strings get more articulations: add pizz and tremolo. All in all it's enough to do composition and orchestration. You'll want more (Opus, Cube...) for those oh-so-subtle exposed melody lines.
I can load the entire 24-bit VGS lib with one instance of GigaPulse and the Bos 290 demo piano in about 2-1/2 minutes. It takes 47% of the memory of my 1.5 GB machine. It's the best sounding single-computer, orchestral scratch pad that I've heard. My wife commented that my recent doodlings sound more realistic than any of the stuff I've done before with my hodgepodge of samples.
I also own GPO. I much prefer the VGS sound overall. It's more airy and magical to my ears. The thing missing in VGS that GPO has is the great mod-wheel expression filtering. Maybe I just haven't found the hidden tricks in VGL. I have yet to open it in the editor. I hope to be able to do some tricks with the parametric filters to be able to add some pseudo-GPO expression. We'll see when I can get around to it.
Also included is one octave of the clarinet and one octave of the Viola Ensemble in VSL legato form. These sound wonderful. They make it worth composing between C4 and C5 if you want great legato on the cheap.
Overall, I plan on using VGS exclusively for basic composing and orchestrations. I'll add other libs when it's time to take the performance to the next level. And I'll do multiple passes of GigaPulse when I want the mix just so.
It's great to have such a great sounding scratch pad with all the sounds ready to go on a single PC. But don't let the "scratch pad" thing sell this short. If your melodies don't require any tricky articulations, you can do the whole performance with VGS and fool most any amateur into thinking they are hearing a real orchestra. If you're really good with your touch and expression control, you might even fool me.
Thank you SOOO much for that great review. I really appreciate it.
I use QLSO Gold right now, and I'm curious about getting the upgrade and trying the VGS stuff to see if its worth getting Opus immediately for what I'm trying to do (Film / Large Orchestral stuff, as opposed to smaller works)
Its guys like you who make this forum the excellent informative place it is.
As a VSL user I agree with most of what's been said, but do remember that what you are getting with GS3 is a tiny proportion of what is available. As a relatively new GS user I have only just started to unlock the editing power, but I have found that filters, release times etc. all controllable from the keyboard are very easy to programme.
However, be aware that other formats of VSL are in production, so if you're happier with a sampler other than GIga, then you might wish to wait. Also I believe that Gigapulse will shortly be released as a plug-in, so you won't even miss out on that.
> "As a VSL user I agree with most of what's been said, but do remember that what you are getting with GS3 is a tiny proportion of what is available."
So true. Only having staccato and sustain works in some situations, but can be lacking. I expect to layer the two at times. For instance, the attack on the string sustains is very slow, so you can't do legato with long durations at all. Layering the two articulations should work, but I haven't tried it yet.
The articulations in Gold (according to the list - I don't own it) are much deeper than VGS. From a keyswitcher's point of view it would be a step backwards. I won't compare the sounds. That's too subjective.
> "Also I believe that Gigapulse will shortly be released as a plug-in, so you won't even miss out on that."
Yeah, GigaPulse VST. I'm not sure when that will be released. I'd hope that Tascam will have a better feel for their schedules after the experiences of the past year or so.
Like I said, I already own GS2.5 (not even installed atm), and use QLSO Gold as my main 'orchestra' package. I'm just considering getting GS3 to test VSL pretty much. Also, Gigapulse would be nice, and sometimes its just nice to have GS kicking around. Assuming it will play nice with sequencers on the same machine now
> For instance, the attack on the string sustains is very slow, so you can't do legato with long durations at all. Layering the two articulations should work, but I haven't tried it yet.
How on earth are you supposed to get legato at all if the long durations have attacks which are too slow and the staccatos are obviously too short for many passages?
Layering the two does not sound too promising to me either. After all, the sound which we are mixing with the sustained sound - the staccato - is exactly the sound we are trying to avoid in order to get a legato feel.
I also own GPO. I much prefer the VGS sound overall. It's more airy and magical to my ears. The thing missing in VGS that GPO has is the great mod-wheel expression filtering.
What does the mod-wheel expression actually do? I read that it controls the volume AND the expression. Sounds a bit strange to me, but that is probably so to make it as playable as possible. Personally I don't really need playability at all and would much rather control all of the parameters separately, but that's just me... And what I gather these controllers can not be mapped at all in GPO.
But what does this expression do to the sound? Are there different layered samples which are controlled with the mod-wheel or what is it? And if so, then how are these samples different.
The Garritan Personal Orchestra expression controls timbre and volume together by using filtering. Instead of multiple velocity layers, there is only one.
I think it is programmed like this: there is only one sample for each note. It is ff and has a hard attack. The samples are many seconds long, then have multi-second loops. I don't think they use releases.
Attack is controlled by velocity. If you play hard, you get the natural attack of the sample. If you play quietly, the sampler player applies a slow attack.
Volume is controlled by the mod wheel. When you push the mod-wheel to it's limit, the sample is unfiltered. When you reduce the volume with the mod-wheel, a low pass filter is employed. This has a similar effect to a crossfade as you might find in Project Sam Horns. The advantage is that there are no phasing issues, so it can work on solo instruments, and it is very efficient.
When I use the expression control, I'd like to have some filtering along with it. To simply change the volume sounds fake to me. When the timbre changes with volume, it sounds more lifelike.