Jump to content


Please note: You can easily log in to MPN using your Facebook account!

MIDI 2.0: a tantalizing first look


Recommended Posts

Thanks to the timely intervention of my friend Dean Swan, who interacts a fair bit with members of the MIDI Manufacturers' Association, I got clued into the press demo of new MIDI 2.0 tech that was being given in a small demo room in the Yamaha space at the Marriott at NAMM. I will write more about it in the coming days, but wanted to describe the two or three very simple demos I got to see.

 

The first one was a Roland synthesizer which had been equipped with a TouchKeys overlay, controlling a prototype Korg hardware synth (a modded minilogue xd, I think) and a Yamaha FM software program, with side to side and forward/back key gestures. Everything was being connected and routed with external microcomputers (Raspberry pi etc), as nothing's been internalized yet.

 

The demonstrator was obviously not an MPE-savvy player, and the resulting "music" was kind of painful to listen to, but the demonstration worked. There was nothing shown that an MPE controller can't do just as well, but that wasn't the point. The point was: all of the data being sent was MIDI 2.0 data, absolutely bog-standard, now baked into the language! Every note now has properties that are applied when the note is struck, as well as per-note control of multiple parameters while it's held down. No more multichannel logic or obscure commands that not all gear can recognize.

 

MIDI 2.0 takes advantage of modern connection protocols. It's cable-agnostic (no more 5-pin plugs with UARTs that get fried if you zap your synth with static electricity), and works over any high-speed bidirectional cable, starting with USB 2 and working up to Ethernet and Thunderbolt 3. One of the things that's built into the spec now is something called a Profile. Each device can have a Profile on board, saved in the firmware, which can be communicated automatically to anything you plug it into. Profiles can be written by anyone and can serve pretty much any purpose you can think of within the realm of "hello, you other device, I am this device, and I can do the following and want you to work with me on this."

 

One example is a synth that comes with a patch editor built in. Plug it into your computer and boom, up comes a screen with the editing software ready to use, no need to install or authorize or update. On a simpler level, imagine never having to document all your hardware synth data separately as part of a DAW session any more: a synth's full state is saved with your project and instantly recalled when you fire it up, without having to send a bunch of carefully prepped SysEx to each box.

 

Last but not least, here's a movie you might find interesting. This shows a fader (on the right) being moved with traditional 7-bit MIDI Control Change messages, vs. a fader (on the left) being moved with the new UMP protocol in MIDI 2.0. No more zippering or ratcheting; all data now have a stupidly-overspec'd 32-bit resolution. Pretty trippy!

 

 

Dr. Mike Metlay (PhD in nuclear physics, golly gosh) :D

Musician, Author, Editor, Educator, Impresario, Online Radio Guy, Cut-Rate Polymath, and Kindly Pedant

Editor-in-Chief, Bjooks ~ Author of SYNTH GEMS 1

 

clicky!:  more about me ~ my radio station (and my fam) ~ my local tribe ~ my day job ~ my bookmy music

Link to comment
Share on other sites



  • Replies 30
  • Created
  • Last Reply

Top Posters In This Topic

Top Posters In This Topic

One example is a synth that comes with a patch editor built in. Plug it into your computer and boom, up comes a screen with the editing software ready to use, no need to install or authorize or update. On a simpler level, imagine never having to document all your hardware synth data separately as part of a DAW session any more: a synth's full state is saved with your project and instantly recalled when you fire it up, without having to send a bunch of carefully prepped SysEx to each box.
:ohmy:

 

Samuel B. Lupowitz

Musician. Songwriter. Food Enthusiast. Bad Pun Aficionado.

Link to comment
Share on other sites

I caught most of the MIDI 2.0 introductory session at NAMM on Friday the 17th (that needed a much bigger room!) and was very impressed by what MIDI 2.0 will be. Like the various presenters said, it's exciting times. Just like we didn't know what MIDI was going to be able to do, MIDI 2.0 may take us some amazing places and result in some gear no one has even conceived of yet. :2thu:

"I'm so crazy, I don't know this is impossible! Hoo hoo!" - Daffy Duck

 

"The good news is that once you start piano you never have to worry about getting laid again. More time to practice!" - MOI

Link to comment
Share on other sites

I caught most of the MIDI 2.0 introductory session at NAMM on Friday the 17th (that needed a much bigger room!) and was very impressed by what MIDI 2.0 will be. Like the various presenters said, it's exciting times. Just like we didn't know what MIDI was going to be able to do, MIDI 2.0 may take us some amazing places and result in some gear no one has even conceived of yet. :2thu:

That's the hope. If we remember that the first MIDI demos only had one synth playing another, there's lots of room for 2.0 to take us way beyond what we're looking at now. Watch this space...

 

Dr. Mike Metlay (PhD in nuclear physics, golly gosh) :D

Musician, Author, Editor, Educator, Impresario, Online Radio Guy, Cut-Rate Polymath, and Kindly Pedant

Editor-in-Chief, Bjooks ~ Author of SYNTH GEMS 1

 

clicky!:  more about me ~ my radio station (and my fam) ~ my local tribe ~ my day job ~ my bookmy music

Link to comment
Share on other sites

This is very exciting. I'm hoping the potential for auto CC assignments or whatever new message protocol may result between one manufacture's device and another is well implemented. VIP software was used by several manufacturers including Akai and Novation but seemed to fall well short of the advertised capability. I always ended up manually configuring everything.

 

I'm also looking forward to 30 years from now when 7 bit "zippering and ratcheting" is all the rage and featured on every new release. Hang on to those soon to be vintage midi 1.0 synths. :D

Link to comment
Share on other sites

This is very exciting. I'm hoping the potential for auto CC assignments or whatever new message protocol may result between one manufacture's device and another is well implemented. VIP software was used by several manufacturers including Akai and Novation but seemed to fall well short of the advertised capability. I always ended up manually configuring everything.

 

I'm also looking forward to 30 years from now when 7 bit "zippering and ratcheting" is all the rage and featured on every new release. Hang on to those soon to be vintage midi 1.0 synths. :D

Yeah, I can't wait, cough hack.

 

Autoconfiguration shells were always tricky. Everyone tried them but they rarely worked. Novation dumped their protocol some years back; no clue if the InMusic brands (Akai, Alesis, M-Audio) are still pushing VIP...

 

Dr. Mike Metlay (PhD in nuclear physics, golly gosh) :D

Musician, Author, Editor, Educator, Impresario, Online Radio Guy, Cut-Rate Polymath, and Kindly Pedant

Editor-in-Chief, Bjooks ~ Author of SYNTH GEMS 1

 

clicky!:  more about me ~ my radio station (and my fam) ~ my local tribe ~ my day job ~ my bookmy music

Link to comment
Share on other sites

  • 10 months later...

It is done....

 

https://www.musicradar.com/news/midi-20-spec-confirmed-the-biggest-advance-in-music-technology-in-decades?fbcli

 

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

Yamaha CP88, Casio PX-560

Link to comment
Share on other sites

Great News, only ...

 

You'd better start saving right away so that you can afford to replace ALL of your old MIDI 1.0 gear with that super-duper MIDI 2.0 technology.

 

I may be wrong but I don't think there'll be an upgrade for my AN1x, nor my VL70m, nor my Akai EWI4000s, nor my PX560, nor my ... etc.

Then there'll be all my DAWs and all my VSTi's and ... etc.

 

Hope your pockets are deep enough, mine aren't.

 

(I know ... wet blanket!)

 

But then I'm still running Windoze 7 on a quad core laptop and loving it.

No OS breaking updates for me. Wonderful.

 

Have fun.

Akai EWI 4000s, Yamaha VL70m, Yamaha AN1x, Casio PX560, Yamaha MU1000XG+PLGs-DX,AN,VL.

 

Link to comment
Share on other sites

Just realised, I'm wrong about the DAWs.

 

Since there's no MIDI 2.0 file format specified yet, there's no standardised way of recording all these wonderful new solutions that MIDI 2.0 is going to give us. ... Oh well!

 

Unless, that is you know differently!

 

And ... as the only connection specification available so far is USB, are we always going to need a 'box' in between two MIDI 2 devices to provide the host to slave / slave to host connections?

Is the fact that all these messages are going through such a pasthrough USB controller going to introduce latency?

Don't expect them to be too far apart, the USB spec needs the cables to be relativley short compared, that is, to the Din cables we're used to. :-(

 

Or are we going to use an Ethernet connection in the future to provide multi-connectivity?

Will I then have to insert an Ethernet controller into every MIDI 2 compliant device with the appropriate protocol stack (IEEE 802.whatever) implemented in the firmware?

 

Better save a little bit more than you were expecting for your new keyboard!

 

Or am I totally wrong?

Akai EWI 4000s, Yamaha VL70m, Yamaha AN1x, Casio PX560, Yamaha MU1000XG+PLGs-DX,AN,VL.

 

Link to comment
Share on other sites

MIDI 2.0 is backward-compatible with MIDI 1.x. IOW, 2.0 devices can simplify their language for 1.x devices, and the 1.x devices don't get mad and feel spoken down to.

 

There are some really good videos and such about MIDI 2.0 out there. I would suggest you check out MIDI.org, notably the video "Introduction to MIDI 2.0" on the home page.

"I'm so crazy, I don't know this is impossible! Hoo hoo!" - Daffy Duck

 

"The good news is that once you start piano you never have to worry about getting laid again. More time to practice!" - MOI

Link to comment
Share on other sites

Thanks Joe,

 

Yes, I am aware of the MMA web site.

In fact, when they revamped the site some years ago, because of my contributions to the original MIDI forum and my own forum, I was made a moderator of the MMA forum.

I've been working with MIDI since the late eighties.

 

My background is as a data communications protocol analyst since the early seventies, from modems in the early days to satellite in the early noughties, working for mainframe manufaturers, telcos and global satellite companies, before I retired a few years back.

It's this background that has taught me that just having a very detailed MIDI protocol is totally insufficient, we need the physical network over which the protocol will flow that will provide the connectivity that people will need.

For example, how will I implement a MIDI 2 patch bay? The only solution I see MIDI is over Ethernet with an Ethernet switch and structured cabling, that will provide sufficient multiconnectivity and cable lengths suitable for running around a studio.

I'm probably in the dark about what is happening here, but in my view the protocols and the interwiring need to happen hand in hand.

USB provides a host to slave architecture, and whilst that is okay for a simple link, it is more problematic when we're talking slave to slave. It needs an intervening processor to forward data from one terminal to another.

 

I've been following the development of MIDI 2 for some time (downloaded and read and hopefully understood all the published specifications) and whilst I applaud the efforts made so far in terms of protocol development, I am disappointed with the lack of apparent progress in terms of connectivity.

Perhaps I need to learn a little more patience?

 

I also find it extremely short sighted not to have addressed what I consider to be a relatively straightforward issue, namely the ability within the protocol to address the fact that enharmonic notes are only the same on a fixed pitch instrument e.g. a piano.

They're not, for instance, on a violin. That is A# is not the same, or needn't be the same, as Bb. This could be quite easily resolved by utilising the extra bit available in the MIDI note number, now that we aren't restricted to 7 bit data fields.

 

A much easier tuning system could have been introduced too based on the widely implemented Scala temperament files.

See http://huygens-fokker.org/scala/

 

But there you go, I'm just a lone voice crying in the wilderness, not a member of some acknowledged music instrument company or a journalist.

Akai EWI 4000s, Yamaha VL70m, Yamaha AN1x, Casio PX560, Yamaha MU1000XG+PLGs-DX,AN,VL.

 

Link to comment
Share on other sites

It's this [John G11's] background that has taught me that just having a very detailed MIDI protocol is totally insufficient, we need the physical network over which the protocol will flow that will provide the connectivity that people will need.

For example, how will I implement a MIDI 2 patch bay? The only solution I see MIDI is over Ethernet with an Ethernet switch and structured cabling, that will provide sufficient multiconnectivity and cable lengths suitable for running around a studio.

 

I haven't been studying MIDI 2.0, just following its general progress in terms of what it provides to the user. Personally, I think that it will be a long time before even some of the musical additions such as higher resolution volume and pitch variations will find creative use. I could debate that point, but I don't want to forget why I replied to this one point in your post.

 

There's no reason why a MIDI patch bay can't be electro-mechanical like an analog patch bay. That always works. And why will we need a MIDI patchbay anyway? We can already assign any MIDI output to any MIDI input through software. I had (still have, in a rack somewhere) a Digital Music Corp MX-8 MIDI routing switcher that let me send keyboard data to any one or combination of sound modules or the computer, or from the computer to sound modules. Today we do things differently, without all the hardware. You can send a MIDI track (or copies of that track) in a DAW to any number of virtual instrument plug-ins, and that same setup - or maybe a simpler software program - can be used in live performance.

 

I recognize that there are other applications for MIDI (despite the original meaning of "MI" in the name) than playing musical instruments, and some of those applications may indeed require Ethernet-like routing techniques. But music is by far the greatest application for MIDI, so I believe that's going to be supported first, and also be first to say "OK, that's enough work for now, let's move on to other applications."

 

But more along the line that you're thinking, Dante and a few other audio over IP protocols are getting more popular and less expensive, and that's just the sort of configuration that you're describing. You might want to look at the possibility and practicality of MIDI over Dante. It seems like a natural path.

Link to comment
Share on other sites

 

There's no reason why a MIDI patch bay can't be electro-mechanical like an analog patch bay. That always works.

 

I'm afraid, Mike, that with only USB defined as the connecting mechanism for MIDI 2, there's every reason why a simple mechanical interface won't work. USB is a polled protocol. Some 'intelligence' has to do the polling for the USB data packets and then extract the MIDI command(s) from them, rerouting them to the appropriate port or channel. That's why in the demo's there is a raspberry pi between devices, it's acting as the USB host to the keyboard slaves, constantly polling them both for data and forwarding it to the other device. Unless, that is you implement both host and slave functions in every device and then, in effect, we've gone back to an 'in' and 'out' configuration. And more expense. And MUCH shorter cable lengths.

 

And why will we need a MIDI patchbay anyway? We can already assign any MIDI output to any MIDI input through software. I had (still have, in a rack somewhere) a Digital Music Corp MX-8 MIDI routing switcher that let me send keyboard data to any one or combination of sound modules or the computer, or from the computer to sound modules. Today we do things differently, without all the hardware. You can send a MIDI track (or copies of that track) in a DAW to any number of virtual instrument plug-ins, and that same setup - or maybe a simpler software program - can be used in live performance.

 

Your MX-8 won't handle MIDI 2 commands encapsulated in USB packets and won't poll USB devices. It's MIDI 1.

Your DAW (at least I haven't seen any yet) doesn't support MIDI 2 commands. So you can't route them anywhere.

At this current time we can ONLY use USB as the transport mechanism. USB is a polled protocol interface totally unlike MIDI 1 Din. IMV USB poses significant interconnection problems.

 

I recognize that there are other applications for MIDI (despite the original meaning of "MI" in the name) than playing musical instruments, and some of those applications may indeed require Ethernet-like routing techniques. But music is by far the greatest application for MIDI, so I believe that's going to be supported first, and also be first to say "OK, that's enough work for now, let's move on to other applications."

 

No, I'm only thinking about music applications.

For example currently, I create a few of split points on my keyboard, I have channel one and two going to a Yamaha MU1000 equipped with three PLG cards, DX, AN and VL, another chanel going to an ancient Roland SC8850 sound module all via MIDI Din.

I can either cascade via MIDI din, or go via a MIDI router such as yours.

Assuming I can get MIDI 2 replacements for the sound modules, and my keyboard firmware is updateable to MIDI 2 (however unlikely) how do I do that over one USB out from the keyboard?

I need a Rasberry PI in the middle that examines the incoming USB data, extracts the MIDI 2 commands from the packets, then re-encapsultes the MIDI data into new USB packets, then reroutes them to the appropriate USB port.

It can be done (in fact I could probably code it myself), but when will it?

Just how much latency and/or jitter will that introduce?

 

I haven't heard of any DAW makers make any predictions/promises about MIDI 2 yet. Have you?

 

But more along the line that you're thinking, Dante and a few other audio over IP protocols are getting more popular and less expensive, and that's just the sort of configuration that you're describing. You might want to look at the possibility and practicality of MIDI over Dante. It seems like a natural path.

I must admit I don't know Dante. I'll do some research.

It needs to be an open standard like Ethernet, or USB are, not proprietary.

 

Having said all that, MIDI 2 promises to offer a lot more flexibility, especially to synth users.

It's just that it's an awful long way from being implemented in anything that is half way usable. IN MY VIEW.

 

I'm not trying to put it down, although it may appear that way I admit, I just would like people to have realistic expectations about it.

 

In my seventy something years I've been promised the earth and been disappointed too many times to take the marketing hype at face value any more.

 

Back in the days Dave Smith and Ikutaro Kakehashi got it right. (IMHO)

 

P.S. It seems Dante is audio and video not MIDI (either 1 or 2).

Akai EWI 4000s, Yamaha VL70m, Yamaha AN1x, Casio PX560, Yamaha MU1000XG+PLGs-DX,AN,VL.

 

Link to comment
Share on other sites

I'm just a casual observer throwing out ideas and letting others solve the problems.

 

With 5-pin MIDI, we could use hardware to route and split data. How about a box that's like an Ethernet router, but for USB instead. If you could plug a several USB devices into it and tell it "I want this keyboard over here to play that synth over there" and "this DAW track will play that other synth." Or maybe the routing can be done by a computer application.

 

Dante is an open standard - they're happy to have more new applications - but there's a license associated with its use. AVB is similar in function, and is free. Both were designed with audio and video in mind because there's a big market for that. But they're all data. I don't mean to imply that it's going to be easy to interconnect devices with new and old protocols. I've seen enough development cycles to know better. But when somebody important enough wants to do something, it'll get done.

 

And don't start buying system components before the system design is finished. I've seen Government agencies that I've worked for do that a few times, over my objections. Contractors' proposals are great. Government specifications aren't always all-encompassing.

 

 

r6ZBu.jpg

Link to comment
Share on other sites

FYI, Craig Anderton over at the SSS forum is the current MMA President so you might be interested in talking with him.

"I'm so crazy, I don't know this is impossible! Hoo hoo!" - Daffy Duck

 

"The good news is that once you start piano you never have to worry about getting laid again. More time to practice!" - MOI

Link to comment
Share on other sites

I'm not trying to put it down, although it may appear that way I admit

It really does, actually. But maybe that's just me.

I just would like people to have realistic expectations about it.

There's a difference between realism and pessimism. But maybe that's just me.

 

In my seventy something years I've been promised the earth and been disappointed too many times to take the marketing hype at face value any more.

On the other hand, in my nearly sixty years (OK Boomer), I've been pleasantly surprised enough by what's eventually delivered to not knock the marketing hype too heavily.

 

25 years as a pro audio journalist also tempered my opinion of marketing types. The last company that pushed a new product while admitting caveats destroyed its US business from millions of dollars down to zero in a matter of months. Anyone remember Opel?

 

Back in the days Dave Smith and Ikutaro Kakehashi got it right. (IMHO)

And has it been forgotten that MIDI wasn't really off and running with its survival assured until early 1985? Manufacturers who had their own interfacing systems (e.g. Oberheim) were scathing about how poor a solution MIDI was going to be, and respected industry leaders (e.g. Bob Moog) stated in print that while MIDI 1.0 was a wonderful gesture of industry cooperation, there was no guarantee that it would stick. It wasn't until the 1986 NAMM Show that it was obvious that this new MIDI thing was here to stay.

 

I'm giving 2.0, and all the people currently getting into all the stuff it will be able to do, at least that long before I decide that wide-eyed optimism isn't the way to go. But maybe that's just me.

Dr. Mike Metlay (PhD in nuclear physics, golly gosh) :D

Musician, Author, Editor, Educator, Impresario, Online Radio Guy, Cut-Rate Polymath, and Kindly Pedant

Editor-in-Chief, Bjooks ~ Author of SYNTH GEMS 1

 

clicky!:  more about me ~ my radio station (and my fam) ~ my local tribe ~ my day job ~ my bookmy music

Link to comment
Share on other sites

Well, Dr Mike, I dare say it does, I dare say it does.

Call me old and disillusioned, for that's what I am.

 

I've been heavily inlvolved in data communications from the early seventies and have a number of 'firsts' to my name in that field, which I won't go into here as it''s not appropriate.

Suffice it to say that I've worked on and off on various national and international standards committees to do with LAN's, WAN's and telephony over the years, and the implementation of networks on an international basis, both terrestrial and via satellite.

This on prototype Ethernet, ISDN BRI, ADSL, GAN and BGAN networks.

 

To date, my experience has been that the development of protocols and line codes typically goes hand in hand with the development or designation of the physical network over which they will run.

 

It is what made MIDI 1 great. The Din interface and the MIDI command structure.

It dismays me that the same, at least apparently, is not happening with MIDI 2.

USB, whilst possibly answering the home studio and simple set up requirement, simply doesn't cut it, for example, in a studio.

 

How, for instance does MIDI 2 plan to replace the piece of kit shown below?

How, using USB, could this be done and allow the length of connection that this MindBurner allows?

 

Mindburner_midi_expander_setup.jpg

 

I look at MIDI 1 equipment and what do I find?

Manufacturers of modern, fairly expensive devices not using the 14 bits available for pitch bend, in many cases not even 10 bits, a few not even 7 bits, in a area where the human ear is most sensitive.

And do we honestly expect them suddenly to implement even 16 bits for note velocity?

Somehow I have my doubts.

 

I hope I'm wrong, Dr Mike, I really do, but I'll remain pessimistic until I see some signs of reality emerge from the MMA.

We need an open Ethernet solution.

We need some signs of a MIDI 2 file specification.

 

They could, so easily, have solved the enharmonic note differences instead of thinking only of fixed pitch instruments.

 

I'm just disappointed. Very disappointed.

Protocols aren't everything.

(I can't believe I'm saying that, when most of my life I've worked with IBM 2780, ICLC03, TCP/IP, IEEE Q.9321/931, 2B1Q, etc.)

 

[/rant] Sorry.

 

JohnG.

1721.jpg.1616fd8d55ee4f85f98a0ba2928cb072.jpg

Akai EWI 4000s, Yamaha VL70m, Yamaha AN1x, Casio PX560, Yamaha MU1000XG+PLGs-DX,AN,VL.

 

Link to comment
Share on other sites

Perhaps I should add, I remain committed to MIDI as a moderator of the MMA forum, and daily help to remove the flood of spammers.

And, from time to time help implementors get kit working or give advice on MIDI file editing.

 

I'm not just a moaner.

Not all the time anyway. ;-)

Akai EWI 4000s, Yamaha VL70m, Yamaha AN1x, Casio PX560, Yamaha MU1000XG+PLGs-DX,AN,VL.

 

Link to comment
Share on other sites

What is the "MIDI Burner" that you pictured? What's its function? There may not be a drop-in replacement but there may be a functional replacement, or could be if anyone wants to buy it.

 

There's always some evolution with new standards. Give it time to grow. In the mean time, you can continue to use what you know.

Link to comment
Share on other sites

What is the "MIDI Burner" that you pictured? What's its function? There may not be a drop-in replacement but there may be a functional replacement, or could be if anyone wants to buy it.

 

There's always some evolution with new standards. Give it time to grow. In the mean time, you can continue to use what you know.

 

This one is a twenty something port MIDI 1 thru/expander.

Take a look here where Rara found a used(?) one.

https://yamahamusicians.com/forum/viewtopic.php?f=4&t=16942

Akai EWI 4000s, Yamaha VL70m, Yamaha AN1x, Casio PX560, Yamaha MU1000XG+PLGs-DX,AN,VL.

 

Link to comment
Share on other sites

What is the "MIDI Burner" that you pictured?

 

This one is a twenty something port MIDI 1 thru/expander.

Take a look here where Rara found a used(?) one.

 

I wasn't curious enough to register in order to read the whole message, but it sounds like what we call in modern technospeak a "boutique" piece - I'm not sure if it qualifies as "vintage" but from your photo, it looks like it's certainly an "industrial" build. Is there more than one product that does its job? It appears to have solved a problem for one user, but what problem did it solve? I assume that it's more than a hub, that it offers some routing capabilities between ports. And is there a compelling need for this user (or you) to switch to MIDI 2 right now? When MIDI 1 was released, there were a number of manufacturers who had their own interconnection system and it took several years for them to integrate MIDI in their products. And one of the first of such interconnection systems for music devices, DC control voltage (CV), with the fresh interest in modular synths, is probably hotter today than in "its" day.

 

Seems like (says me, the non-programmer) that a box with a bunch of USB ports and a Raspberry Pi or similar be able to solve that user's problem? Or a custom IC? If there's a sufficient need, someone will make it.

 

I'm not trying to start an argument - I get what functionality you're after, and it's certainly useful to some users. I'm just saying that I think you're stirring the pot before all the ingredients have been put into the soup.

 

And remember that there's more MIDI being used by individual musicians and composers whose functional setup for having multiple interconnected synths is a DAW with a bunch of virtual instrument plug-ins. It's not the ideal setup for stage performance - for that you want roadworthiness and reliability, but it offers a wide range of sounds and sound design. So when commercial MIDI 2 products start hitting the market, they'll probably be supporting the small time users first because that's where the most customers are.

Link to comment
Share on other sites

JohnG11, these are all fair points, and you're certainly entitled to skepticism, but I maintain my optimism -- all while holding on to my remaining MIDI 1 DIN gear, such as it is.

 

I got rid of a fair bit of it over the years, because simple connectivity and much tighter timing made MIDI 1 over USB were big draws. Hell, even MIDI over Bluetooth has tighter timing than MIDI over DIN, and in a quiet studio at ranges of under 10 feet it's pretty much bulletproof. (I admit I am not yet confident enough to use it, much less rely on it, on stage in a potentially data-noisy venue. Same for wi-fi without my own dedicated router, which I do take with me to gigs as it's barely larger than a smartphone.)

 

Things do change, and updates do happen even within an existing spec, but builders who get into MIDI 2 are going to (for better or worse) leapfrog over the existing limitations of MIDI 1 (such as hardware that doesn't even send 14-bit data when it could) and get into the new and exciting stuff. My background is in alternative controllers, and support for specs that will take the MPE standard and run hard with it can only be a good thing as far as I'm concerned. There will be hiccups, but that's the nature of tech, and it will be a long damn time before we get enough market penetration for people to even think of setting aside MIDI 1 hardware.

 

Just my 5 Lindens' worth. :)

Dr. Mike Metlay (PhD in nuclear physics, golly gosh) :D

Musician, Author, Editor, Educator, Impresario, Online Radio Guy, Cut-Rate Polymath, and Kindly Pedant

Editor-in-Chief, Bjooks ~ Author of SYNTH GEMS 1

 

clicky!:  more about me ~ my radio station (and my fam) ~ my local tribe ~ my day job ~ my bookmy music

Link to comment
Share on other sites

Sorry Mike, I got the product name wrong, it's a MindBurner 20 way MIDI thru splitter, studio edition. It has 20 assignable MIDI DIn sockets. Just Google mindburner MIDI.

The price? An emminenly affordable 84 pounds 99p.

Put simply, it allows you to direct MIDI channels to various devices.

 

Yes, a box with a processor and a number of USB ports could do it.

The problem is that unpacking potentially multiple MIDI commands from within USB packets and repackaging them into other USB packets for onward transmission is a LOT more complex than simply rerouting MIDI messages.

With MIDI 1 it's a doddle. I wrote some code for an Atari way, way back that did something similar. Very straightforward.

 

To add to the confusion, the MIDI 2 messages can now include a 4 bit "group" number as part of the messsage as well as channel number.

A MIDI 2 format mesaage can include a MIDI 1 channel message OR a MIDI 2 channel message, which further complicates routing.

See M2-104-UM_v1-0_UMP_and_MIDI_2-0_Protocol_Specification" sections 3 and 4.

And even when one has solved the routing issues, the cable length for USB is totally unsuited to studio use, not to mention the earthing issues.

 

As you said "if there's sufficient market for it", and there I think you have hit the nail on the head.

How many people need 16 bit velocity or other 16 bit CC's? A few synth users certainly, but elsewhere?

If major manufacturers can't be bothered any more to put even 10 bit pitch bend in their market leading products, what hope for all this extra complexity?

I'm guessing they'll all go for Property Exchange (so they can claim MIDI 2 conformance) and the message they'll send is "we send MIDI 1 messages".

 

I can hear you saying "you old cynic you!" And I am.

 

Dr Mike.

 

To be honest I can see some of the problems being ironed out over time, the question is "how long?"

Will I still be around?

Well, I'm 74 now, so maybe, maybe not.

The brain still seems to fuction okay (apart from the moaning) I'm not so sure about the physical side tho'.

 

A question. How many turns of an encoder do I have to make to go from zero to sixty five thousand five hundred and thirty five? I.e. 16 bits.

 

And, to be brutally frank, I can't see anything in MIDI 2 that I yearn for.

 

There are several other issues which I could expand upon, but I think you've all probably had enough of this moaning, cantankerous(?) old git.

I'll go back to sleep for a while.

Akai EWI 4000s, Yamaha VL70m, Yamaha AN1x, Casio PX560, Yamaha MU1000XG+PLGs-DX,AN,VL.

 

Link to comment
Share on other sites

MindBurner 20 way MIDI thru splitter, studio edition. It has 20 assignable MIDI DIn sockets.

Put simply, it allows you to direct MIDI channels to various devices.

 

According to the web page description, it's a 1 in 20 out splitter. To me, that means that every output gets a copy of the data that comes in. And it's one way only - you can't tickle the keys of the synth plugged into Port 3 and play the synth plugged into Port 15. You could send out the 20 slugs of the data byte on, say MIDI Channel 5, and, when set to the Poly mode (I think that's right) only the synths set to respond only to Channel 5 will play. Or if you have a multi-voice synth set to Mono mode, the channel in the data slug will define which of its voices will play. And if everything is in the omni mode, then you'll have the cacophony that you might be after.

 

I don't see any reason why a USB hub couldn't work in the same way. Now if you want something smart like my Digital Music MX-8, you can do all sorts of things to the data on a port-by-port basis whether it's coming or going. And you could make something like that with a processor inside a box of USB ports. And if the channels-and-modes doesn't give you enough flexibility, as long as you have a processor in the box you can send very specific information to each port,

 

Yes, a box with a processor and a number of USB ports could do it.

The problem is that unpacking potentially multiple MIDI commands from within USB packets and repackaging them into other USB packets for onward transmission is a LOT more complex than simply rerouting MIDI messages.

 

Agreed, but why do you have to do that? Or maybe a better question is how is this different from a USB hub used to split a port out to several ports? How much decoding and reassembling the data is needed for the computer to understand that what's coming in through the port that's connected to the audio interface should be available to an audio program? And when you move the mouse, the synth doesn't play a note unless a program in the computer (not the hub) tells it to.

 

But nobody ever said that MIDI was simple.

 

And even when one has solved the routing issues, the cable length for USB is totally unsuited to studio use, not to mention the earthing issues.

 

Huh? I have a 50 foot MIDI cable that seems to work OK. I believe there's an official limit to cable length but that doesn't mean you can't improve on it with low capacitance cable, or by putting a repeater in line. It's 5 meters for USB 2.0, which I agree won't get you from the live room to the control room in a big studio. The faster you go, the shorter the cable - USB 3.2 is only 3 meters, but they have active cables that extend that range. And if you need to go to the next building, you there are USB-to-Fiber converters.

 

As you said "if there's sufficient market for it", and there I think you have hit the nail on the head.

How many people need 16 bit velocity or other 16 bit CC's? A few synth users certainly, but elsewhere?

 

Yes. And how many people need 24-bit 192 kHz audio files? The ones who really need them got them, and the others "need" them because they're available. And I'm a cynic, too, and have a few years over which makes me a potentially bigger cynic. But if they went to the trouble of adding more features to MIDI, there's no reason not to explore them if you can find a use for them. And, yeah, it will take the hardware that provides what the new standard can support a while to come along. And like the early MIDI instruments, they will initially be pretty expensive - another filter.

 

And, to be brutally frank, I can't see anything in MIDI 2 that I yearn for.

 

But that's no reason to piss on it. On the other hand, availability (and non-availability = declared obsolescence) of software is the only reason why I haven't kept all of my computers on Windows XP. I'm finding software that I'd like to play with that requires Windows 10, so I'm out of luck until I decide (as I did with Windows 7) to bite the bullet. But in reality, I can't think of anything I really need my computer to do that I can't do with the computers that I have now. But others want more, so they can get it.

Link to comment
Share on other sites

But that's no reason to piss on it.

Is that what I was doing?

It certainly wasn't what I thought I was doing.

(Strange how one can be misunderstood, isn't it?)

 

Dr. Mike got it when he said that I made some fair points, which is what I intended.

 

Before I go on to answer some of the points you make, Mike, let me make an analogy.

I'm using an aircraft analogy because my first job, as a teenager, was as an electronics engineering apprentice at B.A.C. formerly Vickers Armstrong where I helped to build VC10's and BAC1-11's.

Cockpit and radio bay electronics. But machines of past decades.

 

You may have heard of a aircraft designer called J.R.Mitchell, he designed a world beating aircraft called the Supermarine S.6B which won the Schneider trophy in 1931.

This incredible (for its time) aircraft went on to become the iconic machine of its time, the '40s, the Supermarine Spitfire.

Still seen as the iconic shape of WWII today.

 

And what I hear you ask has this got to do with MIDI 2?

 

Well, this airframe would be nothing, absolutely nothing without the Rolls Royce Merlin engine that powered it.

1296 cubic inches capacity delivering initially 1100 horse power allowed this plane with the pilots who flew them successfully to defend the UK.

 

The point?

The airframe is brilliant but without the engine it would never have defeated the ME109, another brilliant design, simply the Spitfire would not have been up to the job.

 

MIDI 2 is an exceedingly fine protocol specification, but without the appropriate connectivity mechanisms, it becomes like a Spitfire with the engine of a Bugati Veyron.

Great engine, but not up to the job.

USB 3 is the Bugati Veyron, fantastic new networking technolgy, just not the right one for MIDI 2. Or not if you want fly.

 

Am I still, as you say, urinating on MIDI 2?

I hope not.

The point i'm trying to make is that USB, whilst offering some connectivity, does not give us the flexibility that the current DIN interface allows in certain more demanding environments.

Many professionals need that connectivity.

 

In another post, wihich will take a little while to prepare, I'll try to explain the difference between the way MIDI 1 DIN and MIDI over USB works and why it's considerably more complex with MIDI 2.

Okay?

 

JohnG.

 

Oh! And whilst I remember you said "But nobody ever said that MIDI was simple."

But I did, back in 1988 when I first started messing about with it.

It is simple, incredibly simple, that's why it's been SO successful.

Simple interface, simple command structure, designed to create excellent sounding music. BRILLIANT!

That's why it's lasted nearly forty years, and will go on maybe for another forty.

 

You should compare the MIDI command structure with ISDN Basic Rate Interface Delta channel signalling protocol and the 2B1Q line code of the PSTN line.

Now there's a specification to make even the most expert protocol analyst go white at the gills. Aaaaargh!

Good, in its time, before ADSL, for those who needed it. And still available, I believe, as a satellite data comms link via a BGAN terminal.

Akai EWI 4000s, Yamaha VL70m, Yamaha AN1x, Casio PX560, Yamaha MU1000XG+PLGs-DX,AN,VL.

 

Link to comment
Share on other sites

In another post, wihich will take a little while to prepare, I'll try to explain the difference between the way MIDI 1 DIN and MIDI over USB works and why it's considerably more complex with MIDI 2.

Okay?

Looking forward to it, John!

Dr. Mike Metlay (PhD in nuclear physics, golly gosh) :D

Musician, Author, Editor, Educator, Impresario, Online Radio Guy, Cut-Rate Polymath, and Kindly Pedant

Editor-in-Chief, Bjooks ~ Author of SYNTH GEMS 1

 

clicky!:  more about me ~ my radio station (and my fam) ~ my local tribe ~ my day job ~ my bookmy music

Link to comment
Share on other sites

Sorry for the delay, other things to deal with.

 

Here we go with some background.

 

Before I begin to describe how the data transmission mechanism works, I thought I'd give a quick overview of how a current MIDI controller keyboard works. It will help those not familiar, I believe, to understand the rest of the process.

Naturally, those who know can skip this section.

And, it's not meant to be all inclusive, just an overview.

______________________________

 

In a modern computer we will typically have a QWERTY keyboard attached, and the way this works is by an electronic mechanism scanning the keys very regularly to see if one has been pressed.

If one has, its identity is forwarded to the relevant program. Each key has a simple, spring loaded on/off switch.

 

A MIDI keyboard works in the same way, that is by being scanned (probably thousands of times a second), but each key has two switches (at least) underneath it.

One which triggers at the beginning of the down stroke, a second one whih triggers at the end of the down stroke.

 

Let's say that we have an 88 note controller and we've told it to transmit all notes on MIDI channel 1.

 

So, say we play a C major chord on our keyboard based on middle C.

As the scanning mechanism surveys the switches it will detect that, e.g. the C key has just begun its downward movement and the first switch has triggered.

(we may think we play all the notes simultaneously, but in terms of microsecond timing, we don't. Hand shape, length of fingers, rotated wrist, etc.)

 

So the scanning mechanism now knows a MIDI Note On has occured, it's on Channel 1, and it's a middle C or MIDI Note #60, but it doesn't yet know how strongly the key has been played.

It starts a note timer. From the timer it will derive how quickly the key travels from top to bottom.

 

Meanwhile scanning continues and G (top switch) is detected so it can do the same as above but G above middle C is MIDI Note #67. Start another note timer.

 

Scanning continues and E, top switch is detected, so MIDI Note #64 is put in a third location with the other information and a third note timer is started.

 

Scanning.

 

Middle C second switch occurs the MIDI Note Velocity can be derived.

We now have the entire contents of a MIDI message created so it can be sent to the output buffer for the UART (a piece of electronics) to begin its transmission through the MIDI DiN out socket.

 

Scanning.

As above but with E.

Scanning.

As above but with G.

 

So, we've now started to play our chord but we haven't released the notes yet.

 

Simply put, our MIDI Note On message looks like this:

Note On indicator, - 4 bits

Channel number, - 4 bits

Note number, - 7 bits

Velocity. - 7 bits

 

The scanning continues and the release of the second switch for each note will be detected, note by note, and then ditto for the first switch of each note.

These will generate MIDI Note Off messages for each key.

 

(More of this process, perhaps later. There's more to Note Offs than meets the eye.

____________________________

 

So a MIDI 'Note On' message arrives at the ouput buffer.

The UART sends a 'start bit' to wake the other end and establish accurate timing, and then sends, one by one, the bits in the first byte to the other end.

Then a stop bit, followed by a start bit for the second byte, the byte and a stop bit, then a third start bit, the last byte (bit by bit) and the final stop bit.

Job done.

 

The actual transmission is the waggling of an electrical signal up and down between pins four and five of the Out interface.

 

With this mechanism the transmission can start just as soon as the MIDI Note On message arrives in the output buffer, providing that the UART is not in the process of transmitting another message.

If there's one or more messages queueing, it's just put in the queue.

 

In this case it's highly likely that the three messages will be sent out just as soon as they reach the buffer.

Three three byte messages (that's thirty bits each (including start and stop bits)) at 31,250 bits/sec (that's a thousand messages per second) will take three thousands of a second, or 3 ms for all the Note.Ons.

(That's if I can still do mental arithmetic, and not make mistakes!

 

There's a clever wrinkle in the MIDI 1 spec. which says that if succeeding messages are the same type and channel, e.g. all Note On messages on channel 1, then theres no need to send the first byte, just the next two.

So our 9 bytes for 3 messages can be reduced to 7, thus reducing the overall transmission time for the second two messages. Clever eh?

 

And there's yet another clever wrinkle.

We can't have a Note On with a velocity of zero, it simply doesn't make sense.

So we can use Note On velocity zero to signify Note Off.

 

With our C chord it now means that the three Note Offs can be transmitted as Note Off C, E & G with velocity zero and we've saved another three transmission bytes.

 

Not clear? Try reading it again.

 

And that's it for part one.

In part two I'll look at the receive end of the transmission, or what happens at the MIDI In Din interface.

 

E&OE of course.

See ya later.

JohnG.

 

P.S. Please let me know if you spot any typos, etc.

Or GLARING errors.

Akai EWI 4000s, Yamaha VL70m, Yamaha AN1x, Casio PX560, Yamaha MU1000XG+PLGs-DX,AN,VL.

 

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Unfortunately, your content contains terms that we do not allow. Please edit your content to remove the highlighted words below.
Reply to this topic...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...