MIDI Association Feb 23rd news on MIDI 2.0

[synkrotron]

6 minutes ago, JohnG said:

That was the original problem, it was too small to be read when I did a direct upload. Let me try again.

I'm curious, what has Garittan ARIA Player got to do with MIDI 2.0, or even plain old MIDI?

 

[JohnG]

13 minutes ago, synkrotron said:

I'm curious, what has Garittan ARIA Player got to do with MIDI 2.0, or even plain old MIDI?

 

Well, the standalone ARIA player can be coupled up to a MIDI keyboard in the usual way and played.

One can load a MIDI file into it (16 track only) and it will do a high speed rendering of it or just play it back with the chosen sample set.

The VSTi version can be loaded into CbB in multiple copies for a big orchestra.

BUT, most importantly, each VST loaded can be set individually to a pre-supplied set of different A tuning and any one of thousands of Scala (.scl) files can be loaded to provide an appropriate temperament. The example has A=415Hz and the Werckmeister III temperament added. See the screen dump towards the lower RHS.

I do quite a few orchestral backing tracks for singers of Bach cantatas.

Any good?

JohnG.

Edited February 28, 2019 by JohnG

[synkrotron]

2 minutes ago, JohnG said:

Any good?

Not really, John.

I see all this MIDI 2.0 stuff more important for peeps who use MIDI hardware stuff, not software instruments.

 

[JohnG]

1 minute ago, synkrotron said:

Not really, John.

I see all this MIDI 2.0 stuff more important for peeps who use MIDI hardware stuff, not software instruments.

 

Yep ... if anybody implements it!

Question. Why so few keyboards with polyphonic aftertouch?

Answer. (I believe) insufficient people were prepared to pay the premium for it.

I may be a grumpy old fool, correction, I Am a grumpy old fool, but I don't see much market for MIDI 2.0.

I hope I'm proved wrong.

But I can get all the tuning variations I need from existing MIDI 1.0 compatible products.

Sufficient dynamic range, etc., etc.

[InstrEd]

4 minutes ago, synkrotron said:

I see all this MIDI 2.0 stuff more important for peeps who use MIDI hardware stuff, not software instruments

Me too.

I'm thinking more what will ROLI be able to do with it.  Seaboard

[synkrotron]

1 minute ago, JohnG said:

but I don't see much market for MIDI 2.0

That depends, of course... For me personally, I couldn't give a monkeys if MIDI 2.0 never gets off the ground. I have three pieces of MIDI hardware synths and two hardware effects units that accept MIDI for sync'ing tempo and for patch changes. I ain't going to be investing in new hardware anytime soon, I've got enough, and if I do then it will be CV/Gate stuff, not MIDI.

I also don't do micro tonal stuff. Hell, I don't even use the black notes except by accident.

But there may be many that are looking forward to it, and I am happy for them.

cheers

andy

[synkrotron]

This is a quote from the PreSonus site, regarding "MIDI and beyond MIDI"

Quote

MIDI—and beyond MIDI.

Studio One works with all the standard MIDI gear you know and love, from controllers to tone modules. But once MIDI data arrives into Studio One, it’s converted to a high-resolution, 32-bit internal format. That means no zipper noise on instruments, smoother controller changes and pitch bends, more detailed automation, and other benefits when working in the Studio One environment. And if you need to drive external MIDI gear, you’re covered there too—Studio One translates its high-resolution format back into standard MIDI data if it needs to return to the outside world.

 

[TheSteven]

1 hour ago, synkrotron said:

Not really, John.

I see all this MIDI 2.0 stuff more important for peeps who use MIDI hardware stuff, not software instruments.

 

Being that software imitates hardware I can see possible benefits in how MIDI 2.0 could lead to changes in how software instruments/effects, hardware (not just controllers) and the DAW interact.

For example:

• An increase in MIDI resolution directly translates in a higher resolution for Automation so you will have finer control.
  This is not a possibility but a definite if a DAW supports MIDI 2.0.

  Quote

  MIDI 1.0 messages are usually 7 bit (14 bit is possible by not widely implemented because there are only 128 CC messages). In MIDI 2.0 velocity is 16 bit and the 128 control change messages, 16,384 Registered Controllers, 16,384 Assignable Controllers, Poly and channel pressure and Pitch Bend are 32 bit

• Automapping of VST outputs.
  For example Kontakt or a drum VST recognizing that it was connected to a stereo buss or multiple tracks and automatically routing it's outputs accordingly eliminating have to have mono, stereo, x8, x16 output versions of the DLL.
• Automatic controller mapping  i.e. your VSTs automapping (like what Omnisphere is kind of doing with hardware mapping) to whatever hardware controller you're using so while MIDI learn would still be there that 90% of time you would never need to use it.

I see a lot of stuff that is currently being handled on an ad hoc basis, being standardized and being able to be globally controlled.
For example your DAW being able to switch ALL your instruments instantly to the same microtuning scale.

So who's for taking bets that Steinberg will be announcing a spec for VST4 at the same time MIDI 2.0 is officially adopted?

Edited February 28, 2019 by TheSteven

[Øyvind Skald]

3 hours ago, TheSteven said:

Being that software imitates hardware I can see possible benefits in how MIDI 2.0 could lead to changes in how software instruments/effects, hardware (not just controllers) and the DAW interact.

For example: 

• An increase in MIDI resolution directly translates in a higher resolution for Automation so you will have finer control.
  This is not a possibility but a definite if a DAW supports MIDI 2.0.
• Automapping of VST outputs.
  For example Kontakt or a drum VST recognizing that it was connected to a stereo buss or multiple tracks and automatically routing it's outputs accordingly eliminating have to have mono, stereo, x8, x16 output versions of the DLL.
• Automatic controller mapping  i.e. your VSTs automapping (like what Omnisphere is kind of doing with hardware mapping) to whatever hardware controller you're using so while MIDI learn would still be there that 90% of time you would never need to use it.

I see a lot of stuff that is currently being handled on an ad hoc basis, being standardized and being able to be globally controlled.
For example your DAW being able to switch ALL your instruments instantly to the same microtuning scale.

So who's for taking bets that Steinberg will be announcing a spec for VST4 at the same time MIDI 2.0 is officially adopted?

I really hope so!

[Øyvind Skald]

4 hours ago, JohnG said:

Well, the standalone ARIA player can be coupled up to a MIDI keyboard in the usual way and played.

One can load a MIDI file into it (16 track only) and it will do a high speed rendering of it or just play it back with the chosen sample set.

The VSTi version can be loaded into CbB in multiple copies for a big orchestra.

BUT, most importantly, each VST loaded can be set individually to a pre-supplied set of different A tuning and any one of thousands of Scala (.scl) files can be loaded to provide an appropriate temperament. The example has A=415Hz and the Werckmeister III temperament added. See the screen dump towards the lower RHS.

I do quite a few orchestral backing tracks for singers of Bach cantatas.

Any good?

JohnG.

It seems cool.  But can it make Kontakt or even Komplete Kontrol change micro tonal and temper?

[JohnG]

18 hours ago, ØSkald said:

It seems cool.  But can it make Kontakt or even Komplete Kontrol change micro tonal and temper?

No, the Garritan libraries used to be Kontakt based some years ago but, in my understanding of some of the shortcomings of Kontakt, they decided to move to get their audio engine and sample player designed and implemented by Plogue. That's the ARIA engine.

Their first commercial ARIA offering was, I believe, GPO4. This meant they were very much more in control of the GUI (controlled by XML) and also of other elements, such as tuning and temperament. It also meant that the sample player could be based on the open standard SFZ. It meant also that temperament could be user defined by simple creating a new Scala (text based) file. There are already hundreds of such files, some even with Wendy Carlos' tunings.

For me that means I can rewrite the GUI should I need to,  and I do sometimes want to add another controller, e.g. an ADSR envelope for a brass instrument, or modify the loop point in a sample.

And it all works fine without a trace of Kontakt.

Edited March 1, 2019 by JohnG

[TheSteven]

Lol took me a while to find this original thread.

Below is part of an email that went out to MIDI Association members on Mon Jan 27th.
**********************************
MIDI 2.0 Adopted at Winter NAMM 2020!

 

Hi ,

At the Annual Meeting of the MIDI Manufacturers Association during Winter NAMM 2020,  MMA members in attendance unanimously completed the adoption of the core MIDI 2.0 specifications including five core documents.   

• MIDI Capability Inquiry (Update)
• Specification for Universal MIDI Packet (UMP) Format and MIDI 2.0 Protocol
• Common Rules for MIDI CI Profiles
• Common Rules for MIDI-CI Property Exchange
• Property Exchange Foundational Resources and Basic Resources

At the annual meeting afternoon session, there were demonstrations by Korg, Roland and Yamaha of prototype MIDI 2.0 devices sending and receiving MIDI 2.0 protocol messages.

Details about MIDI 2.0™, MIDI-CI, Profiles and Property Exchange

[craigb]

I wonder if we'll get MIDI 3.0 in 2055...

[User 905133]

1 hour ago, TheSteven said:

Lol took me a while to find this original thread.
Below is part of an email that went out to MIDI Association members on Mon Jan 27th.

Thanks for taking the time to find the thread; a Coffee House thread is more relevant to my comments than the Deals thread on MIDI 2.0. So, thanks for calling this to my attention.  BTW, the link you provided here is the same one that Google turned up for me this morning--the source of my "red flag."  ?

A footnote: Several years ago (4 - 5 +/-? ) I was doing midi tests with an arduino and discovered that the arduino could pump out midi data at 115200 baud to my PC and softsynth via the hairless serial midi tool going into the RX of an audio card joystick port midi interface cable.  It was a custom Arduino C test program to generate notes and CCs on 16 midi channels based on some scales (acceptable note arrays, with randomized CC values and randomized octaves within a 2 or 3 octave range).

I am not sure what did the serial rate step down (to 31250 baud), but MIDI 1.0 hardware (sound modules) had no problems. I didn't investigate where the step down took place (Windows XPSP3 PC), but it was clear that going in, it was 115200 and that the data was being passed around internally at a faster speed than my midi interface was sending it out to the hardware. (Not sure if it was my parallel port 8x8 Motu or another interface.) 

An opinion: Considering that an arduino is super slow compared to later processors, I hope that MIDI 2.0 gear manufacturers design gear, drivers, and OSes with the potential for digital communication speeds that will be around in 30 years.

 

Edited February 1, 2020 by User 905133

[JohnG]

16 hours ago, craigb said:

I wonder if we'll get MIDI 3.0 in 2055...

Don't care, I certainly won't be around any more!  ?  ?

[JohnG]

16 hours ago, User 905133 said:

Thanks for taking the time to find the thread; a Coffee House thread is more relevant to my comments than the Deals thread on MIDI 2.0. So, thanks for calling this to my attention.  BTW, the link you provided here is the same one that Google turned up for me this morning--the source of my "red flag."  ?

A footnote: Several years ago (4 - 5 +/-? ) I was doing midi tests with an arduino and discovered that the arduino could pump out midi data at 115200 baud to my PC and softsynth via the hairless serial midi tool going into the RX of an audio card joystick port midi interface cable.  It was a custom Arduino C test program to generate notes and CCs on 16 midi channels based on some scales (acceptable note arrays, with randomized CC values and randomized octaves within a 2 or 3 octave range).

I am not sure what did the serial rate step down (to 31250 baud), but MIDI 1.0 hardware (sound modules) had no problems. I didn't investigate where the step down took place (Windows XPSP3 PC), but it was clear that going in, it was 115200 and that the data was being passed around internally at a faster speed than my midi interface was sending it out to the hardware. (Not sure if it was my parallel port 8x8 Motu or another interface.) 

An opinion: Considering that an arduino is super slow compared to later processors, I hope that MIDI 2.0 gear manufacturers design gear, drivers, and OSes with the potential for digital communication speeds that will be around in 30 years.

Data communications, i.e. outside the processor environment, was/is driven by devices called UARTs (Universal Asynchronous Receiver/Transmitters).

Back in the day, these devices were "clocked" by a crystal oscillating at 1MHz. The 31,250 b.p.s. is derived as a subdivision of this clock (keep dividing by 2). Your Motu has UARTs inside it, probaby one per socket, although today they may be collected into one more powerful chip.

It's not so much what's inside the computer as what connects the computer to the external networking gear, or one MIDI device to another. One of the reasons that the original MIDI 1 was so widely adopted was because the addition of a UART or 2 and its supporting electronic components and DIN sockets was dirt cheap. Ergo, why not?

Currently we have data transmission technologies like USB, firewire, Ethernet etc. It will be these and the adoption of these technologies within our devices that will provide the higher speeds. Not, sadly, DIN plugs and cables. Some of these technologies provide severe limitations on the distance one object can be from another.

 

Edited February 2, 2020 by JohnG

[User 905133]

1 hour ago, Finnbogi Ragnar Ragnarsson said:

This is the essence of the two way communication.

From Midi.org

"1.1 MIDI Capability Inquiry (MIDI-CI)

The additional capabilities that MIDI 2.0 brings to devices are enabled by MIDI-CI. The basic idea is that if devices have a bidirectional connection, they can exchange their capabilities with each other. Devices can share their configuration and what MIDI functions are supported.

Devices use a MIDI-CI discovers and configures device features using 3 categories of inquiry: Profile Configuration, Property Exchange, and Protocol Negotiation. to configure MIDI features when both devices agree to support that feature. MIDI-CI discovers and configures device features using 3 categories of inquiry: Profile Configuration, Property Exchange, and Protocol Negotiation.

If a device does not support any new features, it uses the MIDI 1.0 as usual. Devices connected to that device will continue to use MIDI 1.0 in communication with that device. "

Thanks for contributing this to the discussion.  I have put this quote here because this seems more relevant to a generalized discussion of  MIDI 2.0 as opposed to the "Deals" subtopic. 

For me, it really depends on how gear builders, software makers, and others implement MIDI 2.0 and choose to make things (or not make them) fully compatible with existing stuff, esp. for me, all my existing gear.

Quote

MIDI-CI discovers and configures device features using 3 categories of inquiry: Profile Configuration, Property Exchange, and Protocol Negotiation.

I haven't read the details, but this sounds to me like handshaking stuff. If its done via usb, it involves serial communication. Time-stamp handling routines should be fascinating to see how they handle information--for example information sent later but received before other information.

I am curious to know if there are parallel interfaces out there or are on the drawing boards or if any hobbyists are working on parallel processing of MIDI 2.0 data.

 

 

Edited February 3, 2020 by User 905133
to add a comma

[User 905133]

12 hours ago, JohnG said:

Data communications, i.e. outside the processor environment, was/is driven by devices called UARTs (Universal Asynchronous Receiver/Transmitters).

Back in the day, these devices were "clocked" by a crystal oscillating at 1MHz. The 31,250 b.p.s. is derived as a subdivision of this clock (keep dividing by 2). Your Motu has UARTs inside it, probaby one per socket, although today they may be collected into one more powerful chip.

It's not so much what's inside the computer as what connects the computer to the external networking gear, or one MIDI device to another. One of the reasons that the original MIDI 1 was so widely adopted was because the addition of a UART or 2 and its supporting electronic components and DIN sockets was dirt cheap. Ergo, why not?

Currently we have data transmission technologies like USB, firewire, Ethernet etc. It will be these and the adoption of these technologies within our devices that will provide the higher speeds. Not, sadly, DIN plugs and cables. Some of these technologies provide severe limitations on the distance one object can be from another.

Thanks for the insights. I think you are right, it was may have been my MOTU that did the stepping down. But now I am thinking, with my Arduino Mega 2560, is there enough program memory to write some handlers to take 4 MIDI 1.0 sound modules, wrap the data in MIDI 2.0 packets and do something with it?  Glad I got the model with the quad UART, though from what I remember, there is code to simulate UARTs via software.

 

Edited February 3, 2020 by User 905133

[Toddskins]

With the discussion posts above about various tunings, I'm pretty sure all the high-end workstations provide numerous intonation tunings without the need of a new MIDI spec. 

All my Kurzweils going way back including my latest, have 16 different intonation maps to select from.  Now perhaps all those maps are using an internal MIDI file to achieve this, I don't know how it's done.

But the drawbacks that we've all complained about since almost MIDI 1.0's beginning was the latency the occurs when you daisy chain more than 3 keyboards (or other MIDI devices).  The new MIDI 2.0 spec should be vastly better and welcome.  And the part in the spec about syncing, too, will be great.

Edited February 3, 2020 by Toddskins

[JohnG]

8 hours ago, User 905133 said:

Thanks for the insights. I think you are right, it was may have been my MOTU that did the stepping down. But now I am thinking, with my Arduino Mega 2560, is there enough program memory to write some handlers to take 4 MIDI 1.0 sound modules, wrap the data in MIDI 2.0 packets and do something with it?  Glad I got the model with the quad UART, though from what I remember, there is code to simulate UARTs via software.

I dare say the Arduino is okay to process the code, but the reason the Motu kit is there is to buffer the message(s) and provide precise timing AND the appropriate voltage and current to drive the physical interface.

In the case of MIDI 2 the packetised MIDI data will have to be inserted into the packet structure of the transmission mechanism being used, i.e. USB, LAN, etc.

There's no point trying to send MIDI 2 data to your MIDI 1 sound modules as they will not understand it. Only with a major firmware upgrade (not necessarily possible) are they likely to make any sense of the new packet structure.

(Retired Principal Consultant/Data Communications). Last job was writing satellite data comms courses and delivering them to telco's worlwide.