MIDI.ORG recently released information about MIDI 2.0 and how the additional capability provided by MIDI 2.0 will enhance the creation and production of music.

Luckily, all the information has been consolidated into one webpage with embedded links for those desiring all the details. I've decided to call the page +++ MIDI 2.0 Explained +++

Enjoy!

Thanks again, Jim. I’ve been following this.

The changes seem incremental to me, but two features really stand out: 256 channels, and tighter timing. Reverse compatibility is supported, which is great.

I’m still puzzled MIDI 2.0 will not include chord symbols. That seems a real missed opportunity.

Especially since so many music apps now include their own proprietary support of chords. While that's fine, having chordal support in a MIDI file will allow for easier movement of files between applications, both notation and DAW.

John, right.

You said “will allow”. Did you mean “would allow”? As far as I can tell, the new standard doesn’t plan to have chord data, unless I missed something.

"Would" would be the grammatically correct term.

Thanks. Just trying to find out if you knew something about this I didn't.

Depending on what you expect MIDI "chords" to do, include or be it may not be that easy given the different manufactures involved and considerations like:

how many times on this forum alone folks have asked "why does this chord not always play the 5th, or dominate 7th, or..." or why is sus2, add2, 9th, ... not all the same or worse ARE the same, etc.

and should teh MIDI spec send only chord notes (or should receiving device only play chord notes) that only jazz players would use on large complex chords (e.g., 13's with other alterations); only chord notes that a 10 fingered piano player would player, or simple rocker interpretations; or ...

just sayin'

Larry

Good questions, Larry, for something that is far from an exact science.

For me, I would want MIDI 2.0 to realize the promise (not always implemented well) of Music XML, namely to transfer existing chord info. Anything else is gravy. But to ignore this function entirely seems a terrible oversight.

It may be I am dense and don't really get how you and John visualize it.

Do you mean read a MIDI 2.x file and have something like Musescore, Finale, etc. read it and make a score or an XML file (i.e., translate for notation) out of that .MID file with chords as recorded and "annotated" by the sequencer?

Maybe they looked at it and decided anything like that also transferred in real-time (via a MIDI 2. stream) to our I7's, my Montage, or any pre-MIDI 2.0 gear - might make them choke/freeze, reset, .....

Just playing devils advocate

Or the other way around. For example, save a BIAB file as MIDI 2.0 and have the chords in the song included in the MIDI.

No, I'm saying that the "standard" should support at least the ability to enter a text expresion for a chord (say, Fm7), and every new standard MIDI compliant app be able to read it as such. What the app does with it is up to the app, but my intention is clear that at that point in the song, the chord is supposed to be Fm7. Some apps implement this in non standard ways by using unassigned controllers; I'm suggesting it be standardized. I'm not saying that the MIDI standard needs to understand and differentiate between unambiguous chords (is it an Fm7 or an Ab/F), but that if I enter an Fm7, the MIDI file will present that as my intention to any other app.

Currently, there is no chord support at all in the MIDI standard.

If there were, I could for example load a MIDI into a keyboard arranger and have it automatically play the accompaniment. I could load A MIDI file with just a melody line and chord text into a notation program and create the lead sheet, and more. I can't do that today.

Support for chords just seems to be something that should be basic for any music standard.

OK but that brings me back to what happens if you try and PLAY that MIDI 2 .mid file?

But the more I think about it all of the NEW MIDI 2 protocol extensions had to already consider what it would do to oldest MIDI 1 gear - so most new information sent out is probably in byes that MIDI 1 already ignores - so with respect to my last reply, as Emily Litella would say, "Never mind!"

The MIDI standard has always supported the idea that a receiving device or program will ignore any midi message that the device or program does not understand. MIDI 2.0 follows that tradition.

MIDI 1 is a serial protocol where information only travels one way from master to receiever. The master transmits data and the other devices within the chain receive data. Devices within the chain can add more data but the master controls the data chain.

What has changed in MIDI 2.0 is that devices can converse and data flow is bi-directional. The conversation starts with a question, "do you want to converse in MIDI 2 or MIDI 1?" When devices answer MIDI 2.0, data is formatted in both MIDI 1.0 and MIDI 2.0 formats so older devices will continue to be able to receive data. A MIDI 1.0 device will ignore any MIDI 2.0 data it receives but will echo (re-transmit) all data through it's midi out or through port.