Vienna Symphonic Library Forum
Forum Statistics

181,965 users have contributed to 42,196 threads and 254,639 posts.

In the past 24 hours, we have 5 new thread(s), 9 new post(s) and 50 new user(s).

  • Who invented legato sampling?

    I am interested in knowing who first came up with the idea of dissecting transitional-notes-to-target-notes from start notes,  in order to create legato transitions. This is maybe the most important advance in realism for sampling, but who came up with this idea?    It is a huge stroke of genius!  Was it originated in VSL?   Now everybody does it, but who came up with the idea?  This person should have a statue.  (Or a patent - too late.)


  • The legato transition technology was not possible when the sample libraries was made for hardware sampler with limited RAM (8, 16, 32MB...). Remember PS Advanced orchestra? Then Gigastudio and computer with more RAM came and opened the possibilities.

    As far as I remember, the oldest sample library with legato transitions was the Garritan Strings http://www.soundonsound.com/sos/may02/articles/garritan.asp But I'm not 100% sure, it may be VSL. 


  • last edited
    last edited

    I don't ever remember seeing anything before VSL, and as it was only available for Giga and EXS, and needed a special Performance Tool, I'm sure I would have heard about it, had there been anything earlier.

    In fact most of the stuff we take for granted was invented by VSL, so thanks guys.  🎂

    DG


  • I had the Garritan Strings which did not have actual sampled legato, just a switch for closely connected sustains.   


  • "... just a switch for closely connected sustains."

    Mr. Kersten is correct. They took a segment of the string sound and slipped it in between the transitions like a buffer. GOS called them "masking samples." It was not a recording of a performance passing across an interval. 

    But I remember Garritan and his wares fondly. Nice man, great support, and considering the price, exceptional products. He was among the first to truly capitalize on hard disk streaming. 


  • I would like to know the actual answer to this question nevertheless.  Who came up with this idea?  Is it forever lost to history?


  • last edited
    last edited

    Hi,

    I´d say, the idea for a realistic legato that actually connected notes just like they were recorded was there for quite some time... but there had to be someone who could actually make it reality, which involves software development for third party samplers (remember the Performance Tool?) and VERY accurate recording and editing sessions, with up to 1000 samples for each legato version for each instrument (and you need to cover quite some legatos, check out the Solo Strings as an example).

    Sometimes it´s not just the idea, it´s also the brains, the team effort and the sheer manpower to make things work....

    Best,
    Paul


    Paul Kopf Product Manager VSL
  • I realize a lot of work went into it and remember using the performance tool.  I think my question was prompted by the fact that nowadays so many other companies are doing all that work to sample the transitions etc. - so I wondered where the specific idea of separated legato transitions attached to target notes first came from but I guess no one knows. 


  • Who the heck knows when the first legato transition was sampled. But VSL sure as well perfected it! ;)

     

    Seriously, VSL's legato (even just the SE basic patches) work more flexibly at slow and fast speeds than any other library out there. I'm a fan of Fanfare trumpets fast legato and some others I've been planning to get for who knows how long now. They are good. But even the entry level stuff outperforms most other libraries IMO.

     

    -Sean


  • That's true they perfected it.  Those fanfare trumpets have a fantastic legato.  I am also a fan of the 8 horn ensemble which has the most perfect sound for legato horns.  

    I think one of the best things about the VSL legato is how you can easily use those transitions from any note that is long enough.  That allows things like crescendo to legato, sforzando to legato, long detache to legato, etc.


  • last edited
    last edited

    I guess we'll never solve this mystery...🌩️

    It's a good question.  Before William started this thread I had always thought VSL invented legato samples.

    Not to change the subject but the mystery of who invented sampled legato is much like the mystery of who came up with the concept of MIDI.

    Now, before you say, "Oh that's a no-brainer. Dave Smith did back in 1983."  You should know that this common knowledge "fact" is, in fact, a myth.   Yes. Dave Smith of Sequential Circuts wrote the protocol or language that we know today as MIDI but the concept of MIDI, or linking electronic musical instruments together, had been available years before 1983.  The problem was, instruments were proprietary to the manufacturer who designed them.  So you could hook up a Roland keyboard with other Roland equipment and/or software but you couldn't hook up a Korg to a Roland keyboard and so on.

    What Dave Smith did in 1981 was write the language which became MIDI.  It was painstaking work that took hours of tedious coding that consumed the entire year.  Then came the hard part; getting all the major players in the industry to sign on.  That consumed all of 1982 and a good part of 83 right up until NAMM convention when it was announced.  So who came up with the idea of linking electronic instruments together with computers?🤔

    The ironic thing about MIDI is that most serious musicians scoffed when it first came out dismissing it as only good for computer geeks making cheesy blipidy bloop video game music a la Asteroids, and Pac Man.  Now, I'm wondering if any of those "serious" musicians are reading this right now while taking a break from their VSL MIDIstrations.😳  


  • last edited
    last edited

    @jasensmith said:

    Not to change the subject but the mystery of who invented sampled legato is much like the mystery of who came up with the concept of MIDI.

    It really is time for an update to the MIDI spec, but I think that we're just so far down the rabbit hole with current offerings that change would be really difficult. Yamaha invented a sort of update with their XP format, but it still used the old limited version. There are various new sorts, but nothing seems to be moving on that front.

    DG


  • last edited
    last edited

    It is an interesting topic about MIDI.  It is old, but still infinite in its capacity, unlike other protocols. I think most problems with MIDI come from sequencers, not MIDI itself.  To me, sequencers have done the same thing that all technology today - except of course VSL 😇 - is doing, which is to add absolutely everything that could possibly be done, though 99% of that is almost never done and the controls to do it are IN THE WAY. So technology becomes more and more clumsy.  Look at a Xerox machine - used to make simple copies 99% of the time  but now made so infernally overcomplicated they constantly crash or refuse to do a single-sided, black and white copy of one page.  I personally have a workaround for this phenomenon by using an antique sequencer that has none of the five million bells and whistles of the current sequencers but like MIDI is old and infinite. 


  • Bill, my main issue is the lack of velocity layers. I'm doing some work on a Disklavier a the moment, and I do notice the difference between MIDI and XP format. It may or may not make a real difference to sample work, but as time goes on and sample player software before every more complex, I think that there will be a few MIDI changes that will become necessary.

    DG


  • What exactly are you referring to with lack of velocity layers?  You mean differential control of layers within a channel? 


  • Anyway I'm sure you're right that changes could be useful.  I simply have noticed how wide ranging and useful MIDI has been compared to other 1980s digital technology because of its design which - at the time - must have seemed absurdly expandable to extremes, which nowadays are not so extreme at all.    An example being the velocity layers of the Vienna Imperial - previously inconceivable detail.   


  • The next midi protocol (HD MIDI) is coming. When? I can't tell. It's being discussed at the midi manufacturers assossiation.

    features proposed for High Definition MIDI:

    • Support for more MIDI Channels and Controllers
    • Greater resolution in data values
    • New messages that were not practical with the MIDI 1.0 protocol
    • The draft proposal allows HD senders to specify a direct pitch, rather than a Note Number. The Direct Pitch field sets the base pitch of the note, overriding the base pitch that would otherwise be selected by the note number, allowing easy implementation of alternate tuning systems.
    • In addition to Note On and Note Off, the draft includes a Note Update message that allows modification of parameters or controllers during the lifetime of a note.

    Note Update feature looks particularly interessting for samples/orchestral music. Maybe like a standardized VST note expression feature.

    Hope Sample library and sample player manufacturer will jump in when the protocol will be completed.


  • last edited
    last edited

    @William said:

    What exactly are you referring to with lack of velocity layers?  You mean differential control of layers within a channel? 

    Sorry, I was typing quickly. What I meant was that with only 128 possible values, it is not enough. The XP format gives 1023, I think. The reason it's not enough, is that the very lowest values are almost never used, and the highest ones aren't either, so in fact the actual normal usable range is really 50-80 steps. Having tested it, even though it sounds fine without, there is something more accurate with the higher resolution.

    DG


  • last edited
    last edited

    @William said:

    What exactly are you referring to with lack of velocity layers?  You mean differential control of layers within a channel? 

    Sorry, I was typing quickly. What I meant was that with only 128 possible values, it is not enough. The XP format gives 1023, I think. The reason it's not enough, is that the very lowest values are almost never used, and the highest ones aren't either, so in fact the actual normal usable range is really 50-80 steps. Having tested it, even though it sounds fine without, there is something more accurate with the higher resolution.

    DG

    I agree. 

    128 values seems grosely inadequate in this day and age.  With more values you can achieve more happy middle grounds for parameters such as expression and attack and action on those parameters would be more forgiving.  With 128, you are constantly battling, what I call, the Goldilocks effect; it's either too much or too little and you're constantly having to go back and re-edit or re-record.


  • "The reason it's not enough, is that the very lowest values are almost never used, and the highest ones aren't either, so in fact the actual normal usable range is really 50-80 steps." - DG

    That makes sense and more smoothness would probably be noticeable depending on the music.