Composers' Forum

Music Composers Unite!

What will MIDI 2 bring to orchestral sounds? Discussion....

I guess many people know that the MIDI 2 standard is out now. Here is a brief overview of the MIDI 2 main specs if your not yet up to speed:

  • It is bidirectional - meaning that the DAW and the MIDI device can talk both ways.
  • It's got 32bit resolution meaning we are no longer confined to 128 degrees of velocity - millions of shades instead.
  • It's backwards compatible with MI(DI 1 and will enhance  MID 1 Instruments (I am not sure how).
  • Its got 'profiles'. Common Classic instruments such as piano, drawbar organ, drums etc, will have a standard mapping, hopefully meaning one can simpllify set up. I understand Violin will come with a list of articulations.
  • Then there is 'property exchange', where states of the instrument can be sent from DAW to instrument and back. Automapping controllers should get easier, even instantaneous.

The MIDI 2.0 Protocol uses the architecture of MIDI 1.0 Protocol to maintain backward compatibility and easy translation while offering expanded features.

  • Extends the data resolution for all Channel Voice Messages.
  • Makes some messages easier to use by aggregating combination messages into one atomic message.
  • Adds new properties for several Channel Voice Messages.
  • Adds several new Channel Voice Messages to provide increased Per-Note control and musical expression.
  • Adds New data messages include System Exclusive 8 and Mixed Data Set. The System Exclusive 8 message is very similar to MIDI 1.0 System Exclusive but with 8-bit data format. The Mixed Data Set Message is used to transfer large data sets, including non-MIDI data.
  • Keeps all System messages the same as in MIDI 1.0.

Building off MIDI CI, Property Exchange lets one device ask another device what parameters it’s got that are available to access. Parameter lists, controller mappings, synthesis parameters, and information about presets are a few of the details that can be communicated - and communicated using meaningful names rather abstract, generic numbers.

I am not sure how much of this can be accessed on a standard MIDI 1 instrument simply by writing a new driver. There seems to be a device called a " Bomes MIDI Translator" which translates MIDI 1 to MIDi" messages and back. I am not sure how it fits in but it may make it possible to use MIDI 1 devices more effectively with MIDI 2, whwn it starts to take a grip of the VST market.

For me, in orchestral terms I have always been frustrated by Orchestral VSTs. I find them all lame. I play many acoustic instruments and keyboards. I play brass and woodwinds. When I play sax for example, I get an ongoing humanised vibrato in every extended note, this is part of the sound, but its very difficult to achieve using MIDI controllers. Using hte Yamama and ROland controllers did not do that much for me, I went back to using keyboard, but that is not good too. Same with Violin, the vibrato on a note is mission critical and should be crafted ituitively by the user, rather than having the same vibrato for each note.

Then there is the issue of articulations. Violin in particular has over 50 different types of articulations, sometimes these articulations blend into eachother and sometimes one articulation (say Staccato) is very very different in different settings. The length of staccato notes vary widely and this is insitctively addressed by the player without it being notated.MIDI 2 has  new attribute for note on/off2 , Attribute Type and Attribute data field. I think this may possibly enable choice of attack.

Let's consider the sound of a trumpet. It is vastly different when heard playing Purcell's trumpet voluntary then when Chet Baker or Maynard Ferguson plays it. Same with Sax, Stan Getz's tone is a world away from Parker or Webster.

There seems to be Four issues when trying to emulate sounds on a computer system.

  • First the mode of entry, the "instrument interface", (keyboard/drumpad/other) must be capable of controlling the sound.
  • Secondly the sound must be available within the VST.
  • Thirdly the whole process should be easy to set up, or even no set up at all.
  • It MUST sound right

I wonder what people's opinions are about teh impact of MIDI 2 on these issues. Where will it take us on this crucial matter? What will the new generation of orchestral VSTs be like?

Views: 49

Reply to This

Replies to This Discussion

Thanks for sharing this important information, ZeroZero!

I read about this when it was announced at NAMM. I don't claim to know everything about it yet. As with any new technology there will be nay sayers and supporters. As time moves along the nay sayers will drop like flies one by one.

The main consideration for me will be how useful it actually is in real use. Since it is a standard that doesn't require total adoption by manufacturers we might see partial adoption such as midi controllers and DAWS that automatically configure to one another. This will be a good thing if all controls are already mapped. I don't see this happening right away though. Might take some time to slowly become a reality. 

Much of the 127 levels limitation in MIDI 1 could be gotten around  using good scripting and programming all depending on the application.

Concerning sample libraries, I think there will be a competitive race by the larger sample library makers to be the first to adopt it and therefore have something over the competition. I can see Spitfire, VSL, Native instruments and others all jumping on that band wagon, even if the real benefits to some compositions are negligible.

You seem to be mostly concerned with issues that are really sample playback. I'm not sure that any improvement in MIDI is going to help with that.

Victor the midi controls the samples.Midi has everything to do with that.

Not *everything*. Midi starts the sample. It has no say over what happens inside the sample.

Timothy Smith said:

Victor the midi controls the samples.Midi has everything to do with that.

I guess we could split hairs over this. If you didn't have midi nothing else would work.

I will be interested to see how the multiple levels of increased control will influence the actual samples >sound. Too early to tell just yet. When we play a piano key we aren't deciding at that moment how much pressure we are applying to the keys. Prior to this  MIDI1 decided in 127 different levels based on key speed and pressure.

127 seems as if it should have been plenty and it usually was. I doubt all vsti makers used all 127 points. MIDI 2 should have exponentially more levels than this. It might allow the production of more expressive midi controllers that could play a sax sound for instance with such new realism that you can't tell the difference between it and a real sax.

MIDI has always been the most advantageous to keyboardists and pianists in my opinion because it related so well as a midi input device. Yes we have midi guitars among many other things. To me they never worked quite as well as the keyboard. This is good for composers because many compose using a piano.

I've downloaded the specs and am gradually working through. I'm mostly concerned with the midi editor in my daw and how that will react to the changes. Much of it seems to be about expanding control and bi-directional messaging (all wrapped up in wadges of jargon). I haven't yet got to the structure of a midi2 message and file layout but since word is it's backward compatible with Midi1 then the transition should be easy.

Thank you for the information.

Reply to Discussion

RSS

Sign up info

Read before you sign up to find out what the requirements are!

Store

© 2021   Created by Gav Brown.   Powered by

Badges  |  Report an Issue  |  Terms of Service