MIDI 2.0

Any music gear discussions that don't fit into one of the other forums.

Moderators: lisa, Kent, Joe.

Post Reply
User avatar
trentpmcd
Wiggling with Experience
Posts: 331
Joined: Fri Mar 20, 2015 4:00 pm
Location: NH

MIDI 2.0

Post by trentpmcd » Sun Feb 02, 2020 12:34 pm

First, I put this in General Gear since this will affect all midi gear (eventually). And software. And... Hmmm. Where does it belong? I also did a search and was surprised to not see it any where. let me know if I am wrong and there is another thread about this or it belongs somewhere else (it's not DYI, so not Music Tech DYI, right?).

Anyway, on January 19, the Midi Manufacturers Association adapted MIDI 2.0. That's like 37 years after MIDI became the standard that it is....

I was surprised to see so little about it. I did a search and found this page, which is great: https://www.midi.org/articles-old/detai ... y-exchange

Some things they talk about - 2 way communication, higher resolution (hurray for converting to/from CV!!!), backwards compatibility, general profiles, specific properties, etc.

As I said, I'm surprised that I haven't seen any discussion here since this maybe a big change in the future equipment. Or not. I was just wondering what you all think about it.

User avatar
racooniac
Wiggling with Experience
Posts: 258
Joined: Tue Aug 11, 2015 6:53 am
Location: Austria, Salzburg

Re: MIDI 2.0

Post by racooniac » Sun Feb 02, 2020 1:18 pm

i cant await the now hopefully soon upcoming midi 2.0 hypetrain. :hyper:

roland has the hardware of one of their newer midi keyboards already "midi 2.0 ready" that just needs to be firmwareupdated once the protocol details are final and i am sure a lot of others will follow very soon once the word got around because the features are just so fucking awesome and in parts imho long overdue! (this got mentioned in the namm 2020 video linked below)

who has not seen it already there are two videos about midi 2.0 from the namm 2019 and another one from namm 2020 thats how i heard from it the first time.

p_shoulder
Common Wiggler
Posts: 59
Joined: Mon Oct 16, 2017 9:33 am

Re: MIDI 2.0

Post by p_shoulder » Sun Feb 02, 2020 7:42 pm

trentpmcd wrote:
Sun Feb 02, 2020 12:34 pm
As I said, I'm surprised that I haven't seen any discussion here since this maybe a big change in the future equipment. Or not. I was just wondering what you all think about it.
There are a fair number of people (myself included) who have been fairly enthusiastic about MPE (the polyphonic-expressive hack for MIDI 1.0), and it looks like we're *finally* seeing a lot more hardware manufacturers and software synthesizers support this, as well as a greater amount of controllers that support this. It's still niche (you really don't need MPE to make dance-pop) but it's a welcome one, I personally think adding a more "expressive" element to synthesizers changes the sort of pigeonhole you tend to find electronic instruments in. MIDI 2.0 sounds like a great extension to this expressiveness, IMHO, we'll see how quickly it's adapted (it took a bit for MPE to catch on.)

The other big possible thing I can envision with the 2 way communication is that it will make hardware integrated to DAWs a heck of a lot easier. While other gear has had this sort of integration in the past, it might be more "plug and play" with this protocol.

User avatar
trentpmcd
Wiggling with Experience
Posts: 331
Joined: Fri Mar 20, 2015 4:00 pm
Location: NH

Re: MIDI 2.0

Post by trentpmcd » Mon Feb 03, 2020 5:50 pm

racooniac wrote:
Sun Feb 02, 2020 1:18 pm
i cant await the now hopefully soon upcoming midi 2.0 hypetrain. :hyper:

roland has the hardware of one of their newer midi keyboards already "midi 2.0 ready" that just needs to be firmwareupdated once the protocol details are final and i am sure a lot of others will follow very soon once the word got around because the features are just so fucking awesome and in parts imho long overdue! (this got mentioned in the namm 2020 video linked below)

who has not seen it already there are two videos about midi 2.0 from the namm 2019 and another one from namm 2020 thats how i heard from it the first time.
Cool videos. I hadn't seen them before I like the answer near the end of the 2020 one that we'll see it "soon".
p_shoulder wrote:
Sun Feb 02, 2020 7:42 pm
trentpmcd wrote:
Sun Feb 02, 2020 12:34 pm
As I said, I'm surprised that I haven't seen any discussion here since this maybe a big change in the future equipment. Or not. I was just wondering what you all think about it.
There are a fair number of people (myself included) who have been fairly enthusiastic about MPE (the polyphonic-expressive hack for MIDI 1.0), and it looks like we're *finally* seeing a lot more hardware manufacturers and software synthesizers support this, as well as a greater amount of controllers that support this. It's still niche (you really don't need MPE to make dance-pop) but it's a welcome one, I personally think adding a more "expressive" element to synthesizers changes the sort of pigeonhole you tend to find electronic instruments in. MIDI 2.0 sounds like a great extension to this expressiveness, IMHO, we'll see how quickly it's adapted (it took a bit for MPE to catch on.)

The other big possible thing I can envision with the 2 way communication is that it will make hardware integrated to DAWs a heck of a lot easier. While other gear has had this sort of integration in the past, it might be more "plug and play" with this protocol.
There are several videos on the link I posted including some of the "plug and play" type stuff. One video is real instruments, but the other is still all in software. Still, pretty cool - your DAW will know what the box does, and a controller can customize itself for what it is playing. Also demoes of per-note volume and other other things.

One problem - With the exception of modular, I may wait to purchase new kit until I know it is midi 2.0 compatible to take advantage of some of the new features...

User avatar
Graham Hinton
Super Deluxe Wiggler
Posts: 3000
Joined: Fri Jul 02, 2010 6:28 pm
Location: England
Contact:

Re: MIDI 2.0

Post by Graham Hinton » Tue Feb 04, 2020 12:29 pm

trentpmcd wrote:
Sun Feb 02, 2020 12:34 pm
Some things they talk about - 2 way communication, higher resolution (hurray for converting to/from CV!!!), backwards compatibility, general profiles, specific properties, etc.

As I said, I'm surprised that I haven't seen any discussion here since this maybe a big change in the future equipment. Or not. I was just wondering what you all think about it.
When MIDI 1.0 came out it was proven, after years of hard work out of public view. Talking about incomplete, unproven specs in public is dangerous. Anyone remember what happened to Lone Wolf or MIDILan?

No mention of galvanic isolation or transmission delays.
Most of the features are DAW-centric ignoring using MIDI devices together directly in real time.

One thing that is obvious is that to translate one transport system to MIDI 1.0 is going to add delays. So it does nothing to improve millions of MIDI devices that have been produced in the last 36 years.

16 bit velocity eh? I wonder how they are going to detect key closures to that accuracy when there is not that much information there...

"Just" a firmware upgrade? If you believe that you'll believe anything.

User avatar
racooniac
Wiggling with Experience
Posts: 258
Joined: Tue Aug 11, 2015 6:53 am
Location: Austria, Salzburg

Re: MIDI 2.0

Post by racooniac » Tue Feb 04, 2020 1:08 pm

Graham Hinton wrote:
Tue Feb 04, 2020 12:29 pm
When MIDI 1.0 came out it was proven, after years of hard work out of public view. Talking about incomplete, unproven specs in public is dangerous.
And its totally fine that they do it again this way with 2.0
Graham Hinton wrote:
Tue Feb 04, 2020 12:29 pm
One thing that is obvious is that to translate one transport system to MIDI 1.0 is going to add delays. So it does nothing to improve millions of MIDI devices that have been produced in the last 36 years.
i havent calculated exactly how much more precise 2.0 is in comparision to 1.0 in terms of resolution and bandwidth but i imagine that an 2.0 capable interface can spit out a much tighter 1.0-midi than midi interfaces which have been built to be able to output midi 1.0 only (midi clock jitter anyone!?)
if you speak about standalone translator-boxes then yeah we have added latency but i expect any midi 2.0 capable interface to be just able to spit out midi 1.0 "directly" so there should be really no need for such boxes imho.
Graham Hinton wrote:
Tue Feb 04, 2020 12:29 pm
"Just" a firmware upgrade? If you believe that you'll believe anything.
If you build the hardware of a midi keyboard with the resolution specs of midi 2.0 already in mind, why should it take more than a firmware upgrade to support it? you have already all the hardware on board you need .... processing power is not expensive anymore i am sure it has enough free ressources left so they dare to promise such a thing.

miminashi
Wiggling with Experience
Posts: 293
Joined: Tue Aug 19, 2008 3:14 pm
Location: SLC, Utah

Re: MIDI 2.0

Post by miminashi » Tue Feb 04, 2020 4:18 pm

Graham Hinton wrote:
Tue Feb 04, 2020 12:29 pm
One thing that is obvious is that to translate one transport system to MIDI 1.0 is going to add delays. So it does nothing to improve millions of MIDI devices that have been produced in the last 36 years.
The jitter-reduction timestamps could be beneficial for MIDI 1.0 devices being controlled via a 2.0 interface, if the interface is using the timestamps to intelligently serialize the messages instead of splatting them out every millisecond as often seems to be the case with with USB MIDI interfaces. I'd gladly pay whatever small increase in latency that incurs, since latency is easy to account for but the same thing can't be said about jitter.

p_shoulder
Common Wiggler
Posts: 59
Joined: Mon Oct 16, 2017 9:33 am

Re: MIDI 2.0

Post by p_shoulder » Wed Feb 05, 2020 2:37 pm

Graham Hinton wrote:
Tue Feb 04, 2020 12:29 pm
16 bit velocity eh? I wonder how they are going to detect key closures to that accuracy when there is not that much information there...
The luxury of using bigger power of 8 byte standards (eg 64 bit messages for general MIDI data) is that you get room for things that need it. 7 bit is fine for initial velocity, but for controller data, it's been pretty clear for a long time that 7 bit will cause audible steps or other similar phenomenon unless a smoothing algorithm is applied. There will probably be much less effort needed for smoothing algorithms with the proposed 32 bit resolution (probably none at all to be honest). I think the same would apply actually to a lesser extent to continuous velocity function (eg aftertouch), too.

Now, one can argue that 32 bit for controllers *also* is overkill in most cases, and it probably is. However, recall that the current default bit rate for "integers" for most computer systems these days is 64 bit. Even more basic digital equipment such as microcontrollers these days is usually 32 bit. So I think *some* of the reason for some of these choices is simply because we can -- and, also, that it aligns more with the number of bits of the variables you're going to end up storing parameter data in the first place.

User avatar
daveholiday
Veteran Wiggler
Posts: 642
Joined: Thu Feb 19, 2015 4:35 pm
Location: Augusta, GA

Re: MIDI 2.0

Post by daveholiday » Wed Feb 05, 2020 7:48 pm

I am very excited about the new developments for the MIDI protocol. But to be honest, I don't think it will change anything for me! I have been using MIDI devices for 30 years now.....started out with a computer and a GM Roland synth. I learned over the years that MIDI has many strengths and several limitations. Since the vast majority of what I do is studio recording, I know how to optimize MIDI data to and from the computer to get jitter/latency/sloppyness to a point that I don't perceive any timing issues. Lab conditions vs live conditions, a big difference I know!

Timing I would hope will be the biggest improvement. I am not discounting resolution, but zippering from CC data filter sweeps never even played a part with me. It fact, sometimes I would thin the CC data to "stairstep" in time to create more rhythmic variation of just a simple, long filter sweep. I understand that this might be more of an issue with other types of music aside from what I like to do. But playing back 10-12 MIDI tracks across as many instruments with accurate timing would be nice! Or having a MIDI track going to a single synth containing note,velocity, 10 different CC tracks, blah blah blah, without worrying about overloading would be a welcome improvement for sure.

I know that 2.0 will have little benefit to old devices, but perhaps 2.0 will help tighten up the interaction between DAW and MIDI interfaces that will support it?
It is all shits and giggles,

until somebody giggles and shits.

p_shoulder
Common Wiggler
Posts: 59
Joined: Mon Oct 16, 2017 9:33 am

Re: MIDI 2.0

Post by p_shoulder » Thu Feb 06, 2020 11:23 am

MIDI 2.0 has a proposed "jitter reduction (JR)" timestamp message that will apparently even work on MIDI 1.0 gear. Not sure of the details beyond that but it sounds like they are considering improving timing as part of the spec.

Post Reply

Return to “General Gear”