Where are we headed?

[nikolas]

I'm not sure if this has been discussed before. It certainly must have been, but still I can't find any link and well, it seems very interesting to me.

I have aboslutely no intensiong of starting any flame war, or fight or anything. This is about philosophies and the future!

So:

It appears to me that there are two completly different tendencies about music at this time and age.

I. The sampling one, with as great detail as possible. Pianos are the example. "I" is 45 Gb for 2 pianos!. "V" is 55 GB for 1 piano and "E" is 150 GB for 4 pianos! ~~~~ (<-will be censored right?) this is A LOT! I mean even with 500 GB disks these take half the disk more or less (in the later case). But then again with 64-bit and 128 GB RAM, plus the new Disks that can hold 3.9 TB (3900 GB) things are tiny again. Or maybe not? Speed is moving really fast as well with Quad cores etc... But is this the solution?

II. Then again you get the exact opposite at 45 Mb (Mega, not Giga) and the "T" library for a piano. Or 65 Mb for a whole orhcestra from "S"! And of course the middle way, to whome we owe this forums Gary, who is doing what appears to me to be revolutionary and perfect in any way, combining the two, with a great Symphonic orchestra at 3 Gb (or is it 4?), and a strad with unparallel realism at 1 Gb (but required less that 100 Mb of RAM which again is great)!

So? what is the way of the future?

Because in my search I have found that bigger and more expensive does not necessarily mean better. Although to be honest in most cases you get what you pay for, and this apply to everyday life things, like super market shoping and whatever else...

But then again I cannot but be greatly appreciative at the strad 2 (which I have), and for example the totally indie Upright Bass by gregjazz (who just came out from the shadows here in the forums), and costs just 30-40$ but sounds just GREAT!

For the record: I have Ivory as my main piano, but can't help but feel overwhelm with the new libraries, and kinda small... then I try to use it and it just... crumbles with my 2 GB RAM and relatively slow hard disk (which is at fault definately!)

[teebones]

I hope i'm understanding your discussion point, if not, ignore this post :D

The way will be a mix of both sides.
More realism pushed into smaller size. Still the big libs will still exist.
This is because, you have realism, and well eh, realism.
There is allway the need for the perfect imitation/sample in certain cases. That's when the big boys come to play. But for the normal (yet realistic) imitation/sample orchestra, libs will become a bit larger (as hardware power increases) but not to the extreme. I think most orchestra libs will be (mainstream libs, not the biggies) arround 5 to 10 gigs by default.

[JonFairhurst]

Tascam has announced a new piano that has a single layer and uses advanced filtering only. They have also announced an older piano that has had its dynamic range expanded with advanced filtering.

I haven't heard these, but my concern is that a piano has three strings. How they interact with each strike is always different. If you use a single sample as the basis, you will always have the same interaction - lets say the first cancellation point is at 200 ms and the first total superimposition point is at 750 ms on a given note. Regardless of the filter, that note will always have those events at 200 ms and 750 ms.

That's the cool thing about lots of samples and rules like round robin - you can avoid hearing the same exact sample back to back.

I think advanced filtering on pianos has its place. If it's well applied you can get an infinite number of dynamic levels (okay 127 with standard MIDI), and if you combine this with multiple samples, round robins and other randomization techniques, you can add the needed variations. And you don't need a giant hard drive to host it.

My guess is that Tascam is following the GigaPiano model - they make a reference instrument available to developers who can then apply these tricks to their own libs. GigaPiano II wasn't all that great, but Tascam was able to prove the concept and show developers how to apply the techniques. And lots of 3rd party pianos have been all the better for it!

[joaz]

Interesting Topic. !!!

I think a determining factor on where we are heading with virtual instruments, will be the over-arching question of where we are heading with midi.

Here is some interesting info on HD midi.
http://www.cvnc.org/reviews/2005/features05/Zenph.html

If this research leads to solid results, that are embraced industry wide,( and I admit, that these are some pretty big if's ).................

Then I would expect that those companies that utilise ensemble-building as part of their approach
(Audio Impressions, Garritan ) would benefit noticably.

Although interestingly AI seem to be going down the ultra-Brute Force route, whereas Garritan products seem to aim at a minimal footprint, and ingenious programming.

The talk of extracting the artistic DNA from a Horowitz ( Ashkenazy, or a Joshua Bell), is bewildering, frightening, and also curiously attractive. :wow:

regards Joe