MIDI Association Feb 23rd news on MIDI 2.0

[JohnG]

7 hours ago, Toddskins said:

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.

Most devices I've worked with, from the Cheetah MS6 to date have  sets of tables held in EPROM with, usually, three bytes per note for the possible 128 MIDI notes.

Something of the detail can be seen in the MIDI 1.0 Detailed Specification (downloadable from the MMA as a pdf) starting on page 47 MIDI tuning. There it describes the Frequency Data Format and also the SysEx message(s) that can be sent to alter the tuning as a four byte message per note. This allows tuning as fine as 0.0061 of a Cent (a Cent=1/100th of a semitone.)

About halfway down page 49 is the spec. for the Single Note Tuning Change (Real-Time). So you can alter your sound module's tuning by simply sending one or more SysEx messages.

There were some extensions to the MIDI Tuning Standards released in '99 which can be seen in RP-020, CA-020 and CA-021 also made avaialble on the MMA site in the specs section.

So we don't need MIDI 2.0 for highly accurate MIDI tuning. (Assuming, in one's opinion, that 0.0061 of a cent is accurate enough.) ;-)

As long as the device one is using actually supports such accuracy.

Most modern MIDI hardware (that I've used) doesn't support the full accuracy of the MIDI 1.0 specification. For instance, there's not one keyboard which I've tested, that gives a pitch bend accuracy greater than 10 bits, despite the 14 bits allowable by MIDI 1.0. Many only offer only 7 or 8 bit accuracy. :-(

I have several MIDI modules too, 2 x Yamaha MU1000 with VL, AN,  DX and PF cards in them, as well as an ancient Roland SC-8850, all gathering dust.

[craigb]

6 hours ago, JohnG said:

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.

 

Noice!  I love me some good geek speak!  ?

I used to be a DEC SME contracting to Verizon Wireless.  I was part of the 1xRTT project (the first generation of Internet on mobile phones) and I wrote the fuzzy logic to match call records as they are handed off from cell tower to cell tower.  I also briefly worked on the guidance system for the Tomahawk Cruise Missile back in the early 80's.  Add in almost 20+ years of writing or rewriting accounting and invoicing software, and the above were probably the funnest projects I had...

 

Edited February 3, 2020 by craigb

[JohnG]

Good to see another geek on the forum.

I started as an electronics apprentice in 19... (a long time ago) and this was what I worked on for a while:

Later on (still in the sixties) I moved over into the computer department of B.A.C. and this was what I became familiar with for a while:

These are both pictures of the RCA301 machine, rebadged in the UK as an ICT1500.

Later on I worked on early Ethernet and Ungermann Bass equipment, then ISDN both Basic and Primary rate before becoming 'involved' with digital telephony and the Mitel digital exchanges. Then BT and Basic and Primary rate (ISDN2 & ISDN30) then DSL and DSLAMs and so on,

then with a few RBOCs in the US and Deutsche Telekom and the Scandinavian Telcos, etc. Finally satellite data comms with InMarSat. The last one wore me out so much I decided to take early retirement , moved to Germany for ten years, where I taught English to business people as a hobby (hence being a grammar nasty), now back in the UK (with my German wife) and doing as little as possible in the greatest amount of time.

 

Edited February 4, 2020 by JohnG

[craigb]

Definitely got me beat!  I didn't start programming until 1970 (HOLY CRAP!!!  I JUST REALIZED THAT'S 50(!) YEARS AGO!!!). ?

I was never into the hardware side of things (until about four years ago) and I was that 19 year-old hacker working for a defense contractor back during the Cold War. ?

I did wire this up recently though (orange were for employees, red for tenants).

[JohnG]

Ah! A good old structured cabling cabinet.

I had to put together a course for British Telecom pre-sales and support employees teaching them how this worked for telephony, ISDN and Ethernet back in the mid 90's. That and modems, ISDN TAs, a bit of early DSL, plus a little html. Ended up giving up to 40 four days courses every year from '96 to some time in Y2K.

I beat you by only a couple of years. Started assembler programming in '66 then COBOL in '68. After that it was all sorts of machine code as I worked on FEPs and other mini computers and early servers. After a bit of training with Intel, I got involved with DOS, early OS/2 and Windoze, device drivers and stuff like that.

Twenty years with a major UK computer manufacturer, now a part of Fujitsu, doing early real-time and on-line systems. Left in '92. Re-employed as a contractor after that for another ten years or so.

All intensely geeky stuff majoring in data communications protocols, etc. My wife reckons that if you could look inside my head it would all be binary and octal and Hexadecimal. ... All neatly structured into TCP/IP packets!

Strangely, I never did get involved with mobile data comms.

[Finnbogi Ragnar Ragnarsson]

On 2/3/2020 at 1:13 AM, User 905133 said:

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. 

 

I chose to have only a short comment as it was getting more and more out of a "deals" discussion, but previous posts were about the communication, although I could say a lot about the opportunities ahead.