[sdiy] MIDI 2.0

[Brian Willoughby]

On Jan 28, 2020, at 3:52 AM, paula at synth.net wrote:
> My worry is that it needs USB currently (they've not agreed any other transport layers yet), and that a lot of smaller manufacturers will struggle to implement this.

That was probably true a decade ago, but things are pretty good now. There are a lot more cheap USB chips capable of Full Speed, and now they all have the basic USB firmware available for free. What’s missing is the USB-MIDI part, but that’s simple enough for any decent firmware developer to create on top of the basic USB support that is provided. I say this as someone who has developed USB-MIDI firmware for small, three-person companies, as early as 2008. To be honest, I hate the limitations of USB-MIDI, but it works for what it is.


> Alongside that, you'll need to be registered with the MMA to get an ID so the CI portion will work properly.

It’s always been true that you need to pay to register with the MMA to get an ID for the System Exclusive portion. Of course, it’s probably a little easier to get by without SysEx. If you can get by without CI, then you probably don’t need MIDI 2.0 anyway. I think it’s cheaper to get an ID from the MMA than it is to get a VID from the USB organization, so there’s that to consider. We’re lucky that the MMA charges money to somebody, otherwise they would have disappeared long ago, and MIDI might not have made it to its 30th birthday, much less it’s 37th.


> Also a lot of the new features seem to have already been covered by MPE.

That’s not saying very much. Every feature covered by MPE already existed in MIDI 1.0
MPE is really just a branding effort by ROLI and other new companies that didn’t want to bother learning what Dave Smith and folks enabled with the original MIDI. I mean, the ancient 1984 Oberheim Matrix-12 is capable of MPE, or at least it can handle three dimensions of expression per note without a firmware upgrade. The whole MPE effort would have been better spent summarizing the existing MIDI capabilities in a way that a term like “MPE” could be used as a quick MIDI Implementation Chart checkbox.


> I get that in theory this is better than 1.0, by a long way, however I worry for the smaller manufacturers that this may push things too far.

On this, I agree. I started to study the MIDI 2.0 spec until I got to the JSON part of it. At that point I realized that it was a mess.

Brian Willoughby


> Paula
> 
> On 2020-01-28 11:35, Neil Johnson wrote:
>> (try again...)
>> I see MID 2.0 is now out:
>> https://www.midi.org/articles-old/details-about-midi-2-0-midi-ci-profiles-and-property-exchange
>> Neil
>> On Tue, 28 Jan 2020 at 11:34, Neil Johnson <neil.johnson71 at gmail.com> wrote:
>>> Hi all,
>>> I see MIDI 2.0 is now out: