However, what is the future of sampling?  What is the ultimate goal?  What happens if you have a perfect sample library?  What will that accomplish?   What is all this leading to?

These are very good questions, William.

Personally, I feel the ultimate orchestral library is to enable people who cannot afford to hire a real orchestra , to have their compositions that would otherwise never be heard "rendered as though by an orchestra" .

God forbid that it would replace an orchestra, or reduce the viability of real orchestras. For me, it is a miracle that 80 odd people with nothing more than wood & reed & brass & gut can make these awe inspiring sounds. And I feel it would dimish their talents and genius, if midi orchestrations just become something that can be knocked off in a couple of hours by guy with a computer and an expensive library.

In some ways it is heartening that it took Jay so long to do this. It means talent still matters; the library should empower the talent - it should not dimish need for talent.

I hope that good midi orchestrations will help give birth to a new classical repotoire and make real orchestras MORE viable, and re-invirgorate the genre.

Well, sample libraries are never going to put symphony orchestras out of business, as they are always almost out of business.  They were before samples, and they are now.  They are not affected by this use at all, and exist precariously in their own world.  In fact symphony players probably get more jobs by playing samples.  

However, sample libraries and synthesizers have definitely put stage orchestra players and studio musicians out of business. I am one of them.  But obviously they are themselves a form of performance - a very complicated issue.

Your point about the 80 odd people (some of them are VERY odd, I can attest to that)  is sometimes true, but only about a great symphony orchestra.  Despite what everyone thinks,  great orchestras are not common just because there are CDs of them all over the place.  Amateur orchestras are the usual, and it is not a miracle to hear them play - it is often painful and they tend to slaughter your music.  So I agree partly with what you're saying, but not all of it because just getting a live orchestra to play your music is not the holy grail.  It has to be a great live orchestra to be better than samples. 

 "In some ways it is heartening that it took Jay so long to do this. It means talent still matters; the library should empower the talent - it should not dimish need for talent." - drew buchan

Your post had a lot of content, so I also noticed this and think it is a good point.  Everyone looks at sample libraries as a substitute, but there is no substitute for actual talent, which the accuracy and depth of VSL definitely does empower.   if this sound is perfected to the ultimate degree, then you will have every choice that a conductor of the London Symphony or the New York Philharmonic has.  But what will you do with it? An idiotic jingle?  A trite and imitative film score?  Or a great symphony or innovative new work of art?  Unfortunately most people will not do the latter or even want to.

 Contrary to what many people think, I actually DO beleive that at some point samples (or more likely some other form of technology) will be able to sound like the real deal.

The past 10 years have shown quite some improvement, but we still have a loooong way to go. Not only in the quality of the sound, but as importantly in terms of space, room and blending. Right now it is virtually impossible to capture the sound and sparkle of a well recorded orchestra. But I'm convinced we will get there.

Orchestras will still have their place, but more for the incredible joy and experience of watching a symphonic performance.

I agree with that. Maybe MIR will be the first time this is going to be accomplished with samples.   It is the physical sensation that an instrument is actually located in a specific place, almost being able to see it with your eyes closed, that gives that added sense of reality. 

The live orchestra will never be obsolete of course, any more than movies made live theater obsolete.   The whole point of a play is that it is happening LIVE.  And the same applies to an orchestral performance.

However, with recordings, it is a different story. 

"However, what is the future of sampling?  What is the ultimate goal?  What happens if you have a perfect sample library?  What will that accomplish?   What is all this leading to?"

My 2 cents:

I think the samples or physical models of instruments is only part of the story. The other part is the reverb, like MIR. I think we are getting to the point where technology is really close to making these 2 aspects a non issue. But, I feel there are 2 more areas, that if conquered will allow us all to express ourselves creatively on even a higher level.

With the current state of sampling if you are great composer but a poor programmer your music suffers. Not everyone has the time to sit and try every sample and articulation until they find the right one for a phrase and do that for every phrase and every instrument, then mix and eq the instruments and finally know how to use and tweak reverb. 

So, what I think needs to happen along side these advancements in sampling is a notation program (preferably by VSL :-) that works like the program Notion or the Sound Set feature in Sibelius. You click in the notes onto the page and write pizz, tr, Arco, trem, or draw in any markings and the correct samples load and play just for that note. Where you don't have to load gigabytes of samples for 4 notes of pizz and then have to "optimize". Only load what you need and when you need it.

The other thing that needs to happen is "virtual paper", for lack of a better term. Basically, a tablet display with a stylus (like the Wacom stuff) and the above described notation software's ability to read hand writing and convert it in realtime to a score. Now you can draw in slurs, notes, markings...by hand and it gets analyzed immediately and converted to the proper digital version right in the score in realtime and the software loads the correct samples in the background based on these markings.

Until this happens the quality of samples almost does not matter because its the skill of the programmer that matters more than the samples. We've all heard bad mockups with great software and great mockups with inferior software so the problem is not only in the samples its in the whole process.

Once we can go back to writing music the traditional way but in a digital fashion via "virtual paper" + specialized notation software and wonderful VSL samples underneath it all the quality of the music being produced by all of us will increase because all these extra steps between the creative process and the realization of that creativity in physical form as audio coming out of speakers will disappear. 

So, to me samples and reverb are only half the battle. A specialized input device a la virtual paper, and specialized notation software that can read handwriting and immediately convert it plus load the proper samples in the background is what I'd like to see.

Think about this. You open the specialized notation software and you have this tablet + stylus and you start writing in the name of the piece, tempo, "composed by", date, "allegro", "largo", time signature, and as you write the software recognizes the terms and recreates them as if you typed via keyboard and mouse. Then you start writing in the notes and as you write they appear on the staff, in the background only those samples gets loaded, not the whole instrument and while its loading you can keep writing. Now, load times go away.

If someone wants to sit down and play everything in via midi controller thats fine, afterwards they can add all the markings by hand via tablet/stylus.

As far as mixing and reverb...well there are already some set of standards for these. You can define all this in the specialized notation program every time you start a new piece or use a template. We are already doing that now.

I believe that the above is the final frontier and until then great samples can still sound fake and bad because its all about the programming, and the programming is so tedious and requires so much work that it kills creativity unless you've defined a workflow process that you've refined over many years.

The best thing VSL can do is to make the process easier and more simplified without taking away the quality they have achieved. A specialized notation (it would have to be a full-on DAW really because we would need 3rd party plugins and audio/video features) software would a move in that direction.

Its no small task but until this happens the whole process is very limiting and very few people have the time it takes to make their compositions sounds as good as they want.

Am I crazy or do you agree?

DM33

 I agree with that - that is a great post.  

Carrying what you said into fantasy (at least for now), the ideal would be a sheet of electronic paper with staff lines that were midi sensitive.    You would set it in a dock and the notes would load according to the orchestral setup on the paper you selected.  Of course crescendo, tempos, dynamics, etc. would also be recognized. It would then play back with intelligent algorithms you selected as appropriate for the particular piece.

That is an extreme example, but shows how right now you are forced to conform to the equipment, which is huge and clunky compared to a pen, paper and brain.  Also, I have always liked the concept that music itself should be the programming language.  It is actually that, though it uses "equipment" that has existed for centuries.

"electronic paper with staff lines that were midi sensitive"

MIDI paper!!!! That's genius! You get the idea William!

I would love to incorporate the technology and at the same time go back to brain --> pen --> paper composing model.

I enjoy the immediacy that software and technology offer (when I'm improvising and creating themes, progressions, melodies) but I still compose the traditional way and then program it. It takes me a weekend to compose a piece and a week to program it...well I could have composed 3 more pieces in that time. Of course I'm very impressed about the sound and quality of the finished work, whether my music is good or not I still know its going to sound great. But what a serious amount of work it is!

Even if we took one step in this direction:

- a way to input music via handwriting recognition software using a tablet + stylus.

or 

- a perfect marriage between samples and a notation program

It would increase the quality of music all of us compose. For instance, right now if I have a great musical idea but I know its going to take to much work to realize it with samples I don't even bother with it and set it aside. I have many pages of music that I don't have time to program. I always have to simplify the music and curb my creativity. Then we all sit around and complain about how there is not enough good music...its a catch-22.

VSL, have you considered designing a notation/DAW program like "Notion" (where samples are automatically linked to the score) but with all the bells and whistles a composer needs today...3rd party plugin support, video scoring, audio plugins and mixing? That would be a good start, you can later add the handwriting recognition features later :-)

You are somewhat already there with VE3...just add everything else, its easy :-) If anyone can do it VSL can.

Thanks for listening,

DM

I guess in a way "midi paper" already exists in the form of Sibelius and photoscore.  I use Sibelius to compose, or scan music in - it will even cope with handwritten scores to an extent, which I then export to my DAW and hone to perfection (sic).  OK so it is an imperfect system - but it does work pretty well. 

The performance marks written in Sibelius are rendered as midi messages, the parameters of which you can edit, and I guess that the flexiblilty and power of this feature will evolve.  With the advent of touch screen displays I suppose Sibelius could eventually accept input via a pen on a "paper" screen.  That would be cool.

 Yes, you can do that in Finale as well but again, you are forced to conform to how the equipment makes you work - at a computer terminal, scanning, etc. etc. 

But the physical aspect of this is more what I meant - the concept of using book-like forms that are actually digital is being done already as well as the development of flexible, paper-like LCD (or similar) screens. 

Another device I have wanted is a handheld USB recorder that transcribes a whistled or hummed tune into notation and then uploads into a computer.  This would replace the scrawled notes for melodies or motifs that I am often putting on napkins or scraps of paper and losing.  Of course then I would lose the recorder with a year's worth of humming and really be disgusted.

As I said in my post, I agree with you that Sibelius etc are imperfect, but with the advent of touchscreen LCDs and digital paper - which i am sure will arrive fully in the next five years (windoze 7 is rumoured to have some touchscreen capabilities) - all our DAWs and software will have a more tactile interface. I am sure the technologies will converge and develop.

As regards your second point about a hummed or wistled tune - once again it exists in nascant form in Melodyne - the latest version will have the ability to turn more than one line into midi....

Now..... what about a thought wave reader, to turn what we are thinking into midi....

No wait!  We don't want it to be tooooo easy - everyone will be doing it then

Those are more like peripherals to computer systems. 

I meant a completely different base of operations that is exactly like traditional musical creation as much as possible.  Not taking a computer and making it a bit similar, like a touch sensitive screen, or putting a program that can transcribe sounds,  but instead completely remodeling the basic GUI so that it is radically different as in the case of music paper that was completely internally digital and had no cables, etc.  This is seen in something like the Amazon "Kindle", which is not a glowing backlit computer screen (which irritates long-term book readers) but is actually illuminated by the ambient light in the room, exactly like book paper.

 Sorry this got so far away from the topic.

I was wondering if Jay could discuss the percussion parts on this, and what difficulties there might have been, how many tracks were used, which instruments.  I had the impression that the tam-tam part would be difficult with muted fortissimo, etc.  Also, if all the instruments asked for in the score were available.

Everything we have talked about here will happen but in do course.

I have recently seen a robot made by Toyota connected wirelessly to a human. When the man had a thought about moving his foot the robot moved his foot. This was repeated but with various tasks like walking, jumping...the man was in a chair but the robot was doing what the man was thinking. It was amazing.

I know that everything I wrote about above will become available, its just a matter of time.  For me they can't happen soon enough and the discussion about samples is only half the battle. Making the process of music dictation easier is what needs to happen. Right now the samples are better sounding than this process allows.

I see your point about how it will become easy and everyone will be doing it then, but music composition is unlike the other arts like dance, or painting, sculptor in the way that we (the audience) only care about the final product and not process whereas how a painting was painted or statue sculpted is part of those artistic disciplines. Its not really the composing we are making easier but the process of getting it from our brains to the world. We still need the musical training and understand orchestration, counterpoint, melody....

What I care about is how good my music is and how fast I can compose the next piece and I want to take away all the steps that are not required in the process, the natural organic process: think, compose, perform/record. And that process is simply dictation, once the musical idea hatches. Many of the steps we have to take right now to get the idea into the computer are completely useless to the creative process and in fact blocks creativity itself.

I look forward to the future where I can put on a helmut (instead of using a keyboard and mouse) and conceive music as I hear it in my head and it gets scored right into some software for me to hear it back immediately.

Don't get me wrong, I LOVE writing music on paper at the piano. If I had all the money in the world that is how I would compose. I would MUCH rather compose that way as opposed to wearing a helmut or using a mouse or keyboard, I would also much rather hand the finished piece of music to real musicians and rehearse them and conduct the piece myself and ditch the computer completely. But, since I am not wealthy I have to resort to this unnatural process with computers just to hear my ideas as they would sound with a real orchestra.

I am sorry to have hi-jacked this thread so maybe we can move these last few messages to a new thread and continue there if need be.

DM33

Jay, this is an astonishing piece of work with meticulous attention to detail. Well done. Thank you for sharing the MIDI file and VE set-ups with us, they are very useful resources.

"The second challenge for me was acquiring technology powerful enough to get the job done. The size and complexity of the Rite quickly brought my old computer (Windows 32-bit XP with 4 GB of RAM) to its digital knees. I would have to brave the new world of 64 bits. I got a book on building my own computer and bought all the separate components. Then, fearing that I would somehow fry my expensive circuits with a single ungrounded touch, I carefully assembled my 16 GB, quad-core silicon juggernaut. Surprisingly, the building process was much easier than I feared. And within a few days I was able to run more than 100 VSL MIDI tracks on a single computer with only the occasional hiccup."

if you're not too busy sequencing the entire works of Wagner, could you kindly spare the time to elaborate and tell us what the components, OS (etc.) of your new computer are? (Sorry if this is posted elsewhere and I missed it.)

Here are the main specs for the system that I used to for the Rite of Spring.  Looks like I'm going to have to build an even faster one to run MIR.   I'll post those specs too in about a month. 

--Jay

OS: Vista Home Premium 64 bit
Processor:  Intel Core2 Quad Q6600 @2.40GHz
Memory: 16 GB
Graphics: NVidia GeForce 8800 GT
Sound: Echo Digital Audio Mia Midi
Network Adapter: Marvell Yukon 88E8001/8003/8010 PCI Gigabit Ethernet Controller (Ethernet 802.3)
Main Drive: Seagate Cheetah 15K 300GB

2 500GB eSATA hard drives

 Thanks Jay!

Hi again Jay, I downloaded the two big MIDI files and the VE set-ups for your fabulous arrangement and am having great fun studying it. One question - in Part 1, I found what appears

 to be a glitch in the time signature and tempi tracks. Having started at bar 11, they both seem to be missing data for 48 bars from bar 12. This may well be a side-effect of Logic 7's inept attempt to import the file, but if someone has the time could they kindly check the original MIDI file (available here:

http://vsl.co.at/downloader.asp?file=/Sounds/MP3/VI_DEMOS/Rite_of_Spring/Rite-Part1_ProjectFiles.zip )

... and let me know if it seems OK? Many thanks. Such a great piece of work, it's wonderful to be able to hear all the individual parts! [Y]

 Just rechecked midi file.   Looks fine in Sonar 8.

--Jay

 Thanks Jay, must be a Logic thing.